Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Studies in Fe bioavailability : co-limitation of primary productivity by iron, light, and nitrate in… Taylor, Rebecca Lynn 2011

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
24-ubc_2012_spring_taylor_rebecca.pdf [ 13.12MB ]
Metadata
JSON: 24-1.0053264.json
JSON-LD: 24-1.0053264-ld.json
RDF/XML (Pretty): 24-1.0053264-rdf.xml
RDF/JSON: 24-1.0053264-rdf.json
Turtle: 24-1.0053264-turtle.txt
N-Triples: 24-1.0053264-rdf-ntriples.txt
Original Record: 24-1.0053264-source.json
Full Text
24-1.0053264-fulltext.txt
Citation
24-1.0053264.ris

Full Text

STUDIES IN FE BIOAVAILABILITY: CO-LIMITATION OF PRIMARY PRODUCTIVITY BY IRON, LIGHT, AND NITRATE IN THE BEAUFORT SEA, AND DIRECT FE-SIDEROPHORE UPTAKE MECHANISMS IN FE DEFICIENT PHYTOPLANKTON  by    REBECCA LYNN TAYLOR   B.Sc., University of Guelph, 2008    A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF   MASTER OF SCIENCE   in   The Faculty of Graduate Studies   (Oceanography)    THE UNIVERSITY OF BRITISH COLUMBIA  (Vancouver)   November 2011    © Rebecca Lynn Taylor, 2011  ii Abstract This study explores two relevant questions in the realm of iron (Fe) bioavailability to phytoplankton. First, does Fe availability limit (or co-limit) growth of indigenous plankton communities in the Arctic Ocean? Second, can phytoplankton internalize ferrated siderophores with a non-reductive uptake mechanism? To address the first question, an 8-day grow out experiment was conducted in the Beaufort Sea in early September 2009, during which light, Fe, and nitrate (NO3-) levels were manipulated. Bottles were sampled on days 0, 2, 4, 6, and 8 for accessory pigments, size-fractionated chlorophyll a, phytoplankton abundance and composition, nutrients, Fe quotas and uptake rates. It was found that NO3- was limiting plankton growth at the time of sampling. The community also appeared to be light limited. Additionally, co-limitation of primary production by Fe and light at light levels ≤ 10 % surface irradiance (Io) was observed. These results have interesting implications about how the seasonality of NO3-, light, and Fe availability may control primary productivity in the Beaufort Sea. To address the second question, I investigated the potential of a non-reductive Fe uptake mechanism for siderophore-bound Fe in the model diatom Thalassiosira oceanica and the in situ plankton communities along Line P in the subarctic Pacific Ocean in late summer 2010. To do this, we radiolabeled the siderophore desferrioxamine B (DFB) by methylating its terminal amine group with radioisotope 14C methyl iodide. Internalization of 55Fe14DFB was observed both in phytoplankton cultures and field communities along Line P, suggesting the presence of a non-reductive Fe uptake mechanism in phytoplankton. However, the results are inconclusive due to the inability to purify and verify the concentration of 14DFB.  The overarching goal of this investigation was to gain a better understanding on the bioavailability and acquisition of Fe by phytoplankton. This is imperative in order to predict the role of Fe in future primary productivity, and subsequently the fate of phytoplankton communities and the biological carbon pump, as our oceans respond to global warming.  iii Preface  The work in Chapter 2 describes a field project conducted in the Beaufort Sea. Maite Maldonado, Chris Payne, and David Semeniuk accompanied and trained me during the GEOTRACES expedition on August 27 – September 12, 2009. All on-board work was conducted collaboratively. Once we returned to The University of British Columbia, Chris Payne analyzed nutrient samples, Constance Couture analyzed HPLC and flow cytometry pigment samples, and Maite Maldonado interpreted HPLC sample data with CHEMTAX software. Jay Cullen’s group, from the University of Victoria, analyzed dissolved Fe samples from station L1. David Semeniuk was a valuable resource to discuss ideas and results. Beyond these extremely generous contributions, I have been responsible for the analysis and interpretation of data for this chapter. Chapter 3 describes both laboratory and fieldwork, the latter of which occurred along Line P in the subarctic Pacific Ocean from August 17 – September 2, 2010. David Semeniuk conducted the initial experiment that inspired this investigation in the summer 2007; his work is described in the Introduction of Chapter 3. Tom Ruth and Jian-Ming Lu were responsible for coordinating and labeling the commercially available siderophore Desferal® with 14C-methyl- iodide. David Semeniuk also accompanied me on the fieldwork component of this Chapter. I was responsible for experimental design, setup, and analysis, and welcomed Dave’s help in filtering while aboard the John P. Tully. Keith Johnston analyzed Fe samples from each station of the expedition. Josiane Mélançon analyzed chlorophyll a samples while aboard the ship. Beyond these much-appreciated collaborations, I have been responsible for the analysis and interpretation of data for this chapter.    iv Table of Contents  Abstract.......................................................................................................................................... ii	
   Preface........................................................................................................................................... iii	
   Table of Contents ......................................................................................................................... iv	
   List of Tables ................................................................................................................................ vi	
   List of Figures................................................................................................................................ x	
   Acknowledgements ..................................................................................................................... xii	
   Dedication ................................................................................................................................... xiii	
   Chapter 1: Introduction ............................................................................................................... 1	
   1.1	
   A brief early history of phytoplankton Fe nutrition research .......................................................... 1	
   1.2	
   The role of phytoplankton in the world ocean................................................................................. 2	
   1.3	
   Phytoplankton physiological Fe requirement .................................................................................. 2	
   1.4	
   Geochemistry of Fe in oceans ......................................................................................................... 4	
   1.5	
   Siderophores .................................................................................................................................... 6	
   1.6	
   Fe uptake mechanisms of marine phytoplankton ............................................................................ 9	
   1.7	
   From physiology to field and future .............................................................................................. 11	
   1.8	
   Research objectives ....................................................................................................................... 12	
   Chapter 2: Co-limitation by Light, Nitrate and Iron in the Beaufort Sea............................. 14	
   2.1	
   Introduction ................................................................................................................................... 14	
   2.2	
   Materials and methods................................................................................................................... 17	
   2.2.1	
   Station parameters and experimental design.......................................................................... 17	
   2.2.2	
   Biological and chemical parameters ...................................................................................... 20	
   2.2.3	
   Fe uptake rates and Fe:C ratios .............................................................................................. 24	
   2.2.4	
   Statistical analyses ................................................................................................................. 26	
   2.3	
   Results ........................................................................................................................................... 27	
   2.3.1	
   Overview of results ................................................................................................................ 27	
   2.3.2	
   Station parameters .................................................................................................................. 28	
   2.3.3	
   Plankton community growth and limitation by nitrate or light.............................................. 30	
   2.3.4	
   Plankton growth co-limitation by Fe and light ...................................................................... 33	
   2.3.5	
   Macronutrient drawdown ....................................................................................................... 35	
   2.3.6	
   Plankton community composition.......................................................................................... 37	
   2.3.7	
   Short-term Fe uptake rates ..................................................................................................... 42	
   2.3.8	
   Fe:C ratios .............................................................................................................................. 44	
   2.4	
   Discussion...................................................................................................................................... 45	
   2.4.1	
   The plankton community at L1 in the Beaufort Sea during the summer ............................... 45	
   2.4.2	
   Primary nitrogen source ......................................................................................................... 46	
   2.4.3	
   Co-limitation by light and Fe ................................................................................................. 48	
   2.4.4	
   Light limitation....................................................................................................................... 50	
   2.4.5	
   Decoupled seasonalities of Fe, light, and NO3- ...................................................................... 52	
   2.4.6	
   Oceanographic implications................................................................................................... 53	
     v Chapter 3: Non-reductive Fe transport from an Fe-ligand complex by the model diatom Thalassiosira oceanica and in situ plankton communities along Line P in the Subarctic Pacific Ocean ............................................................................................................................... 56	
   3.1	
   Introduction ................................................................................................................................... 56	
