3.3 Incubation design and sampling
Incubation regeneration experiments were carried out in the subantarctic Southern Ocean south of Tasmania (IN2018_V02) in acid-cleaned 20 L LDPE cubitainers with Teflon-lined caps. Three regeneration experiments were conducted, with differences in incubation temperature and the particle collection depth (supplemental material). Each experiment consisted of two treatments, each in triplicate. Cubitainers were filled with filtered (0.2 µm) surface (~5 m) seawater collected using a trace metal-clean tow fish, and aliquots were collected for initial (T0) nutrients, [Cr] and δ53Cr before particle inoculation. Filled cubitainers were stored in the dark at 1° C until inoculating with particles (within 8 hours of filling). Ambient suspended particle samples from the oxic water column were collected onto acid-cleaned 1 µm polycarbonate filters using in-situ pumps (McLane). Particles were resuspended from these filters, forming a particle-rich slurry that was added to the individual cubitainers to result in particulate concentrations approximately 2-4 fold above natural levels. Because only resuspended particles were added to the incubaitons, the high Cr blanks associated with many filter materials (e.g. Scheiderich et al., 2015) will not impact the incubations.
Particle-inoculated cubitainers were incubated in the dark at either 4° C (ambient) or 1° C (cold treatment) for five days, with three subsampling time points. For subsampling, the cubitainers were transported to a Class 100 clean van, gently mixed, and subsampled into acid-cleaned LDPE bottles (for trace metals) or polycarbonate tubes (for macronutrients). Subsamples for trace metals were transported to a laminar flow hood and filtered through acid-cleaned 0.4 µm polycarbonate filters using a Teflon vacuum filtration apparatus (Savillex). Loss of Cr to container walls has been shown to be minimal over short timeframes (e.g. Semeniuk et al., 2016) and was not considered here. Chemically-labile particulate metal and P concentrations were determined with magnetic-sector ICPMS following solubilisation with a heated mixture of reductant (hydroxylamine hydrochloride) and weak acid (acetic acid) designed to dissolve biogenic material and Mn and Fe oxides (Berger et al. 2008). Analytical procedures followed those previously described in Rauschenberg and Twining (2015).