Biogeochemical Subsidies from Glacier Runoff into Alaska Coastal Marine Food Webs, Gulf of Alaska, 2012-2013

Collection methodology: Bulk plankton samples were collected with a 150 µ mesh ring net on a 50 m vertical haul and the net contents were placed in filtered water for 3 hours to evacuate any gut contents. After ~ 1 ml of bulk plankton was placed in a vial, 20-40 large copepods were individually removed from the remaining bulk plankton sample and placed in a vial and frozen with ~1 ml of seawater. In Prince William Sound, euphausiids (Euphausia pacifica, Thysanoessa spp.), mysiids (Neomysis rayii), amphipods (Themisto libellula), young of the year (YOY) fish (capelin, Mallotus villosus; Pacific herring, hereafter herring, Clupea pallasii; and walleye pollock, hereafter pollock, Gadus chalcogrammus), and age 1+ fish (capelin; herring; pollock; eulachon, Thaleichthys pacificus; and Pacific sand lance, hereafter sand lance, Ammodytes personatus) were collected with a modified herring trawl towed in the near surface water (< 20 m tow depth). Whole crustaceans were placed in a vial and frozen in the field. Fish dorsal muscle tissue was excised and placed in a vial and frozen in the field, or sometimes whole fish were frozen and muscle tissue was excised in the lab. In the Western Aleutians, euphausiids (Thysanoessa spp.), YOY and age 1+ fish were collected at seabird colonies from adults provisioning their chicks (for details see Hatch and Sanger 1992). Mussels (Mytilus trossulus) were collected in the intertidal zone, at low tide when they were accessible, near kelp and other intertidal organisms in the Western Aleutians and near glacier runoff in Prince William Sound. Kelp is not commonly found in areas with close proximity to glacier runoff, therefore kelp was present at the collection site in the Western Aleutians and not in Prince William Sound. Benthic organisms (marine sponge, Suberites sp.; snail, Fusitriton oregonensis and Colus halli, and spot shrimp, Pandalus hypsinotus) collected in a trawl near tidewater glaciers were frozen whole until they could be processed in the lab. Kittlitz’s murrelets (Brachyramphus brevirostris) and marbled murrelets (Brachyramphus marmoratus) at coastal and oceanic sites were captured using the night lighting technique (Whitworth et al. 1997). Blood was drawn from the tarsal vein of murrelets and frozen in glass vial in the field. Black-legged kittiwakes (Rissa tridactyla) were collected with a shot gun (Prince William Sound only). Black-legged kittiwake liver samples were frozen individually in glass vials. Replicate samples of a subset of biota from each taxon were placed in pre-combusted glass vials and frozen in the field for Δ14C analysis. Freshwater was collected in an acid-washed 4L container from melt water runoff streams and seawater was collected in 4 L Niskin bottles using a Seabird Electronics SBE 19 plus v.2 conductivity-temperature-depth profiler equipped with an SBE55 auto-fire water sampler. To normalize tissue δ2H values due to environmental water, we collected water samples at each location that biota were sampled for stable isotopes in Prince William Sound, and at a single location near Attu Island in the Western Aleutians. To map the water isoscape we also sampled sites throughout marine waters in Prince William Sound. Water samples for δ2H and δ18O analysis were collected from melt water runoff near recently grounded glaciers (as above), and from 2 and 10 m water depths in the coastal waters in 4-L Niskin bottles. In 2013, we also collected a water sample near the bottom of each CTD cast, or to a maximum of 300 m depth. These water samples were stored in tightly sealed 20 ml plastic vials with no headspace at room temperature until they could be processed in the lab.

Lab methodology: POM filters were oven dried at 50 °C for 48 hours, and acidified with concentrated sulfurous acid to remove inorganic carbon. δ13C and δ15N analyses of POM were performed at UC Davis Stable Isotope Facility with standard deviations of < 0.1‰ for δ13C and < 0.2 ‰ for δ15N . Plankton (> 150 µ) and other biota samples (δ2H, δ13C, and δ15N), as well as water samples (δ2H and δ18O) were processed at the Environment Canada lab, Saskatchewan. Samples were thawed, rinsed in distilled water, oven dried (plankton) or freeze dried (all other organisms) for 48 hours, powdered and treated with 2:1 chloroform methanol solution to account for variable lipid content across species and trophic levels. Crustaceans were soaked in 0.1 N HCl to remove carbonates. Stable carbon and stable nitrogen isotope assays were performed on 1 mg samples of homogenized materials by combustion at 850ºC in a Robo-Prep elemental analyzer. CO2 and N2 gases were analyzed with an interfaced Europa 20:20 continuous flow isotope ratio mass spectrometer (Europa, ** USA) with every five unknowns separated by two (egg albumen) laboratory standards. Based on within-run replicate measurements of the lab standard, we estimated measurement error as ± 0.1‰ for δ13C and ± 0.2‰ for δ15N. Stable-hydrogen isotope analyses of organic materials were conducted using the comparative equilibration method described by (Wassenaar and Hobson 2003) through the use of calibrated keratin hydrogen-isotope reference materials (CBS: -197‰, SPK:-121.6‰ , KHS: -54.1‰). δ2H isotope measurements were performed on H2 derived from high-temperature (1350 ˚C) flash pyrolysis of 350 ± 10 ug tissue subsamples using continuous-flow isotope-ratio mass spectrometry (Isoprime, Manchester UK). Within-run (n=5) measurement of the three keratin laboratory reference materials (CFS, CHS, BWB) indicated measurement error < 2‰. Water isotopes (δ2H and δ18O) were determined using off-axis integrated-cavity output laser spectroscopy (Model DLT-100; Los Gatos Research Inc.) using the method described in Lis et al. (2008). All samples were normalized to internal laboratory water standards previously calibrated relative to VSMOW (0‰) and VSLAP (δ2H −428 ‰ and δ18O −55.5 ‰). Natural abundance Δ14C was analyzed for a subset of samples from each focal food web component at Woods Hole Oceanographic Institute by accelerator mass spectrometry using the methods of Raymond and Bauer (2001). Samples for analyses of Δ14C in DOC, DIC, POM and biota were acidified with concentrated sulfurous acid to remove carbonates, combusted at 900 ºC in double sealed quartz tubes, and the evolved CO2 was purified and quantified on the vacuum extraction line (Druffel et al. 1992). Isotope abundance is expressed in δ notation as the deviation from standards in parts per mil (‰) according to the following equation: δX=[(R_sample/R_standard )-1]×1000 (1) where X is 13C, 15N, or 2H, and R is the corresponding ratio 13/12C, 15/14N, or 2/1H. R_standard is based on Vienna PeeDee Belemnite (VPDB) for δ13C, atmospheric N2 for δ15N, and Vienna Standard Mean Ocean Water-Standard Light Antarctic Precipitation (VSMOW-SLAP) for δ2H. Reported Δ14C values were corrected for fractionation using the δ13C values of the samples. Most samples had paired measures, however, δ13C samples were lost in a small number of cases. We substituted the nearest within-fjord δ13C value for two POM samples with missing δ13C values, and substituted the DIC δ13C value of a sample with the most similar δ2H water value for 3 missing DIC δ13C measures. Because there was relatively little change in coastal marine DIC δ13C (- 0.77 at 2 m depth, or 0.17 at 160 m depth) the difference in Δ14C value when substituting one δ13C value or the other was 2 ‰, and less than the range of expected error (5 – 10 ‰).