   3.2	
   Materials and methods................................................................................................................... 61	
   3.2.1	
   Culturing techniques .............................................................................................................. 61	
   3.2.2	
   Laboratory short-term uptake experiments ............................................................................ 62	
   3.2.3	
   Detecting ligand uptake.......................................................................................................... 64	
   3.2.4	
   Developing a 14C-labeling protocol for DFB ......................................................................... 64	
   3.2.5	
   Labeling DFB with 14C .......................................................................................................... 67	
   3.2.6	
   Laboratory testing and use of 14DFB...................................................................................... 68	
   3.2.7	
   Field sites ............................................................................................................................... 71	
   3.2.8	
   Biological and chemical parameters ...................................................................................... 72	
   3.2.9	
   Short-term 55Fe14DFB uptakes along Line P .......................................................................... 73	
   3.3	
   Results and discussion ................................................................................................................... 76	
   3.3.1	
   Laboratory results................................................................................................................... 76	
   3.3.1.1	
   Fe contamination of 14DFB............................................................................................. 76	
   3.3.1.2	
   Non-reductive Fe uptake by T. oceanica ........................................................................ 76	
   3.3.2	
   Field results ............................................................................................................................ 78	
   3.3.2.1	
   Station parameters........................................................................................................... 78	
   3.3.2.2	
   Plankton growth and nutrient drawdown........................................................................ 80	
   3.3.2.3	
   Plankton community composition and nutritional status................................................ 82	
   3.3.2.4	
   Non-reductive Fe uptake by indigenous plankton communities along Line P ............... 84	
   3.3.2.5	
   Trends in non-reductive Fe uptake along Line P ............................................................ 89	
   3.3.3	
   Oceanographic implications and future direction .................................................................. 90	
   Chapter 4: Conclusion................................................................................................................ 92	
   4.1	
   Fe and Arctic primary productivity ............................................................................................... 92	
   4.2	
   Fe and phytoplankton survival ...................................................................................................... 93	
   4.3	
   Fe and global warming .................................................................................................................. 95	
   Bibliography ................................................................................................................................ 97	
   Appendix.................................................................................................................................... 112	
      vi List of Tables  Table 2.1: Control and experimental treatments (marked by X) and their corresponding light, iron, and nitrate levels................................................................................................................... 19	
    Table 2.2: Matrices of pigment : chlorophyll a ratios for eight algal groups utilized in CHEMTAX analysis. The initial pigment ratios matrix (A) was derived from a compilation of numerous pigment investigations throughout the world ocean (Wright et al., 1996; Mackey et al., 1998; Rodríguez et al., 2002; Suzuki et al., 2002; Vidussi et al., 2004; Latasa, 2007; Hashihama et al., 2008; Miki et al., 2008; Fujiki et al., 2009; Kim et al., 2010; Wright et al., 2010; Wright, personal communication). The final pigment ratios matrix (B) represents an average of the best 10 % of 60 iterative matrices of the initial pigment ratios matrix for each analyzed sample, based on the RMS of each estimate; these averaged best matrices varied less than 10 %. Pigment annotation: Chlc3, chlorophyll c3; MgDVP, magnesium 3,8-divinylphaeoporphyrin a5 monomethyl ester; Chlc2, chlorophyll c2; Per, peridinin; Butfuc, 19’-butanoyloxyfucoxanthin; Fucox, fucoxanthin; Neox, neoxanthin; Pras, prasinoxanthin; Violax, violaxanthin; Hexfuc, 19’- hexanoyloxyfucoxanthin; Ddx, diadinoxanthin; Allox, alloxanthin; Zeax, zeaxanthin; Lut, lutein; Chlb, chlorophyll b; Chla, chlorophyll a. ..................................................................................... 22	
    Table 2.3: Average size-fractionated and total chlorophyll a (ng L-1) for triplicate initials on day 0 and duplicate samples for control and experimental treatments on the final day of an 8-day grow out initiated at station L1 (71°05.94’ N, 139°08.87’ W) in the Beaufort Sea. Values reported equal the mean ± standard error for initials and the range for day 8 duplicates; absence of a range indicates the treatment was only analyzed in singlet. Values in brackets represent the times increase from the control treatment..................................................................................... 31	
    Table 2.4: Summary of Fe effects, light (%Io) effects, and Fe × %Io interactive effects, determined by two-way ANOVA statistical analyses, on chlorophyll a, net and maximum nutrient drawdown, and Fe:C ratios. All three Fe levels and three light levels were included in these analyses. Only significant effects are included (p < 0.05). Blank spaces indicate no significant effects; “n/a” indicates test was not possible. ............................................................. 34	
    Table 2.5: Average NO3-, PO43-, and SiO42- drawdown (µmol L-1) for triplicate initials on day 0 and duplicate samples on all sampling days of an 8-day grow out initiated at station L1 in the Beaufort Sea. Initial values reported equal the mean ± standard error. Net drawdown values are equal to the day 8 nutrient concentration subtracted from the initial nutrient concentration. Maximum drawdown values are equal to the lowest nutrient concentration observed throughout the experiment subtracted from the initial nutrient concentration; the day of lowest nutrient concentration is reported in brackets. All net and maximum NO3- drawdown values are calculated with the initial + NO3- value, save for the control treatment which is calculated with the initial value.............................................................................................................................. 36	
        vii Table 2.6: Mean abundance (%) of eight algal groups on day 8 of an 8-day grow out initiated at station L1 (71°05.94’ N, 139°08.87’ W) in the Beaufort Sea. Values represent the mean ± the range for duplicate treatment bottles; those without an error value were detected in one bottle per treatment. Mean abundance of Prasinophytes Types 1 and 2 were combined. Mean abundances below detection are noted with “bd”. Data was unavailable for those noted with “-”. ................ 38	
    Table 2.7: Short-term (24 h) volumetric (×10-12 mol Fe L-1 d-1) and carbon-specific (×10-6 mol Fe (mol C)-1 d-1) Fe uptake rates of size-fractionated phytoplankton on day 6 of an 8-day grow out initiated at station L1 (71°05.94’ N, 139°08.87’ W) in the Beaufort Sea. Iron was precomplexed to ligands DFB or EDTA in the ratios 100 nmol L-1 Fe : 105 nmol L-1 DFB with 5 % of total Fe as 55Fe, or 100 nmol L-1 Fe : 1000 nmol L-1 EDTA with 1 % of total Fe as 55Fe. Values are equal to the mean ± the range for duplicate bottles. ............................................................................... 42	
    Table 2.8: Size-fractionated Fe:C ratios (µmol Fe (mol C)-1) of phytoplankton on day 6 of an 8- day grow out initiated at station L1 (71°05.94’ N, 139°08.87’ W) in the Beaufort Sea. Iron was precomplexed to DFB in the ratio 100 nmol L-1 Fe : 105 nmol L-1 DFB with 5 % of total Fe as 55Fe. Carbon was added to each bottle to a final concentration of 20 µCi L-1 14C. Values are equal to the mean ± the range for duplicate bottles. ............................................................................... 44	
    Table 3.1: Short-term Fe uptake rates (×10-22 mol Fe µm-2 h-1) of a single culture of Fe-limited (1.28 nmol L-1 Fe) T. oceanica. Fe was precomplexed in a 10 nmol L-1 Fe : 15 nmol L-1 DFB ratio with each of desferrioxamine mesylate (brand name Desferal®) and hydrogen chloride- labeled DFB. The growth rate of a 1 L culture of T. oceanica was 0.69 doublings d-1, cell density was 29 × 104 cells mL-1, and mean cell size was 4.8 µm when harvested. Reported values are equal to the mean ± the range for pseudoreplicates. ..................................................................... 66	
    Table 3.2: Short-term Fe uptake rates (×10-22 mol Fe µm-2 h-1) of Fe-limited (13.7 nmol L-1 Fe) T. oceanica cultures. Fe was precomplexed to DFB in a 1 nmol L-1 Fe : 10 nmol L-1 DFB ratio. Reported values are equal to the mean ± the range for pseudoreplicates. The growth rate of a 0.5 L T. oceanica culture for experiment 1 was 0.95 doublings d-1; cell density was 23 × 104 cells mL-1, and mean cell size was 4.5 µm. For experiment 2, the growth rate of a 1 L culture was 1.2 doublings d-1, the cell density was 23 × 104 cells mL-1, and mean cell size was 4.3 µm. Experiment 3 was characterized by a growth rate of a 1 L T. oceanica culture of 1.0 doublings d- 1, a cell density of 13 × 104 cells mL-1, and a mean cell size of 4.2 µm. ...................................... 67	
    Table 3.3: Control and experimental treatments utilized in short-term 55Fe14DFB uptake rate experiments along Line P. Mixtures of Fe and DFB (and Ga if applicable) are outlined for each treatment; components marked with an X were not added to the respective treatment. Each treatment contained a 1 : 2 molar ratio of Fe : DFB. .................................................................... 74	
