Sampling Sites:
Plankton was collected between 2007 and 2015 adjacent to shore over the continental shelves in the West Antarctic Peninsula (WAP), Amundsen Sea, and Ross Sea, as well as from open water off the continental shelf. Ninety-four discrete samples, comprising a variety of zooplankton taxa (average of four taxa per station), were collected at 34 stations over this region, including: ten offshore stations, ten nearshore stations off the WAP, seven stations in the Amundsen Sea, and seven stations in the Ross Sea. Details of the cruise durations are described in Brault et al. (in press).

Sample Collection:
Phytoplankton samples from the PAL-LTER surveys were collected using an 80 µm ring net towed through the upper water column (≤ 50 m depth) for ~ 30 minutes. Samples was rinsed into a pre-cleaned plastic tub, re-concentrated by sieving through a 25 µm mesh, and then frozen at -80 °C. A sub-sample of each tow was examined under a compound microscope to determine the dominant species (diatoms in all cases); microzooplankton were removed manually. Phytoplankton were collected during the Oden cruises of 2007/08 and 2010/11 via vertical tows from depths of ~ 20 m with a 30 µm ring net. Samples were similarly re-concentrated and frozen at -80 °C. The 2010/11 samples were dominated by Phaeocystis antarctica and once again any microzooplankton were discarded manually. The 2007/08 samples were not evaluated under a microscope to identify the dominant phytoplankton species.

Krill obtained during the PAL-LTER sampling, mixed zooplankton samples collected during the Oden cruise in 2010/11, and one sample of Clione limacina from a 2007/08 Oden cruise were derived from oblique tows (700 µm square-frame net) in 120 m and 400 m water depth for PAL-LTER and Oden cruises, respectively. Samples were transferred from the cod end into pre-cleaned buckets, re-concentrated by sieving through 700 µm mesh (retaining the retentate), and frozen at -80 °C. Samples were identified to the lowest taxonomic group possible prior to freezing. Zooplankton samples from the L. M. Gould cruise were obtained with open oblique hauls of a 505 µm mesh net from ~150 m to the surface using a 1.8 m Isaacs-Kidd midwater trawl. Samples were filtered through a 505 µm mesh sieve, sorted by species, and frozen at -20 °C. Mixed phytoplankton and zooplankton samples from the 2011/2012 cruise aboard the RV Nathaniel B. Palmer were collected with 200 µm bongo net tows in the upper water column (0-200 m). The samples were stored in a 4 % formaldehyde-seawater mixture at 4 °C.

Taxonomic Groups:
All phytoplankton samples were treated together as "phytoplankton". Zooplankton taxonomic categories from the Ross and Amundsen Seas were (1) copepods, (2) gammarid and hyperiid amphipods, (3) euphausiids (larval, juvenile, adult), (4) Salpa thompsoni, and (5) pteropods Clione limacina (naked) and Limacina helicina (shelled). The WAP and Drake Passage samples consisted of euphausiid species E. superba, E. crystallorophias, E. frigida, E. triacantha and Thysanoessa macrura, hyperiid amphipod species Themisto gaudichaudii, Vibilia antarctica and Primno macropa, Salpa thompsoni and the pteropod Spongiobranchia australis (naked).

Sample Preparation:
Samples were kept frozen until laboratory preparation, except for the formaldehyde-preserved plankton samples. Phytoplankton collected on the Oden cruises and phytoplankton and krill from the PAL-LTER cruises were freeze-dried at the Virginia Institute of Marine Science (VIMS, Gloucester Point, VA) with a Labconco Freezone Plus 6 at -80 ˚C for ~ 72 hours. Krill were homogenized with a Virtis “45” tissue homogenizer (Virtis Co., Inc.) before freeze-drying. Phytoplankton samples were manually homogenized after freeze-drying. Zooplankton from the Oden and L. M. Gould cruises, sorted by taxon at each station were freeze-dried at the University of California, Santa Cruz (UCSC) using a Labconco Freeze Dry System (Lyph Lock 4.5) at -40 ˚C for ~ 48 hours and then manually homogenized. All freeze-dried phytoplankton and zooplankton samples were stored in a dessicator after drying.

