Seawater for five shipboard incubation experiments was collected using a surface towfish (Mellett and Buck 2020) at ~2 meters (m) depth on the RSS Discovery between 5 May 2021 and 26 May 2021. All experimental setup and sampling was conducted in a trace metal clean (TMC) lab using TMC methods. Water was prefiltered through a 150-micrometer (µm) mesh to remove large grazers. For all incubations, 20-liter (L) acid-cleaned carboys were filled in sets based on the replicate number (e.g. replicate 1 carboys for all treatments were filled in succession). For LT1, individual carboys were completely filled prior to filling the next. For all other experiments carboys were filled "round robin", filling each 1/3 at a time until all replicates in the set were full. Triplicate control and treatment carboys were amended for each experiment as indicated in Table 1 (see Supplemental File "EXPORTS_2021_dy131_table1.pdf"). In addition, triplicate T0 carboys were also collected. T0 carboys were sampled immediately after the experimental start. Control and treatment carboys were sampled at the end of each incubation. T0 carboys are shared between LT2 and ST2 experiments.

Short-term incubations were carried out for 24 (ST1A and ST2) and 42 hours (ST1B). Long-term incubations were carried out for 6 (LT1) and 5 (LT2) days (Table 1). Short-term nutrient amendment treatments were selected to alleviate potential in situ nutrient stress. The +Si and +N treatments were amended to final concentrations of 20 micromolar (µM) silicate and nitrate respectively, in addition to in situ nutrient concentrations. The +Fe treatment was amended to a final concentration of 0.005 µM ⁵⁷Fe. The long-term experiments were amended with all key macro and trace nutrients (N, P, Si, Fe) except one, with the aim of inducing a specified nutrient stress. For LT1 and LT2, we refer to the +20 μM nitrate +1.25 μM phosphate +20 μM silicic acid treatment as "AllButFe" and the +20 μM nitrate +1.25 μM phosphate +0.005 µM ⁵⁷Fe treatment as "AllButSi".

Nutrient samples were filtered through 0.2 μm polycarbonate filters and frozen at -20 degrees Celsius (°C). Samples for biogenic silica concentrations were size fractionated by serial filtration through 5 μm and 0.6 μm polycarbonate filters Filters were stored frozen at -20°C. Particulate organic carbon and nitrogen were measured on samples from experiments examining the effect of added Fe and Si on carbon fixation. These samples were filtered through precombusted GFF filters placed in glass scintillation vials and frozen at -20°C.

Biogenic silica concentrations were determined by NaOH digestion followed by colorimetric analysis of the resulting dissolved Si. Particulate organic carbon and nitrogen samples were analyzed by Dumas combustion. Nutrient concentrations were determined by flow injection using a Lachat Instruments QuikChem 8500 Series 2 analyzer.

Metatranscriptomes were sequenced for experiments LT1 and ST1B. Metatranscriptomic samples were collected on 5 μm 47-millimeter (mm) polyester (PETE) membrane filters using a peristaltic pump. Filters were then preserved in 2 milliliters (mL) of RLT buffer (RNeasy Mini Kit, Qiagen, Germantown, MD, USA), flash frozen in liquid nitrogen, and stored at -80°C until RNA extraction. PolyA selection was performed on sample libraries to enrich for eukaryotic sequences and libraries were sequenced on an Illumina NovaSeq 6000 platform.

For more information on nutrient, particulate, and chlorophyll measurements see the Protocol documents attached as Supplemental Files and https://msi.ucsb.edu/facilities-services/analytical-lab/services.