    Table 3.4: Coordinates, estimated incident irradiance (Io; µE m-2 s-1), mixed layer depth (MLD; m), seawater temperature at pumping depth (°C), dissolved Fe (nmol L-1), NO3- (µmol L-1), PO43- (µmol L-1), and SiO42- (µmol L-1) at four stations along Line P in the subarctic Pacific Ocean. Irradiances, MLD, and temperature were determined from CTD data. Samples for nutrients were collected at the same time and depth (10 m) as pumping for NRU experiments. Station incident irradiances were estimated via logarithmic regression of deep cast CTD data; R2 values from  viii these analyses are also reported. Stations’ MLD are equal to the mean ± standard error for triplicate CTD casts. Dissolved Fe concentrations were determined by Keith Johnson as described in Johnson et al. (2005). Station P4 macronutrient concentrations are equal to the mean ± the range for duplicate samples. Station P20 and P26 macronutrient are equal to the mean ± standard error for triplicate samples. ............................................................................................ 79	
    Table 3.5: Total chlorophyll a (µg L-1), size-fractionated chl a (% contribution to total chl a), relative abundance of diatoms, coccolithophores, and dinoflagellates (%), and cell density of picoeukaryotes and heterophic bacteria (cells mL-1) of four locations along Line P in the subarctic Pacific Ocean. Total and size-fractionated chl a, light microscopy, and flow cytometry samples were collected at the same time and depth (10 m) as pumping for NRU experiments. Station P20 and P26 chlorophyll a concentrations and flow cytometry cell densities are equal to the mean ± standard error for triplicate samples. Station P4 and P16 flow cytometry cell densities are equal to the mean ± the range for duplicate samples. ............................................................. 81	
    Table 3.6: Short-term (8, 16, and 24 h) C-normalized Fe (×10-6 mol Fe (mol C)-1 d-1) and DFB (×10-3 mol DFB (mol C)-1 d-1) uptake rates by size-fractionated phytoplankton at station P4 along Line P in the subarctic Pacific Ocean. Seawater was obtained before sunrise from 10 m depth. During the precomplexation step for each treatment, 1 nmol L-1 55Fe and 4 nmol L-1 Fe were complexed with 10 DFB nmol L-1; 10 nmol L-1 14DFB was also added to the experimental treatments. Additionally, it was estimated that ~ 5 nmol L-1 Fe contamination was present in the 14DFB solution, maintaining a Fe : DFB molar ratio of 1 : 2 for each treatment. Gallium was added in a 1 : 1 molar ratio to Fe where applicable. Rates are ± standard error for triplicate measurements. Significantly different (p < 0.05) uptake rates within each size-fraction for each sampling event are marked with different letters (a,b); rates without letters were not significantly different......................................................................................................................................... 85	
    Table 3.7: Short-term (8-h) C-normalized Fe (×10-6 mol Fe (mol C) -1 d-1) and DFB (×10-3 mol DFB (mol C) -1 d-1) uptake rates by size-fractionated phytoplankton at stations P4, P16, P20, and P26 along Line P in the subarctic Pacific Ocean. Seawater was obtained before sunrise from 10 m depth at each station. During the precomplexation step for each treatment, 1 nmol L-1 55Fe and 4 nmol L-1 Fe were complexed with 10 DFB nmol L-1; 10 nmol L-1 14DFB was also added to the experimental treatments. Additionally, it was estimated that ~ 5 nmol L-1 Fe contamination was present in the 14DFB solution, maintaining a Fe : DFB molar ratio of 1 : 2 for each treatment. Gallium was added in a 1 : 1 molar ratio to Fe where applicable. Rates are ± standard error for triplicate measurements. Significantly different (p < 0.05) uptake rates within each size-fraction for each sampling event are marked with different letters (a,b,c); rates without letters were not significantly different.................................................................................................................... 87	
    Table 3.8: Summary of statistically significant differences (p < 0.05) between stations’ size- fractionated C-normalized 8-h ρFe for each treatment, determined by one-way ANOVA. Statistical differences are denoted by different letters (A,B,C). ................................................... 89	
    Table 3.9: Summary of statistically significant differences (p < 0.05) between stations’ size- fractionated C-normalized 8-h ρDFB for each treatment, determined by one-way ANOVA. Statistical differences are denoted by different letters (A,B,C). ................................................... 90	
    ix Table A.1: Control and experimental treatments (marked by X) and their corresponding light, iron, and nitrate levels................................................................................................................. 112	
    Table A.2: Average size-fractionated and total chlorophyll a (ng L-1) for triplicate initials on day 0 and duplicate samples for control and experimental treatments on the final day of an 8-day grow out initiated at station L1 (71°05.94’ N, 139°08.87’ W) in the Beaufort Sea. Values reported equal the mean ± standard error for initials and the range for day 8 duplicates; absence of a range indicates the treatment was only analyzed in singlet. Values in brackets represent the times increase from the control treatment................................................................................... 113	
    Table A.3: Size-fractionated Fe:C ratios (µmol Fe (mol C)-1) of size-fractionated phytoplankton on day 6 of an 8-day grow out initiated at station L1 (71°05.94’ N, 139°08.87’ W) in the Beaufort Sea. Iron was precomplexed to DFB in the ratio 100 nmol L-1 Fe : 105 nmol L-1 DFB with 5 % of total Fe as 55Fe. Carbon was added to each bottle to a final concentration of 20 µCi L-1 14C. Values are equal to the mean ± the range for duplicate bottles. .................................... 114	
    Table A.4: Short-term (24 h) volumetric (×10-12 mol Fe L-1 d-1) and carbon-specific (×10-6 mol Fe (mol C)-1 d-1) Fe uptake rates of size-fractionated phytoplankton on day 6 of an 8-day grow out initiated at station L1 (71°05.94’ N, 139°08.87’ W) in the Beaufort Sea. Iron was precomplexed to ligands DFB or EDTA in the ratios 100 nmol L-1 Fe : 105 nmol L-1 DFB with 5 % of total Fe as 55Fe, or 100 nmol L-1 Fe : 1000 nmol L-1 EDTA with 1 % of total Fe as 55Fe. Values are equal to the mean ± the range for duplicate bottles. ................................................. 115	
    Table A.5: Average NO3-, PO43-, and SiO42- drawdown (µmol L-1) for triplicate initials on day 0 and duplicate samples on all sampling days of an 8-day grow out initiated at station L1 (71°05.94’ N, 139°08.87’ W) in the Beaufort Sea. Initial values reported equal the mean ± standard error. Net drawdown values are equal to the day 8 nutrient concentration subtracted from the initial nutrient concentration. Maximum drawdown values are equal to the lowest nutrient concentration observed throughout the experiment subtracted from the initial nutrient concentration; the day of lowest nutrient concentration is reported in brackets. All net and maximum NO3- drawdown values are calculated with the Initial + NO3- value, save for the control treatment which is calculated with the initial value. ...................................................... 116	
    x List of Figures  Figure 1.1: Chemical structures of a) a catecholate bidentate-coordinating group and b) a hydroxamate functional bidentate-coordinating group, both complexing Fe(III). ......................... 7	
    Figure 1.2: Chemical structure of the hydroxamate siderophore desferrioxamine B (DFB) complexing Fe(III). ......................................................................................................................... 9	
    Figure 2.1: Bathymetric map of the Beaufort Sea and Canadian Arctic Continental Shelf. Station L1, (71°05.94’ N, 139°08.87’ W; 1918 m depth) is located north of Yukon, Canada and west of Banks Island. Figure produced using Ocean Data View (Schilitzer, 2004). ................................ 18	
    Figure 2.2: Vertical profiles of (a) density (kg/m3), oxygen (mL L-1) (b) transmissivity (%), PAR (µE m-2 s-1), fluorescence, (c) chlorophyll a (ng L-1), nitrate (µmol L-1), phosphate (µmol L-1), and silicate (µmol L-1) in the upper 150 m at station L1 (71°05.94’ N, 139°08.87’ W) in the Beaufort Sea. Data points in figures (a) and (b) are from a CTD cast conducted at 16:22 MDT on August 31, 2009. Chlorophyll a and nutrient data in figure (c) were obtained from samples collected from Niskin bottles on a rosette deployed with the same CTD cast. ............................ 29	
    Figure 2.3: Average chlorophyll a (ng L-1) concentration for treatment duplicates over the course of an 8-day grow out experiment initiated at station L1 (71°05.94’ N, 139°08.87’ W) in the Beaufort Sea, comparing (a) the no Fe added treatment group, (b) 50 %Io treatments, and (c) 10 %Io treatments. Initial parameters of the seawater collected from 29 m depth include a NO3- concentration, chlorophyll a concentration, and irradiance level of 0.2 µmol L-1 NO3-, 40 ng L-1 chl a, and 6 % of the incident irradiance (Io = 513.73 µE m-2 s-1), respectively. Chlorophyll a concentrations represent biomass collected with 0.22 µm filters on days 2, 4, and 6 and are the summation of biomass collected on 0.22, 1, and 5 µm filters on day 8. Error bars represent the range for duplicate bottles............................................................................................................. 32	
    Figure 2.4: Relationships between cell density of picoeukaryotes (cells mL-1) and total chlorophyll a concentration (ng L-1) for (a) all of days 4, 6, and 8, (b) day 8, and (c) day 6 of an 8-day grow out experiment initiated at station L1 (71°05.94’ N, 139°08.87’ W) in the Beaufort Sea. This figure includes duplicate sample data for all light levels within the no Fe added, + 1 nM Fe, and + 1 nM DFB treatment groups. The no Fe added and + 1 nM Fe treatments at 50 %Io are denoted by ▲, the no Fe added and + 1 nM Fe treatments at 10 %Io are denoted by ■, and all other treatments are denoted by ●. Linear regressions: a. (y = 8482.5x + 1162.2, R2 = 0.84, t0.0001(2)52 = 16.24, p < 0.0001), b. (y = 9802.8x -137.36, R2 = 0.92, t0.0001(2)16 = 13.37, p < 0.0001), and c. (y = 17,577x + 376.04, R2 = 0.77, t0.0001(2)16 = 7.38, p < 0.0001). Note the smaller axes scales in 4c. ........................................................................................................................... 41	
          xi Figure 3.1: Short-term uptake rates of Fe(III) (×10-20 mol Fe µm-2 h-1) bound to ferrioxamine B (DFB) in a 1:2 molar ratio as a function of Fe concentration. Data points are a) duplicates of Fe- limited (10 nmol L-1 Fe) T. oceanica cultures and b) an Fe-limited (40 nmol L-1 Fe) T. pseudonana culture. The average growth rate of duplicate 2 L cultures of T. oceanica was 0.98 ± 0.060 doublings d-1 and average cell density was 14 ± 2 × 104 cells mL-1 when harvested. The growth rate of a 2 L culture of T. pseudonana was 1.7 doublings d-1 and cell density was 33 × 104 cells mL-1 when harvested. ..................................................................................................... 58	