All zooplankton samples were lipid-extracted, except for the Antarctic shelled-pteropods. Carbonate shells of the pteropod Limacina helicina were acidified and decarbonated with a 10% HCl solution. After HCl treatment, the samples were neutralized with Milli-Q water (Thermo Fisher Scientific, Inc.) and freeze-dried in the UCSC Labconco Freeze Dry System as described above. These samples were not lipid-extracted due to sample size limitations but are considered lipid-poor. The PAL-LTER krill were lipid-extracted at the VIMS over three days using a chloroform:methanol (1:2; v:v) mixture via Soxhlet extraction. After lipid extraction, samples were dried and frozen at -80 ˚C until stable isotope analysis. A portion of each zooplankton sample from the Oden and L. M. Gould cruises was lipid-extracted via Accelerated Solvent Extraction (1500 psi; 60 ˚C; 3 cycles) with petroleum ether, according to a lab-established protocol at the UC Santa Cruz. For these zooplankton samples, δ13C values were obtained from lipid-extracted material and δ15N values were obtained from the non-extracted material.

To remove formaldehyde-seawater solution from the Ross Sea plankton samples, samples were transferred to 50 ml BD Falcon centrifuge tubes, centrifuged (15 min, 10,000 rpm), and decanted. The pellet was rinsed with Milli-Q water and centrifuged (15 min, 10,000 rpm) three times, discarding the supernatant between rinses. Samples were then transferred to 10 ml borosilicate vials and dried at 60 ˚C.

Isotopic Analysis:
For δ15N and δ13C analyses, ~ 1 mg of samples were weighed into tin cups (Costech, 3×5 mm) for elemental analysis-isotope ratio mass spectrometry (EA-IRMS). The PAL-LTER krill were analyzed at VIMS on a Costech ECS 4010 CHNS-O Elemental Analyzer (EA) (Costech Analytical Technologies, Inc.) coupled to a Delta V Advantage Isotope Ratio Mass Spectrometer (IRMS) with a Conflo IV Interface (Thermo Electron North America, LLC). The δ15N and δ13C values were referenced to AIR and V-PDB standards, respectively. Blanks and international standards (USGS 40 [L-glutamic acid with δ15N and δ13C values of -4.5 ‰ and -26.4 ‰, respectively] and USGS 41 [enriched L-glutamic acid with δ15N and δ13C values of 47.6 ‰ and 37.6 ‰, correspondingly] ) were analyzed on the EA-IRMS after every ten samples (standard deviations were < 0.1 ‰ for both δ15N and δ13C, and C/N deviations were <0.05).

All other phytoplankton and zooplankton samples were analyzed at the Stable Isotope Lab at UC Santa Cruz using a Carlo Erba EA 1108 EA coupled to a Thermo-Finnigan DeltaPlus XP IRMS referenced to AIR and V-PDB standards for N and C, respectively. On a day-to-day basis, we measured and calibrated analyses with a laboratory IU Acetanilide standard (δ15N = 1.18‰, δ13C = -29.52‰, %N = 10.36%, %C = 71.02%) and a laboratory gelatin standard (δ15N = 5.60‰, δ13C = -12.60‰, %N = 16.44‰, %C = 44.02‰). The isotopic and concentration value of these laboratory standards are known by calibration to international standards (IAEA601, IAEA-346, USGS24, USGS25, USGS26, USGS34, USGS35, USGS41). We applied mass and drift corrections during each instrument session using the gelatin standard. Standard deviations for standards were < 0.1 ‰ for both δ15N and δ13C and <0.05 for C/N (seven PUGel standards analyzed at the start of each session and a PUGel and an Acetanilide standard analyzed after every eight samples during the session).