Methodology:

Two sediment traps (McLane PARFLUX, shallow=400 m, deep=2000 m) were deployed on a long term mooring station (P3-Scotia Sea Open Ocean Observatory (SCOOBIES: https://www.bas.ac.uk/project/scoobies/ ), located at: -52.8036, -40.1593 at 3748 m water depth. Sediment traps were deployed in January 2018 aboard research cruise JR17002 and recovered in January 2019 aboard research cruise DY098. Each sediment trap collected 14 samples, open for periods of 7-31 days, which, on return to the laboratory were split into multiple aliquots for subsequent analysis. Splits were taken for particulate organic carbon (POC), particulate nitrogen (PN), biogenic silica (BSi, SiO2), as well as stable isotopes delta13CPOC, delta15NPN and delta 30SiBSi. For full details of laboratory analyses of these splits, see Belcher et al. in review.

Data collection:

Sample collection: McLane PARFLUX Sediment Trap

Sample splitting: McLane Sample Splitter

POC and PN: CE-440 Elemental analyser (Exeter Analytical.285 Inc)

delta13Cpoc and delta15NPN: Thermo Finnigan Delta-Plus Advantage mass spectrometer

BSi: Hach DR3900 spectrophotometer

delta30SiBSi.: Finnigan Neptune Plus High Resolution MC-ICP-MS (Thermo Fisher Scientific)

Data quality:

We note that the mooring itself experienced periods of higher flow in June 2018 and late August/September 2018 and thus particle collection may be biased at this time.

For each flux measurement, we present the maximum and minimum measurement resulting from analyses of 2 replicate splits (for a couple of samples, 3 replicates were measured, and again the maximum and minimum are given here).

Reproducibility of delta13Cpoc and delta15NPN is 0.2 per mille based on repeat measurements of analytical standards.

For delta 30Si, solutions were measured in replicate (n = 2-3) alongside continuous measurement of reference standards Diatomite and LMG-08 to ensure reproducibility and to monitor data quality. Measurements of Diatomite and LMG-08 yielded delta 30Si of +1.23 per mille (SD ± 0.03, n=18) and -3.40 per mille (SD ± 0.05, n=5) respectively, which agreed with published values. Typical reproducibility between the sample pseudo replicates was 0.04 per mille (1 x SD). Not all samples were able to be successfully analysed for delta 30SiBSi. and these are denoted by ND (=No Data).