    Figure 3.2: Ligand exchange Fe uptake mechanism by Aeromonas hydrophilia: a) a ferrated donor siderophore (square) approaches a desferrated siderophore (oval) bound by a receptor protein in the cell membrane (in grey), b) the donor siderophore and membrane-bound siderophore undergo an exchange of Fe(III), c) the exchange of Fe(III) to the membrane-bound siderophore induces a conformational change in the receptor protein, allowing the the now- ferrated, membrane-bound siderophore to move into the periplasmic space, d) the receptor protein then binds the now-desferrated donor siderophore to the cell surface, returns to its initial conformation, and the cycle begins again. Figure based on Stintzi et al., 2000. .......................... 60	
    Figure 3.3: Chemical structures of a) desferrioxamine mesylate (brand name Desferal®), b) hydrogen chloride-labeled DFB, c) methylate salt-labeled DFB, d) 14C-labeled DFB (14DFB), and e) 14DFB complexing 55Fe...................................................................................................... 65	
    Figure 3.4: CAS assay results, including (a) the absorbance spectrum for various CAS-*DFB (DFB or 14DFB) solutions, an (unchelexed) CAS solution, and water, and (b) absorbance as a function of *DFB concentration. 14DFB data in both figures represent an average from duplicate assays. ........................................................................................................................................... 69	
    Figure 3.5: Short-term Fe and DFB uptake rates (×10-21 mol µm-2 h-1) of a Fe-limited (1.28 nmol L-1 Fe) culture of T. oceanica. Uptake was determined for both an experimental treatment and a cold-control treatment, which was subjected to near-freezing temperatures. For both treatments, Fe was precomplexed to DFB in a 5 nmol L-1 Fe : 20 nmol L-1 DFB ratio, with all of added Fe as 55Fe and 50% of added DFB as 14DFB. Additionally, it was estimated that ~ 5 nmol L-1 Fe contamination was present in the 14DFB solution, yielding an estimated final Fe : DFB ratio of 10 nmol L-1: 20 nmol L-1. The growth rate of a 1 L culture was 0.46 doublings d-1, cell density was 28 × 104 cells mL-1, and cell size was 4.3 µm when harvested. All treatments exhibited significantly different slopes. Linear regressions: Fe uptake (y = 5.47x, R2 = 0.98), DFB uptake (y = 7.85x, R2 = 0.93), Fe cold control (y = 0.32x, R2 = 0.04), and DFB cold control (y = 3.44x, R2 = 0.55). Testing the equality of slopes: Fe uptake vs. DFB uptake (t0.05(2)10 = 4.9, p < 0.05), Fe uptake vs. Fe cold control (t0.05(2)10 = 5.8, p < 0.05), DFB uptake vs. DFB cold control (t0.05(2)10 = 4.8, p < 0.05). ................................................................................................................................ 78	
     xii Acknowledgements  I would like to thank Dr. Maite Maldonado for her invaluable guidance, focus, passion, and support throughout my completion of this degree. Throughout many challenges she has been one of the most supportive people around me. I am lucky to have worked with her these last three years.  Thank you to Dr. Philippe Tortell, Dr. Raymond Andersen, and Dr. Roger François for their guidance and feedback throughout this project. Thank you to Chris Payne for his patience and teaching aboard the Amundsen. Thank you to Jay Cullen and Keith Johnson for their invaluable contributions to my field research.  Thank you to all of those in the Tornado Lab Family for their camaraderie and feedback. I have enjoyed countless coffees and lab meetings with all of you. Special thanks to Kristina Brown, a one-woman cheering squad, and to Dave Semeniuk for three years of friendship and teamwork both on and off land.  Finally, I could not have completed this endeavor without the unwavering support of my friends, Warren, and my parents. There have been many occasions that I had to be convinced of my potential from others and skeptically took their word for it – Thank you.       xiii Dedication             I dedicate this thesis to my parents, for 26.25 years of unwavering support (and counting).    1 Chapter 1: Introduction  1.1 A brief early history of phytoplankton Fe nutrition research The integral role of Fe in marine primary production and productivity has been apparent since the late 1920s, as phytoplankton physiological studies began to explore optimal growth conditions and field studies sought to explain regional differences in oceanic primary production. In the laboratory, interest in primary productivity was initially secondary as evidenced by Hopkins and Wann’s (1926) work with the green alga Chlorella, which was used as an easily controllable proxy for higher plants. During his investigation, Hopkins sought to determine the optimal pH to facilitate growth of Chlorella, and suggested Fe as the culprit when Fe precipitates correlated with poor culture growth. This hypothesis inspired subsequent investigations on Fe nutrition and assimilation by phytoplankton (Hopkins and Wann, 1927; Goldberg, 1952; Ketchum, 1954), investigations that were limited by a lack of trace metal clean techniques and analytical sensitivity until the later 1900s. Early field studies identified Fe as a potential explanation for the differential between high phytoplankton productivity in coastal regions and low productivity in open ocean regions (Gran, 1931; Hart, 1934; Cooper, 1935; Harvey et al., 1935). In the summer of 1932, what began as a study investigating biology and chemistry of different areas within the Gulf of Maine ended as an Fe limitation hypothesis, due to the bloom of Skeletonema costatum that was observed upon the addition of a soil extract containing soluble Fe to N- and P-replete in situ cultures of the neritic diatom (Gran, 1933). Subsequent explorations of this theory involved Fe enrichment of diatom cultures with pieces of steel, soil extract, the iron food additive ferric ammonium citrate, and Fe-sulfur compounds (Gran, 1933; Cooper, 1935; Harvey et al., 1935). This early Fe nutrition research ultimately led to substantial theories about Fe supply and demand, precisely  2 that the phytoplankton demand for Fe could not be met by the meager Fe supply in solution.  1.2 The role of phytoplankton in the world ocean Marine phytoplankton are responsible for nearly half of the global primary productivity and play an important role in the global carbon cycle (Field et al., 1998). Actively photosynthesizing phytoplankton initiate the biological carbon pump, whereby inorganic CO2 is converted into organic solutes and particles (Falkowski et al., 1998). This conversion decreases surface water partial pressure of CO2, which is compensated for by a gradient of atmospheric CO2 into the ocean. Particulate C is later sequestered to the deep ocean via the death and sinking of organic matter. Deemed the most important players in this process, diatoms’ ability to aggregate with other cells and their heavy silicate cell walls facilitate efficient sinking (Michaels and Silver, 1988; Lochte et al., 1993; Buesseler et al., 1998; Bopp et al., 2005; Sarthou et al., 2005). As such, these relatively large, single-celled species of autotrophic phytoplankton serve an integral role in carbon sequestration. Additionally, diatoms alone account for at least 20 % of global primary production (Nelson et al., 1995; Smetacek, 1999). When nutrient concentrations are high, diatoms dominate due to their ability to rapidly take up nutrients, and are thus common in coastal and upwelling regions, as well as along melting sea ice edges in polar regions (Smetacek, 1999; Morel and Price, 2003; Bopp et al., 2005). Capitalizing on nutrient-replete environments by storing nutrients in vacuoles, these organisms are able to support continued cell divisions until they reach a point of severe nutrient limitation (Smetacek, 1999).  1.3 Phytoplankton physiological Fe requirement All phytoplankton have an obligate requirement for Fe, a micronutrient that can control  3 phytoplankton growth rate and biomass (Gran, 1933; Goldberg, 1952; Ryther and Kramer, 1961). This trace metal is required for many metabolic functions, such as photosynthesis and respiration, chlorophyll synthesis, nitrate reduction, and nitrogen fixation (Glover, 1977; Rueter and Ades, 1987; Raven, 1988; Raven et al., 1999). In linear photosynthetic electron transport (utilizing both Photosystems II and I), Fe is required by many thylakoid proteins, including ferredoxin and cytochromes b6-f and c6 (Raven et al., 1999). In nitrogen metabolism, Fe is required for the synthesis and operation of both nitrite and nitrate reductase enzymes, which are involved in NO3- assimilation (Cárdenas et al., 1972; Zumft, 1974; Geider and La Roche, 1994). Under Fe deficiency, pigment (chlorophyll a and carotenoid) production, cytochrome concentrations, growth rates (Glover, 1977; Greene et al., 1991), and Fe-containing photosystems I (PS I) and II (PS II) decrease (Raven et al., 1999). As a result of these physiological changes, fluorescence increases and light saturation occurs at lower irradiances, reducing the efficiency of photosynthesis (Rueter and Ades, 1987). In nutrient-replete conditions, one linear electron transport chain contains 23 Fe atoms, whereas in limiting conditions, Fe- containing catalysts can be substituted by Fe-free proteins, lessening this requirement to 20 Fe atoms (Raven, 1988; Raven et al., 1999). Similarly, nitrite and nitrate reductases are decreased under Fe limitation and NO3- assimilation rates subsequently decrease (Verstreate et al., 1980). The relative expression of Cu-containing plastocyanin and Fe-containing cytochrome c6 can change with availability of Cu and Fe in chlorophyta and cyanobacteria (Raven, 1988; Raven et al., 1999); the oceanic diatom Thalassiosira oceanica has also been shown to contain plastocyanin, thus reducing its Fe requirement (Peers and Price, 2006). Under Fe-limiting conditions, metal-free flavodoxin can be substituted for Fe-containing ferredoxin by cyanobacteria, chlorophyta, rhodophyta (Raven, 1988; Raven et al., 1999), and diatoms  4 (LaRoche et al., 1993; 1995). Finally, the ratio of high-Fe PS I to low-Fe PS II and the ratio of these photosystems to the cell can decrease under Fe deficiency (Raven et al., 1999), and has been shown to be intrinsically lower in oceanic diatoms and higher in coastal diatoms, suggesting physiological adaptation to Fe availability in these regions (Strzepek and Harrison, 2004). Though enzyme substitutions and downregulation of Fe-containing machinery can enhance survival under Fe limitation, these compensations reduce the efficiency of nitrate assimilation and primary productivity (Raven, 1988). Severe Fe deficiency cannot be properly compensated for, which eventually leads to cell death, followed by aggregation and sinking out of the mixed layer.  