Dataset: Nitrogen isotopic ratios of individual amino acids extracted from size-fractionated particulate organic matter aboard the R/V Kilo Moana in February (cruise KM1407) and September (cruise KM1418) 2014 at Station ALOHA

Sampling was conducted onboard R/V Kilo Moana at Station ALOHA (22.75°, ‑158.0°) on February 19-28, 2014 (KM1407) and August 29-September 11, 2014 (KM1418). Five to six in situ pumps (WTS-LV standard, McLane Research Laboratories, Inc., East Falmouth, MA) were deployed with sequential 142 mm-diameter filters mounted on tiered mini-MULVFS filter holders (Bishop et al. 2012) to capture submicron (0.2-0.7 µm; sterile-packed polyethersulfone), small (0.7-53 or 1-53 µm; pre-combusted glass or quartz fiber filters), and large (>53 µm; acid- and methanol-rinsed nylon mesh) particles. Pumps were deployed at depths from 25 m to 1205 m, pumping between ~400-2000 L per deployment at rates of 3-8 L min^-1. Filters used in this study were stored in combusted foil and frozen at ‑80°C until further processing on land. Large particles captured on Nitex mesh were thawed briefly and resuspended in 0.2 μm-filtered seawater; resuspended particles were gently vacuum-filtered onto a 47-mm glass fiber filter (GF/F; 0.7 μm pore size), re-frozen, and lyophilized. Dried filters were inspected under a dissecting microscope to remove whole swimmers and loose Nitex fibers, both of which were rare, and to make qualitative observations such as presence of Trichodesmium and carbonate skeletal fragments. All filter samples were subsampled, lyophilized overnight, and quantitatively split by weight for bulk analysis, compound-specific isotope analysis of amino acids (CSIA-AA), and amino acid enantiomer analysis.

Bulk N concentrations and δ¹⁵N values were determined using a Costech Elemental Analyzer interfaced to a Thermo Delta Plus XP isotope ratio mass spectrometer (IRMS) using standard methods in the Popp Laboratory (University of Hawaii).

Amino acid preparation generally followed the methods of Hannides et al. (2009) and references therein. Filter solids were removed by filtering sample hydrolysate through combusted glass wool and 0.2 μm polyethersulfone disc filters. Filter pieces were rinsed 3-5 times with 0.01 N HCl during this process, and the sample was fully extracted by squeezing the filter pieces in a 5 or 10 mL glass syringe plugged with glass wool. Amino acids were purified using cation exchange resin (50W-X8, 100-200 mesh, 1 mL bed volume), eluting in 2 N ammonium hydroxide, followed by reprotonation (0.2 N HCl, 110°C, 5 minutes) and derivatization to trifluoroacetyl/isopropyl esters in a two-step procedure (esterification of carboxyl groups, 4:1 isopropanol/acetyl chloride, 110°C, 1 hour; trifluoroacetylation of amino groups, 3:1 dichloromethane/trifluoroacetic anhydride, 100°C, 15 minutes). Excess salts were removed by liquid-liquid extraction (chloroform/phosphate buffer), and the trifluoroacetylation step was repeated. Samples were stored in airtight vials at ‑20°C in between procedural steps. Immediately before analysis sample aliquots were dried under ultra-purity N₂ gas and redissolved in 6-50 μL of ethyl acetate, depending on concentrations of AAs estimated from bulk N analysis. All glassware was pre-cleaned in 1.2 N HCl and combusted at 500°C overnight before use.

AA δ¹⁵N analysis was conducted using a Thermo Trace gas chromatograph (GC) via a split/splitless inlet operated in splitless mode and equipped with a BPX5 column (SGE, Inc.). Individual amino acids were converted to N₂ gas via a two-stage Thermo GC-C-III combustion (980°C) and reduction (650°C) interface. N₂ sample gas was purified over a liquid nitrogen cold loop and directed into a Delta-V Plus IRMS.

δ¹⁵N values (‰ vs. AIR) were calculated by the Thermo Isodat 3.0 software, using reference gas peaks introduced into each run from an externally calibrated N₂ gas cylinder and/or using co-injected reference peaks of isotopically known norleucine and aminoadipic acid standards derivatized identically to samples. Instrument accuracy was determined by injecting a mixture of amino acids of known δ¹⁵N values every fourth injection and derivatized identically to samples. Individual AA δ¹⁵N values were calculated as the average across 2-3 analytical replicates, and 1σ uncertainty was calculated as the standard deviation of these analytical replicates. For some small particle samples, the reported δ¹⁵N values represent the average and standard deviation across sampling replicates (same depth and near/identical size fraction). These are indicated by data with values reported for the parameter ISO_DateTime_UTC_rep. In some cases, results were averaged for a 0.7-53 um size fraction and a 1-53 um size fraction. In these cases, the value for parameter Size_Fraction_Min is reported as 0.85.

AA concentrations were determined based on IRMS peak area response of individual AA standards and the amount of sample injected. Estimated seawater concentrations of nitrogen deriving from amino acids in particles (nmol N L^-1) were determined as the sum of concentrations of all measured AAs.