Culturing

The marine diatom Thalassiosira weissﬂogii Grunow (strain CCMP 1010) was cultured in sterile filtered natural North Sea water (Schleswig-Holstein, Germany) or Baltic Sea water (Schleswig-Holstein, Germany). The medium was enriched with f/4 concentrations of macro- and micronutrients (nitrate, phosphate, silicic acid, trace metal mixture, vitamin mixture (Guillard and Ryther, 1962). All experiments were performed in sterile 2.1 L Schott Duran glass bottles. These bottles were made of borosilicate glass (filters UV radiation <310 nm) except for the quartz glass bottles (pure silica without UV radiation filter) used in the UV experiment. The cultures were either incubated in climate chambers with 400 –700 nm radiation or 10 cm below water level at low tide in Kiel Fjord in May 2011. Water temperature and light irradiance data were obtained from the weather station maintained by the GEOMAR institute in Kiel, Germany. Growth conditions for the various treatments, i.e. salinity, pH, temperature, and irradiance are given in Table 1. pH values (reported on free scale) were measured with separate glass and reference electrodes (Metrohm) and calculated with equation 3 from DOE 2007 chapter 6b (Dickson et al., 2007) as described in (Bach et al., 2012). Cultures were inoculated with densities of 20 cells ml-1. Cell densities and equivalent spherical diameters were determined with a Coulter Counter (Beckman Coulter) at the beginning and the end of the experiment, respectively. When incubations ended, cells were filtered on 47 mm diameter, 5 um mesh size Nucleopore Track-Etch Membrane filters (Whatman) and frozen at -18 deg C immediately after filtration.

Sediment sampling

Sediment samples were retrieved from a 14.97 m core, station M772-003-2, collected November 26, 2008 by the Meteor cruise M77 at 271 m water depth within the main upwelling area off Peru (15 deg 06.21´S, 75 deg 41.28´W). The Peruvian ocean margin is characterized by a high particle flux and a well-defined oxygen minimum zone. At the time of sampling, the O2 concentration at the seafloor was measured to 1.1 uM, the salinity to 34.9 psu and the temperature to 12.2 deg C.

Prior to analysis, sediment samples were pre-treated with an acid-alkali-acid cleaning with HCl and NaOH (Grootes et al., 2004).)

Analyses

Both diatom and sediment samples were freeze dried prior to isotopic analysis. To prepare aliquots for derivatization of amino acids, we used 3-4 mg of diatoms and 100-150 mg of sediments. The samples were transferred to Pyrex culture tubes (13 x 100 mm), flushed with N2 gas, sealed, and hydrolysed in 1 ml 6N HCl at 110 deg C for 20 h. After hydrolysis, lipophilic compounds were removed by vortexing with 2 ml n-hexane/DCM (6:5, v/v) for 30 sec. The aqueous phase was subsequently transferred through disposable glass pipettes lined with glass wool into 4 ml dram vials. Samples were evaporated to dryness under a stream of N2 gas for 30 min at 110 deg C before being stored at 18 deg C until required for analysis. The derivatisation procedure was modified from Corr et al. (2007) as described by Larsen et al. (2013). In short, the dried samples were methylated with acidified methanol and subsequently acetylated with a mixture of acetic anhydride, triethylamine, and acetone, forming N-acetyl methyl ester derivatives. As a precautionary measure to reduce the oxidation of amino acids, we flushed and sealed reaction vials with N2 gas prior to methylation and acetylation. Another modification from Corr et al. (2007) was that icebaths were substituted with solid aluminum blocks at room temperature. We used known d13C values of pure amino acids prepared and analyzed under the same conditions as the samples to calculate correction factors specific to each amino acid to account for carbon addition and fractionation during derivatization. The derivatised AAs were dissolved in 250 ul ethyl acetate and stored at 18 deg C until required for analysis.

Amino acid d13C values were obtained from Leibniz-Laboratory for Radiometric Dating and Stable Isotope Research in Kiel. We injected the AA derivatives into a PTV injector held at 250 deg C for 4 min. before GC separation on an Agilent 6890N GC. Diatom samples were separated on an Rtx-200 column (60m x 0.32mm x 0.25um, Fig. S1) and sediment samples on a Thermo Trace GOLD TG-200MS GC column (60m x 0.32mm x 0.25um). For both GC columns, the oven temperature of the GC was started at 50 deg C and heated at 15 deg C min-1 to 140 deg C, followed by 3 deg C min-1 to 152 deg C and held for 4 min, then 10 deg C min-1 to 245 deg C and held for 10 min, and finally 5 deg C min-1 to 290 deg C and held for 5 min. The GC was interfaced with a MAT 253 isotope ratio mass spectrometer (IRMS) via a GC-III combustion (C) interface (Thermo-Finnigan Corporation). We obtained consistently good chromatography for alanine (Ala), valine (Val), leucine (Leu), isoleucine (Ile), asparagine/aspartate (Asx), threonine (Thr), methionine (Met), glutamine/glutamate (Glx), phenylalanine (Phe), tyrosine (Tyr), lysine (Lys), and arginine (Arg) with the exception that Asx and Thr partially coeluted with the Rtx-200 column. Serine (Ser) and proline (Pro) coeluted on both columns. The average reproducibility for the norleucine internal standard was ± 0.4‰ (n = 3 for each sample), and the reproducibility of amino acid standards ranged from ± 0.1‰ for Phe to ± 0.6‰ for Thr (n = 3).

Amino acid composition of the diatom samples was determined with the derivative samples used for d13CAA analysis. The amino acids were separated on an Rxi-35SIL MS column (30m x 0.32mm x 0.25um) with an Agilent 6890 N GC with a flame ionization detector. With this column we obtained good chromatography for Ala, Asx, Glx, Gly, Ser, Tyr, Arg, Ile, Leu, Lys, Met, Phe, Thr, and Val. For quantification, we used internal references consisting of pure amino acids (Alfa Aesar, Karlsruhe, Germany). The composition of the amino acids are shown in Table 3 according to the following biosynthetic families: Pyruvate (Ala, Leu, Val), Oxaloacetate (Asx, Ile, Lys, Met, Thr), α-ketoglutarate (Arg, Glx), 3-phosphoglycerate (Gly, Ser), and Shikimate (Phe, Tyr).

Bulk 13C, %C, 15N and %N values of the diatom samples were determined at the UC Davis Stable Isotope Facility using a PDZ Europa ANCA-GSL elemental analyzer interfaced to a PDZ Europa 20-20 isotope ratio mass spectrometer (Sercon Ltd., Cheshire, UK). The dry weight of the samples ranged between 1.5 and 2.5 mg. During analysis, samples were interspersed with several replicates of at least three different laboratory standards. These laboratory standards, which were selected to be compositionally similar to the samples being analyzed, had previously calibrated against NIST Standard Reference Materials (IAEA-N1, IAEA-N2, IAEA-N3, USGS-40, and USGS-41). A sample’s preliminary isotope ratio was measured relative to reference gases analyzed with each sample. These preliminary values were finalized by correcting the values for the entire batch based on the known values of the included laboratory standards. The long term standard deviation is 0.2‰ for 13C and 0.3‰ for 15N.