1.4 Geochemistry of Fe in oceans Concurrent with investigations of phytoplankton physiological Fe requirements, researchers also sought to elucidate biogeochemical cycling of this micronutrient. Field investigations established that the predominant, yet paltry, supplier of new dissolved Fe to most open ocean areas is often dust fallout from the atmosphere (Duce, 1986; Martin and Gordon, 1988; Duce and Tindale, 1991). The spatial distribution of dust fallout is extremely variable, as demonstrated by the North Atlantic Ocean, which receives 43 % of global dust input (Jickells et al., 2005) due to its proximity to the Sahara desert (Prospero, 1996). As such, Fe can be predominantly supplied via upwelling and lateral transport from continental shelves in regions such as the Equatorial Pacific (Johnson et al., 1997) and the subarctic Pacific (Lam et al. 2006; Lam and Bishop 2008), respectively. Along coastal regions, new iron is associated with upwelling, land drainage, and continental shelves which typically dictate higher Fe concentrations than the open ocean (Ryther and Kramer, 1961; Johnson et al., 1997). This  5 horizontal disparity in Fe availability affects not only greater primary production in the coasts, but also influences the species of phytoplankton that bloom, supporting growth of larger cells such as diatoms (Smetacek, 1999; Morel and Price, 2003; Bopp et al., 2005). Field studies have observed that an influx of Fe to Fe-limited areas leads to a community shift from small cyanobacteria to large diatoms (Coale et al., 1996; Cavender-Bares et al., 1999; de Baar and Boyd, 2000; de Baar et al., 2005). Phytoplankton Fe requirements also differ between regions; many oceanic phytoplankton have lower cell Fe quotas (intracellular Fe concentrations normalized to cellular C) than coastal phytoplankton (Brand, 1991; Sunda et al., 1991; Maldonado and Price, 1996; Strzepek and Harrison, 2004; Marchetti et al., 2009). Likewise, growth of oceanic species is less limited by low Fe concentrations than coastal species (Brand et al., 1983; Brand, 1991; Maldonado and Price, 1996; Marchetti et al., 2006). Unlike other nutrients, Fe does not show inter-ocean fractionation (Martin et al., 1993), whereby one would expect there to be less Fe in the Atlantic Ocean due to circulation of deep ocean water from the Atlantic to the Pacific (Martin et al., 1989). This is due to scavenging and the very short residence time (decades) of dissolved Fe in seawater (Parekh et al., 2004). The vertical distribution of dissolved Fe concentrations follows a nutrient-type depth profile, featuring depleted Fe concentrations at the surface due to phytoplankton internalization, yet differs from other nutrient profiles in its stable Fe profile throughout the deep ocean (Martin et al., 1989; Johnson et al., 1997). Regulating this unique vertical distribution is the complexation of Fe with various organic and inorganic compounds that span the world ocean (Rue and Bruland, 1997). Complicating the meager supply of Fe to the world ocean is its redox chemistry and solubility. Above pH 4, Fe is very insoluble in oxygenated seawater, and soluble Fe  6 concentrations rarely exceed a few nanomolar (Falkowski et al., 1998). An early study concluded that Fe was available to plankton only in its free ion state (Hopkins, 1930). This remained the paradigm of Fe bioavailability until analytical techniques improved, allowing for Fe uptake experiments with well-defined Fe speciation. Shortly thereafter, it was reported that the inorganically-bound fraction of Fe was bioavailable to phytoplankton (Anderson and Morel, 1982), yet this concept of inorganic Fe nutrition soon became the subject of a paradox. Over 99 % of total iron in seawater is scavenged from the dissolved phase by sinking particles to form oxyhydroxide minerals (Whitfield and Turner, 1987). The rest of total Fe (< 0.4 µm) is dissolved in seawater as one of three forms: free Fe(III) (< 0.1 %), inorganically bound (~ 1 %), or organically bound (~ 99 %) (Rue and Bruland, 1997). From his study compiling Fe measurements from the world ocean, Johnson et al. (1997) determined the average concentration of total dissolved Fe in the mixed layer to be ~ 0.07 nmol L-1, with most concentrations < 0.2 nmol L-1. The overwhelmingly dominant fraction of the dissolved Fe is predominantly complexed by strong organic ligands; field study has isolated these compounds from seawater (Macrellis et al., 2001) and identified their Fe-binding functional groups to be typical of siderophores (Witter et al., 2000; Barbeau et al., 2001), the most tenacious Fe chelators in nature (Neilands, 1995). Given the average total dissolved Fe concentration in the mixed layer, these complexes thus regulate an inorganic Fe(III) concentration of ~ 0.07 pmol L-1.  1.5 Siderophores Greek for “iron carrier”, siderophores are low-molecular-weight, tenacious Fe(III)- chelating agents synthesized by most microorganisms in aerobic environments to circumvent the insolubility of Fe (Raymond et al., 1984; Neilands, 1995). When bound with Fe, a six-coordinate  7 octahedral complex is formed via multiple sets of bidentate-coordinating groups with electron- donating atoms, typically oxygen (Neilands, 1981a; Raymond et al., 1984; Guerinot and Yi, 1994). Siderophores can be classified into one of three groups based on the chemical structure of the bidentate-coordinating groups that bind Fe: catecholates (Figure 1.1a), hydroxamates (Figure 1.1b), or a combination of these two functional groups (Raymond et al., 1984). The biosynthesis of siderophores is regulated by extracellular Fe concentrations. Siderophores are released into the media to extracellularly chelate Fe(III) and facilitate transport of Fe into the cell via Fe- siderophore specific membrane receptors (Neilands, 1981b). In particular, it has been observed that bacteria have a high Fe requirement for growth, and many bacteria species have been identified to produce siderophores, including Escherichia coli (O'Brien and Gibson, 1970), Pseudomonas sp. (Venturi et al., 1993), Vibrio sp. (Sigel and Payne, 1982), Rhizobia (Carson et al., 1992), and marine bacteria (Reid et al., 1993).   Figure 1.1: Chemical structures of a) a catecholate bidentate-coordinating group and b) a hydroxamate functional bidentate-coordinating group, both complexing Fe(III).  8 Siderophores are ubiquitous in marine microbial isolates (Trick, 1989), and are present in marine seawater via biosynthesis, as well as viral lysis and grazing of microbial cells (Witter et al., 2000). Multiple field studies in the subarctic Pacific Ocean in late summer have determined that under Fe limitation, marine bacteria can account for 20-70 % of Fe uptake (Tortell et al., 1996; Maldonado and Price, 1999) and have Fe quotas 1.5-2 times greater than those of phytoplankton (Maldonado and Price, 1999). Unsurprisingly, it has been observed that production of siderophore-like organic ligands in situ rapidly increases following small Fe additions to an Fe-deficient community (Rue and Bruland, 1997; de Baar et al., 2005). This response plays a considerable role in Fe bioavailability, as rapid binding of new Fe input prevents the micronutrient from precipitating out of solution (Sunda, 1997; Maldonado and Price, 1999). Yet, this crucial aspect of Fe solubility has its disadvantages, as those organisms that produce Fe-binding complexes also compete directly with phytoplankton for this resource (Tortell et al., 1996). In marine phytoplankton physiological research, the commercially available compound desferrioxamine B (DFB, desferrioxamine mesylate, Desferal®), a hydroxamate siderophore (Figure 1.2), is routinely used in Fe uptake experiments. This particular ligand is produced by the bacteria Streptomyces pilous and significantly decreases the pool of inorganic Fe, the most bioavailable pool for phytoplankton uptake (Wells et al., 1994). The high affinity of DFB for ferric iron is demonstrated by a high binding conditional stability constant of 16.5 M-1 (Hudson et al., 1992; Maldonado and Price, 1999).  9  Figure 1.2: Chemical structure of the hydroxamate siderophore desferrioxamine B (DFB) complexing Fe(III).  1.6 Fe uptake mechanisms of marine phytoplankton In addition to regulating the vertical distribution of Fe and the amount of total Fe in the dissolved fraction, the presence of excess siderophore-like organic compounds regulates an inorganic Fe concentration well below that needed to sustain phytoplankton growth (0.1-0.01 pmol L-1; Rue and Bruland, 1995; 1997). Based on the ambient dissolved Fe concentration observed in the equatorial Pacific Ocean (21 pmol L-1), Rue and Bruland (1997) calculated that these tenacious Fe-binding organic ligands regulate an equilibrium concentration of inorganic Fe of ~ 10 fmol L-1, a concentration far smaller than that which limits diffusive supply of inorganic Fe to the cell surface of a small phytoplankton (4 pmol L-1). Furthermore, a field study in the subarctic Pacific Ocean determined that Fe uptake rates by phytoplankton were much faster than the natural rate of Fe dissociation from marine ligands (Maldonado and Price, 1999). Given this inequality between inorganic Fe supply and demand, the paradigm of bioavailable Fe shifted. It  10 was evident that the organically-bound Fe pool must be accessible to phytoplankton by yet uncharacterized Fe uptake mechanisms (Maldonado et al., 2001). In the years following this paradigm shift, lab investigations illustrated that Fe-limited model centric diatom species Thalassiosira oceanica (an oceanic isolate) and Thalassiosira pseudonana (a coastal isolate) can access organically-bound Fe via a reductive process (Maldonado and Price, 1996; 1999; 2000; Maldonado et al., 2001; 2006). This process is similar to the reductive Fe uptake pathway of the baker’s yeast Saccharomyces cerevisiae (Dancis et al., 1994), in that a non-specific plasmalemma ferric reductase reduces the ligand-bound Fe(III) to Fe(II) releasing the Fe from the organic complex, after which a multi-copper oxidase reoxidizes the Fe(II) to Fe(III). Ferric Fe is then transported into the cell via a permease; the reduction and reoxidation steps allow for a high specificity of Fe uptake (Maldonado et al., 2006). This process facilitates “pirating” of Fe bound to siderophores, capitalizing on Fe acquisition strategies of microorganisms (Rue and Bruland, 1997). As such, the fraction of bioavailable Fe approaches the total dissolved Fe concentration in seawater (Morel, 2008). Interestingly, although Maldonado et al. (2006) demonstrated reductive Fe uptake by T. oceanica and T. pseudonana under Fe limitation, a disparity between the species was discovered. It was noted that Fe uptake by T. oceanica was two times faster than by T. pseudonana, despite having virtually equal oxidation and reduction rates in the reductive Fe uptake mechanism. In microorganisms such as yeast Saccharomyces cerevisiae and bacteria Aeromonas hydrophilia, additional Fe uptake mechanisms exist that would explain this disparity. Saccharomyces cerevisiae have been shown to transport the entire Fe-siderophore complex (Lesuisse and Labbe, 1989; Lesuisse et al., 1998; 2001), whereas Aeromonas hydrophilia (Stintzi et al., 2000) acquire Fe via a ligand exchange mechanism. In A. hydrophilia, a desferrated Fe chelator bound by a  11 receptor protein facilitates non-reductive Fe(III) exchange between the cell and a donor ferrated siderophore (Stintzi et al., 2000). Although it has long been considered that organically-bound Fe is not directly internalized by phytoplankton, this disparity must be considered and potential Fe uptake mechanisms by these organisms should be re-examined.  1.7 From physiology to field and future Overall, primary productivity is often severely limited by a lack of Fe; 30-40 % of the world ocean is Fe-limited, particularly in the Southern Ocean, and equatorial and subarctic Pacific Oceans (Moore et al., 2002). These three regions were classified as high nutrients, low chlorophyll (HNLC) regions in the early 1990s, as the re-introduction of the Fe hypothesis (Martin and Gordon, 1988; Martin, 1990; 1992) led to a boom in world ocean Fe-limitation research. In the field, elucidating phytoplankton Fe nutritional status becomes a complicated endeavor as one considers the myriad of other resources that influence Fe bioavailability. Research expeditions have explored regions such as the subarctic Pacific Ocean, Southern Ocean, and Bering Sea, and demonstrated that light, CO2, and N can co-limit primary productivity with Fe (Price et al., 1991; Maldonado et al., 1999; Tortell et al., 2008). Despite this progress, it is yet to be determined whether Fe plays a role in controlling primary productivity in the Arctic Ocean, a region which is changing drastically in response to global warming (Meehl et al., 2005; Smetacek and Nicol, 2005; Stroeve et al., 2005; 2007; Armbrust, 2009). Previously, it has been shown that low light and NO3- levels control primary productivity in the spring and late summer, respectively (Sherr et al., 2003; Hill et al., 2005). Sea ice is integral to these seasonal changes as warmer atmospheric temperatures and increased irradiance melt ice in spring, allowing light to penetrate surface waters. This process facilitates a spring bloom of plankton,  12 drawing down NO3- concentrations that accumulated over winter (Hill et al., 2005). Sea ice also plays a vital role in Fe availability to the high north, precluding direct deposition of aeolian transport in surface waters and entraining Fe-containing suspended sediments during the formation of pack ice, both of which are supplied to surface waters when the melting season begins (Measures, 1999; Isaksson et al., 2003; Krachler et al., 2005). Future field expeditions should investigate whether the seasonal fluxes of NO3- and light interacts with Fe in controlling growth of Arctic phytoplankton as they have in the HNLC Southern Ocean and subarctic Pacific Ocean (Maldonado et al., 1999; Lancelot et al., 2000).  1.8 Research objectives Many questions remain about Fe acquisition mechanisms by phytoplankton and the extent of Fe limitation of phytoplankton growth in the world ocean. Understanding the dynamic relationship between phytoplankton and Fe is imperative for predicting the fate of phytoplankton communities and the biological carbon pump as our oceans respond to climate change. As such, the overarching goal of this thesis is to address questions of Fe nutrition in phytoplankton, both physiological and biogeochemical. This dissertation addresses two relevant questions in the realm of Fe bioavailability to phytoplankton: can phytoplankton acquire organically-bound Fe with a non-reductive uptake mechanism, and does Fe limit (or co-limit) phytoplankton growth in the Arctic Ocean? In the summer of 2009, I conducted an 8-day grow out experiment with an indigenous plankton community in the Beaufort Sea, during which light, Fe, and NO3- levels were manipulated. In the following year, I investigated the potential of non-reductive Fe uptake by T. oceanica, and sought to explore the existence of this physiological mechanism along Line P in the subarctic  13 Pacific Ocean. These investigations have enhanced my understanding of Fe nutrition by phytoplankton. Additionally, the research presented here will contribute to the broader spectrum of accurately modeling changes in oceanic primary productivity and carbon export in the face of global warming, as it describes novel Fe-dependent factors such as acquisition mechanisms and Fe nutritional status in the north Pacific Ocean and the Beaufort Sea, respectively.    14 Chapter 2: Co-limitation by Light, Nitrate and Iron in the Beaufort Sea  2.1 Introduction The Arctic Ocean is undergoing profound changes in response to global warming. Due to the climate sensitivity of sea ice, this large polar expanse is expected to change disproportionately rapid in comparison with the rest of the world (Meehl et al., 2005; Smetacek and Nicol, 2005; Stroeve et al., 2005; 2007; Armbrust, 2009). Already, increasing atmospheric CO2 concentrations have affected numerous changes in physical parameters in the ocean waters of the high north. Reduced sea ice thickness (Rothrock et al., 2003) and extent (Belchansky et al., 2005) have facilitated increased absorption of solar radiation by seawater and a subsequent increase in sea surface temperature (Perovich et al., 2007). In turn, surface water heating will affect a warmer atmosphere via albedo feedback (Morales-Maqueda et al., 1999), thereby facilitating a positive feedback system. Additionally, Arctic surface waters have freshened due to increased river discharge and sea ice melt (Peterson et al., 2002; Yamamoto-Kawai et al., 2009). The increase in freshwater input has strengthened stratification and shoaled mixed layer depths (Toole et al., 2010), changes that will continue to strengthen with time and gradual warming. Unsurprisingly, this myriad of both present and future physical changes has led to predictions about Arctic primary productivity. It is expected that rising surface water temperatures will favour picoplankton growth, as increased stratification will limit nutrient availability (Tremblay et al., 2009). These small organisms are typically recycled in the microbial food web (Smetacek, 1999), and are thus expected to reduce the efficiency of the biological carbon pump, a process that converts inorganic atmospheric CO2 to organic particles that are eventually sequestered to the deep ocean via the death and sinking of organic matter (Falkowski et al., 1998). Additionally, due to thinned and melted sea ice, light penetration may  15 enhance productivity, although increased stratification could limit this phenomena, particularly for diatoms that depend on high silicate- and nitrate-rich waters (Armbrust, 2009). Indeed, the future of Arctic primary productivity is less lucid than that of predicted physical changes. Presently, some areas of the Arctic Ocean exhibit low primary productivity, which has traditionally been attributed to low light levels due to sea ice cover (Sherr et al., 2003; Hill et al., 2005; Tremblay et al., 2006) and the polar night that lasts the winter months (Tremblay et al., 2006). In brighter seasons, field studies in the Chukchi Plateau and western Beaufort Sea have shown that low light and nitrogen levels control primary productivity in these regions in the spring and late summer, respectively (Sherr et al., 2003; Hill et al., 2005). Warmer atmospheric temperatures and increased irradiance melt ice in spring, allowing light to penetrate and facilitate a spring bloom of plankton (Hill et al., 2005). As the summer months progress, nitrogen, typically nitrate (NO3-), becomes the limiting factor due to depletion by growing plankton and stratification, which limits the supply to surface waters (Hill et al., 2005). Nitrate has been identified as the largest fraction of dissolved inorganic nitrogen (DIN) available in the euphotic zone of the Beaufort Sea (Sherr et al., 2003; Hill et al., 2005; Yamamoto-Kawai et al., 2006; Tremblay et al., 2009), contributing to a total DIN of < 0.8 µmol L-1 in surface and intermediate depths (Tremblay et al., 2009). Similarly located field studies have characterized winter NO3- concentrations in the surface waters of the Beaufort Gyre to be as low as 2 µmol L-1 (Tremblay et al., 2006), due to the influx of low-nitrogen waters from the Bering Sea (Varela and Harrison, 1999) that undergo further denitrification over the extensive continental shelves of the Bering and Chukchi Seas (Yang et al., 2004; Yamamoto-Kawai et al., 2006; 2008). Despite these meager concentrations, primary productivity in the Arctic is considered to be primarily NO3- based, reflected in an f-ratio > 0.5 when NO3- concentrations exceed 0.3 µmol L-1 (Tremblay et  16 al., 2006). Globally, primary productivity is largely affected by iron (Fe) inputs, an obligate component of many metabolic functions such as photosynthetic and respiratory electron transport, chlorophyll a synthesis, and nitrogen assimilation (Raven, 1988; Raven et al., 1999). In the high north, aeolian transport and direct deposition of Fe-laden dust in surface waters is precluded by perennial sea ice cover (Measures, 1999). However, sea ice indirectly enhances the Fe concentration of surface waters as it can accumulate this trace metal both above and below the ice (Tovar-Sanchez et al., 2010). On the surface, sea ice collects dust via dry or wet atmospheric deposition, later reaching Arctic waters via seasonal melting (Measures, 1999; Isaksson et al., 2003; Krachler et al., 2005; Tovar-Sanchez et al., 2010). Below the water, suspended sediments entrained during the formation of pack ice can be later mixed into surface waters via summer ice melt and brine channels (Measures, 1999; Tovar-Sanchez et al., 2010). Through the Bering Strait, Fe is supplied to the Chukchi Sea via ice melt and shelf water (Smetacek and Nicol, 2005). Along the shelf region of the Chukchi Sea, a relatively high Fe concentration of 20 nmol L-1 has been observed due to these sources (Measures, 1999). High input close to the coasts is typically scavenged and does not reach the open ocean gyres (Measures, 1999), though lateral transfer over the continental shelves, which represent ~ 30 % of the Arctic Ocean’s surface area (Guéguen et al., 2007), may supply these open waters with Fe from the shelf sediments (Nakayama et al., 2011). Moving eastward, approximately 10 % of the river input to the Arctic is supplied by the Mackenzie River into the Beaufort Sea (Measures, 1999), contributing to a year- round Fe source in Arctic waters. Along a transect from the mouth of the Mackenzie River to the Canada Basin of the Beaufort Sea, average surface water dissolved Fe concentrations ranged 0.8 – 1.3 nmol L-1 and decreased with distance from shore (Giesbrecht et al., 2010). This value is on  17 par, yet slightly lower, than concentrations reported for the Eurasian Basin, a neighbouring open ocean gyre in the Arctic Ocean along the northern coasts of Russia (Measures, 1999). As of yet, it is unknown whether the geochemical and atmospheric supplies of Fe to the Arctic Ocean limit primary productivity. Furthermore, the seasonal effects of light and NO3- may interact with Fe in controlling growth of Arctic phytoplankton as they have in the Southern Ocean and subarctic Pacific Ocean (Maldonado et al., 1999; Lancelot et al., 2000). Elucidating whether Fe independently or interactively controls Arctic primary productivity is crucial to our understanding of the efficiency of the biological carbon pump in these northern waters, both currently and in predicting future effects of climate change. This need is made all the more imperative as it has been suggested that Fe uptake may decrease due to decreasing seawater pH, which subsequently reduces inorganic Fe concentrations and the effectiveness of redox processes involved in Fe uptake from organic complexes (Shi et al., 2010). To illuminate the potential of Fe limitation in the Arctic Ocean, we investigated the controls of primary production in the Beaufort Sea as part of the GEOTRACES International Polar Year program. During a two-week expedition in late summer 2009, an on-deck incubation experiment was conducted, during which light levels and concentrations of Fe and NO3- were manipulated.  2.2 Materials and methods 2.2.1  Station parameters and experimental design Along the CCGS Amundsen’s transect from the shallow Mackenzie River delta (58 m) to the deep Canada Basin (3485 m) in the Beaufort Sea, an 8-day grow-out incubation experiment was initiated at Station L1 (71°05.94’ N, 139°08.87’ W; 1918 m; Figure 2.1). This station lay above the extensive Arctic continental slope, yet out of reach of the Mackenzie River nutrient  18  Figure 2.1: Bathymetric map of the Beaufort Sea and Canadian Arctic Continental Shelf. Station L1, (71°05.94’ N, 139°08.87’ W; 1918 m depth) is located north of Yukon, Canada and west of Banks Island. Figure produced using Ocean Data View (Schilitzer, 2004).  input. On August 31, 2009, at 22:25 Mountain Daylight Time (MDT), a trace metal clean Teflon® peristaltic pump and tubing system, suspended by a Kevlar® line, was used to pump seawater from 29 m depth into a trace metal clean laminar flow hood within a clean room. This depth was approximately 20 m from the base of the ship, eliminating the potential for trace metal contamination by the hull. Unfiltered seawater was transferred to trace metal clean polycarbonate bottles and low-density polyethelene 10 L cubitainers. Earlier during daylight hours at 14:22 MDT, many biological, chemical, and physical parameters of Station L1 were assessed via a rosette equipped with Niskin bottles (OceanTest Equipment) and a Sea-Bird 911plus CTD probe. Surface photosynthetically active radiation (SPAR; 400-700 nm) was measured with a LI-COR  19 sensor (L1-190SA), underwater photosynthetically active radiation (PAR) profiles were determined with a Biospherical Instruments light sensor (QCP-2300), and transmissivity was assessed with a Wet Labs C-Star transmissometer. Macronutrient and chlorophyll a samples were obtained immediately after the rosette cast from the Niskin bottles; these samples were analyzed on board the icebreaker as per the protocol described below. For the 8-day experiment, ten treatments were tested in duplicate, including three light levels and three Fe levels (Table 2.1). A gradient of Fe sufficient (+ Fe) to putative Fe limitation (in situ Fe) to Fe deficiency (+ DFB) was simulated. Iron deficient conditions were simulated via the addition of 1 nmol L-1 of the tenacious bacterial siderophore desferrioxamine B (DFB), thereby reducing the equilibrium concentrations of free and inorganically bound Fe for the duration of the incubation. To complement this treatment, an Fe-sufficient treatment was created by amending the in situ seawater with 1 nmol L-1 Fe bound to 2 nmol L-1 of the synthetic organic ligand ethylenediaminetetraacetic acid (EDTA) to prevent Fe precipitation. Preparation of this FeEDTA complex was done as described in Maldonado and Price (1999); nearly 24 hours following complexation, treatment bottles were amended with the Fe-ligand compound.  Table 2.1: Control and experimental treatments (marked by X) and their corresponding light, iron, and nitrate levels.    Light level   1 %Io 10 %Io 50 %Io + 1 nM Fe  X  X  X no Fe added  X X X  X Fe level + 1 nM DFB   X  X  X   no N added +10 µM N no N added +10 µM N no N added +10 µM N   Nitrate level    20 In addition to manipulating Fe concentrations, the nitrogen content of our experimental treatments was adjusted. Based on the low NO3- concentration measured at 22 m depth at Station L1 (0.13 µmol L-1), we assumed the community to be nitrogen limited. As a result, all treatments apart from the control treatment were enriched with 10 µmol L-1 chelexed NO3- (Table 2.1). Once all treatment additions were completed, cubitainers were sealed with Parafilm® and placed in deck incubators covered with neutral density screening to correspond to three light levels: 50, 10, and 1 %Io. Continuously flowing seawater pumped from 5 m depth maintained in situ temperatures (~ −1.1 °C) within the incubators.  2.2.2 Biological and chemical parameters Bottles were sampled on days 0, 2, 4, 6, and 8 for total chlorophyll a, heterotrophic and autotrophic bacteria flow cytometry, and nutrients. On days 0 and 8, additional samples for high performance liquid chromatography (HPLC) pigments, light microscopy phytoplankton identification, and size-fractionated chlorophyll a were collected. Total chlorophyll a (chl a) samples were collected on 0.22 µm Poretics® polycarbonate membrane filters by gentle vacuum (< 100 mm Hg) and subsequently frozen at −80 °C. In addition, to determine size-fractionated chl a, 0.22, 1, and 5 µm Poretics® polycarbonate membrane filters separated by Millipore nylon mesh drain disks were used. While on the icebreaker, batches of individually-preserved membrane filters were treated with 90 % acetone to extract chl a. Following a 24-hour extraction period during which samples were kept in the dark at −20 °C, concentration of chl a in the acetone was determined with a Water Properties Turner 10 AU fluorometer. Samples for macronutrients (nitrate, NO3-, phosphate, PO43-, and silicate, SiO42-) were collected and analyzed on board the icebreaker, based on Barwell-Clarke and Whitney (1996). Dissolved Fe samples  21 were collected by Jay Cullen’s group and later analyzed in his lab at the University of Victoria; analysis was based on Obata et al. (1993) and De Jong et al. (1998). HPLC pigment samples (~ 4 L) were filtered with Whatman® GF/F filters by gentle vacuum (< 100 mm Hg) and immediately frozen at −80 °C. Upon return to The University of British Columbia (UBC), samples were processed with a Waters Alliance BIO 2796 Bioseparations Module. To calibrate pigment extraction yield, 10 µL of the pigment internal standard TAC were added to each filter. Within a glass vial, 2-3 mL of 90 % acetone were added to sample filters and sonicated for 5 minutes to disrupt cells and extract pigments. Following this, the pigment extract was filtered with a 1 mL syringe (Fisher B309659) and a solvent- resistant 0.2 µm PTFE filter (Fisher) to remove debris. A mixture of 250 µL filtered pigment extract, 50 µL deionized water, and 50 µL acetonitrile was injected into the reversed-phase C8 column of the HPLC system and analyzed according to the protocol outlined in Zapata et al. (2000). Eluent A was a 50:25:50 v:v:v mixture of methanol : acetonitrile : 0.25 M aqueous pyridine solution, and Eluent B was a 20:60:20 v:v:v mixture of methanol : acetonitrile : acetone. Pigments were identified using a mixed pigments standard containing 23 different pigments dissolved in 90 % acetone (DHI Water & Environment). Concentrations for each identified pigment were then quantified with standard curves. Following sample processing, data were interpreted with the chemotaxonomy program CHEMTAX V1.95, an updated version from that first presented in Mackey et al. (1996). This software utilizes the concentrations of biomarker pigments detected in HPLC analysis to determine the contribution of each phytoplankton group to community biomass. The initial pigment ratios matrix (Table 2.2a), which expresses the unique combination of pigment:chl a ratios for each algal group, was derived from a compilation of numerous pigment investigations throughout the world ocean (Wright et al., 1996; Mackey et  22 Table 2.2: Matrices of pigment : chlorophyll a ratios for eight algal groups utilized in CHEMTAX analysis. The initial pigment ratios matrix (A) was derived from a compilation of numerous pigment investigations throughout the world ocean (Wright et al., 1996; Mackey et al., 1998; Rodríguez et al., 2002; Suzuki et al., 2002; Vidussi et al., 2004; Latasa, 2007; Hashihama et al., 2008; Miki et al., 2008; Fujiki et al., 2009; Kim et al., 2010; Wright et al., 2010; Wright, personal communication). The final pigment ratios matrix (B) represents an average of the best 10 % of 60 iterative matrices of the initial pigment ratios matrix for each analyzed sample, based on the RMS of each estimate; these averaged best matrices varied less than 10 %. Pigment annotation: Chlc3, chlorophyll c3; MgDVP, magnesium 3,8-divinylphaeoporphyrin a5 monomethyl ester; Chlc2, chlorophyll c2; Per, peridinin; Butfuc, 19’-butanoyloxyfucoxanthin; Fucox, fucoxanthin; Neox, neoxanthin; Pras, prasinoxanthin; Violax, violaxanthin; Hexfuc, 19’-hexanoyloxyfucoxanthin; Ddx, diadinoxanthin; Allox, alloxanthin; Zeax, zeaxanthin; Lut, lutein; Chlb, chlorophyll b; Chla, chlorophyll a.    A Chlc3 MgDVP Chlc2 Per Butfuc Fucox Neox Pras Violax Hexfuc Ddx Allox Zeax Lut Chlb Chla Diatoms  0.0011 0.19   0.77     0.26     1.0 Dinoflagellates  0.001 0.22 0.79       0.22     1.0 Prymnesiophytes 0.038     0.42    1.5 0.14     1.0 Chlorophytes  0.0017     0.081  0.044    0.074 0.17 0.31 1.0 Prasinophytes Type 1  0.67     0.073 0.24 0.091    0.035 0.011 0.92 1.0 Prasinophytes Type 2             0.0080 0.006 0.74 1.0 Cryptophytes  0.0013 0.17         0.26    1.0 Pelagophytes 0.11 0.0013 0.15  0.99 0.50 0.0050  0.11 0.057 0.15 0.32 0.075 0.12 0.22 1.0 Cyanobacteria             0.56   1.0 B Chlc3 MgDVP Chlc2 Per Butfuc Fucox Neox Pras Violax Hexfuc Ddx Allox Zeax Lut Chlb Chla Diatoms  0.00099 0.20   0.80     0.27     1.0 Dinoflagellates  0.00095 0.25 0.83       0.22     1.0 Prymnesiophytes 0.036     0.42    1.1 0.14     1.0 Chlorophytes  0.0019     0.083  0.045    0.074 0.16 0.32 1.0 Prasinophytes Type 1  0.61     0.070 0.25 0.094    0.035 0.011 0.91 1.0 Prasinophytes Type 2             0.0084 0.0062 0.73 1.0 Cryptophytes  0.0011 0.16         0.27    1.0 Pelagophytes 0.12 0.0013 0.15  0.99 0.51 0.0050  0.11 0.060 0.15 0.32 0.075 0.12 0.22 1.0 Cyanobacteria             0.51   1.0  23 al., 1998; Rodríguez et al., 2002; Suzuki et al., 2002; Vidussi et al., 2004; Latasa, 2007; Hashihama et al., 2008; Miki et al., 2008; Fujiki et al., 2009; Kim et al., 2010; Wright et al., 2010; Wright, personal communication). Utilizing this initial ratio matrix, the software used factor analysis and a steepest descent algorithm to best fit the Station L1 data. Sixty iterations of the initial pigment ratios matrix were calculated to avoid the potential for poor starting choices producing unrepresentative results. These matrices were obtained by random adjustments of each pigment ratio in Table 2.2a by ± 35 % and were applied to each sample, from which the best 10 %, based on the RMS of each estimate, were used to calculate the mean % abundance of 8 algal groups: diatoms, dinoflagellates, prymnesiophytes, chlorophytes, prasinophytes, cryptophytes, pelagophytes, and cyanobacteria. The final pigment matrix, representing an average of the best 10 % matrices, is presented in Table 2.2b. Most of the final pigment ratios differed very little from the initial pigment ratios matrix with the exception of 19’-hexanoyloxyfucoxanthin, which fell from 1.5 to 1.0, but was still within the 35 % variation limit. Light microscopy phytoplankton identification samples were collected in 250 mL amber bottles and preserved with 0.2 % gluteraldehyde. Due to freezing temperatures coupled with on- deck storage, few phytoplankton composition samples survived the Amundsen’s voyage to port. Of the 27 samples that were taken (1 sample for each treatment bottle on day 8 plus 3 initials), 6 survived; these were enumerated with a Zeiss Axiovert 10 microscope. Triplicate samples for heterotrophic and autotrophic bacteria flow cytometry were dispensed into cryovials containing a 10 % formaldehyde, 0.5 % gluteraldehyde and 40 % phosphate-buffered fixative solution based on Marie et al. (2005). These 1 mL samples were immediately stored in the dark for 15 minutes, after which they were flash frozen in liquid nitrogen and stored at !80 °C. Upon return to UBC, a FACSCalibur flow cytometer (Becton-  24 Dickinson) equipped with a 15-mW, 488-nm laser was used to count and characterize heterotrophic and autotrophic bacteria flow cytometry samples. After thawing to above-freezing temperatures, 50 "L were taken for heterotrophic bacteria analysis and the remainder was analyzed for autotrophic bacteria. An addition of FluoroSpheres carboxylate modified yellow- green 1-"m beads (Molecular Probes; ~ 2 # 10-5 mL-1 final concentration) was made to both subsamples as an internal reference. Heterotrophic bacteria flow cytometry subsamples were then diluted 10-fold in 0.02 "m pore-size filtered TE buffer (10 mmol L-1 Tris, 1 nmol L-1 EDTA, pH 8.0) to avoid coincidence and stained with SYBR Green-I (Invitrogen). Heterotrophic and autotrophic bacteria samples were then incubated and analyzed as per Marie et al. (1999) and Martinez-Martinez et al. (2006). The list files generated by the instrument were then analyzed with Cytowin software (Vaulot, 1989), which distinguishes organism populations based on their natural fluorescence and scatter of instrument lasers. The discrimination of autotrophic bacteria was done via scatter plots of orange and red fluorescence; heterotrophic bacteria were quantified via scatter plots of side scatter signal vs. SYBR Green-I fluorescence. Cell density was calculated from scatter plot cell counts, instrument flow rate, and analysis time. Due to expectations of low biomass, day 2 flow cytometry samples were not processed.  2.2.3  Fe uptake rates and Fe:C ratios Short-term Fe uptake rates ($Fe) and Fe:C ratios were determined on day 6 of the grow out using 55FeCl3 (4.95 # 103 Ci mol-1, PerkinElmer) and H14CO3- (9.19 # 10-3 Ci mol-1, PerkinElmer). Rates of Fe uptake were measured in the presence of Fe bound to either DFB or EDTA. Iron bound to DFB is only taken up by Fe-limited phytoplankton (Maldonado and Price, 1999); thus rates of Fe uptake from FeDFB were of interest to provide information about the Fe  25 status of the plankton. Additionally, iron uptake rates from FeDFB ($FeDFB) mimicked in situ Fe uptake rates, as the free inorganic Fe concentration was not disturbed over the short-term (24 h) duration of these experiments. In contrast, the short-term Fe uptake rates from FeEDTA ($FeEDTA) were determined to investigate saturating rates of Fe uptake, which are a direct function of the density of plasmalemma-bound Fe transporters and thus faster for Fe limited phytoplankton (Harrison and Morel, 1986). The combination of $FeDFB and $FeEDTA allowed us to elucidate the Fe deficiency state of the in situ plankton community. Seawater from each of the 20 treatment bottles was dispensed into triplicate trace metal clean 500 mL polycarbonate bottles for each of the ligands and enriched with 14C (20 # 10-6 Ci L-1) and 55Fe within a laminar flow hood. 55FeCl3 was added to the bottles in ratios of 100 nmol L-1 Fe : 105 nmol L-1 DFB with 5 % of total Fe as 55Fe, or 100 nmol L-1 Fe : 1000 nmol L-1 EDTA with 1 % of total Fe as 55Fe. Preparation of these 55Fe-ligand complexes was done as described in Maldonado and Price (1999). DFB was added in 5 % excess of Fe so as to mimic the relative abundance of Fe and its binding organic ligands in seawater. High concentrations of EDTA were chosen so that competitive complexation by DFB in the + 1 nM DFB treatment bottles would be negligible. The high concentrations of Fe : ligand permitted us to calculate Fe uptake rates without a priori knowledge of the in situ Fe concentration. Following addition of isotopes, bottles were tightly capped, sealed with Parafilm®, and returned to the incubators. In situ temperature and light levels were maintained as previously described. Following 24 hours of incubation, the plankton were filtered onto 0.22, 1, and 5 "m pore size Poretics® polycarbonate membrane filters separated by nylon drain disks. Before going dry, filters were soaked with 5 mL of 0.2 "m filtered Ti(III) citrate EDTA reagent for 5 minutes to dissolve Fe oxides adsorbed to cell surfaces (Hudson and Morel, 1989), then rinsed with 5 mL of  26 0.2 "m filtered seawater. Following this, the filters were transferred to scintillation vials and saturated with ScintiSafe Plus 50 % scintillation cocktail (Fisher). Particulate 55Fe and 14C activities were determined with a Beckman Coulter LS 6500 Multi-Purpose Scintillation Counter upon return to UBC two weeks later. Assuming linear Fe uptake during the 24-hour incubation period (Maldonado and Price, 1999), size-fractionated volumetric Fe uptake rates (mol Fe L-1 d- 1) were calculated via total particulate Fe (mol Fe L-1) and the distinct incubation time for each bottle. Uptake rates were also carbon-normalized (mol Fe (mol C)-1 d-1) via size-fractionated day 6 chl a (ng L-1) and a C:chl a ratio of 50 g g-1. Finally, assimilation Fe:C ratios were also calculated from the 55FeDFB and 14C uptake rates (mol L-1 d-1) determined on day 6.  2.2.4 Statistical analyses Two research questions guided the statistical analyses of this experiment. First, we sought evidence of light or Fe limitation, as well as an interactive effect of these two factors, on growth of the indigenous plankton communities at Station L1. Specifically, we were interested in determining whether plankton growth was enhanced due to Fe additions to seawater relative to in situ Fe levels, and whether varying light levels would emphasize these differences. Thus, it was hypothesized that the + 1 nM Fe treatments would enhance plankton growth in the Beaufort Sea. To test for this hypothesis, a 2-way ANOVA was performed for the no Fe added and + 1 nM Fe treatment groups at all three light levels. This test was done separately for each size fraction when appropriate. Second, to further elucidate the interaction between Fe and light, the treatment designed to induce Fe deficiency (+ 1 nM DFB) was also included in a 2-way ANOVA test. Thus, all three light levels of the no Fe added, + 1 nM Fe, and + 1 nM DFB treatment groups were subjected to  27 a two-way ANOVA and subsequent Tukey testing to determine the significance of light and Fe on various biological and chemical variables associated with growth. As above, this test was done separately for each size fraction when appropriate. Additional ANOVA testing was performed to assess the significance of the bivariate fit between flow cytometry cell density and chlorophyll a data. All statistical analyses were conducted using JMP software.  2.3 Results 2.3.1 Overview of results Station L1 was characterized by small cells, with ~ 30 % of initial chl a exceeding 5 µm (Table 2.3), and was dominated by chlorophytes and prasinophytes (Table 2.6). The pumping depth (29 m) for this incubation lay within an intermediate layer marked by the base of the mixed layer (11.6 m) to the base of the euphotic zone (72 m); this layer encompassed maximum dissolved oxygen, chl a, fluorescence, and minimum macronutrients (Figure 2.2). At 30 m depth, dissolved Fe concentration was found to be low at 0.154 nmol L-1 (J. Cullen, personal communication). By the final day of the grow out, all treatments with the NO3- addition (10 µmol L-1) had increased chl a relative to the control treatment (no NO3- added; Table 2.3), supporting previous field studies that suggest NO3- limits phytoplankton growth in summer in the Beaufort Sea (Sherr et al., 2003; Hill et al., 2005). In the treatments with the 10 µmol L-1 NO3- addition, co-limitation by Fe and light was also observed, as total chl a concentrations of the + 1 nM Fe and no Fe added treatments were virtually equal at 50 %Io (900 ± 30 ng L-1 and 860 ± 89 ng L-1 respectively) but differed at 10 %Io (550 ± 33 ng L-1 and 390 ± 92 ng L-1 respectively; Figure  28 2.3). At 1 %Io, the no Fe added and + 1 nM Fe treatments yielded comparatively low total chl a (180 ± 9.5 and 110 ± 33 ng L-1, respectively), indicative of insufficient light for growth regardless of Fe levels. The higher Fe requirements associated with growth at low irradiance were typically supported by 1-2 orders of magnitude higher Fe:C ratios in treatments at 1 %Io relative to those at 50 and 10 %Io (Table 2.8), as well as the fastest rates of organically-bound Fe uptake (Table 2.7). Furthermore, the Fe:C ratios of the + 1 nM Fe treatment were ~ 2-fold higher than the no Fe added treatment at 1 %Io, indicating severe Fe deficiency at in situ Fe (0.154 nmol L-1) and low light (Table 2.8).  2.3.2 Station parameters During the CTD cast and seawater sampling, the weather conditions were mild, with blue skies and scattered clouds. Station L1  (Figure 2.1) was located amidst an area of open water with a large ice floe (~ 5-6 m thick). Surrounding waters were covered (75 %) with drifting, broken, multiyear ice, which was approximately 4 m thick (K. Brown, personal communication). Incident irradiance (Io) was moderate; based on CTD data from the upper 150 m, the average SPAR estimate was 513.73 µE m-2 s-1. The base of the euphotic zone, defined as the depth at which the light level is 1 %Io, was found at a depth of ~ 72 m. Density data (Figure 2.2a) from multiple CTD casts at L1 were used to estimate mixed layer depth (MLD). Using a density difference criteria of 0.125 from the surface (after Levitus, 1982), MLD was determined to be shallow at 11.6 ± 1.5 m. Salinity in the mixed layer was ~ 25.6 (data not shown). Below this, salinity steadily increased to a maximum of ~ 34.9 which persisted throughout the bottom 700 – 1902 m of the deep ocean. Within the upper 150 m, temperature ranged from !1.1 to !0.55 °C (data not shown). At the MLD (11.6 m), the  29  Figure 2.2: Vertical profiles of (a) density (kg/m3), oxygen (mL L-1) (b) transmissivity (%), PAR (!E m-2 s-1), fluorescence, (c) chlorophyll a (ng L-1), nitrate (!mol L-1), phosphate (!mol L-1), and silicate (!mol L-1) in the upper 150 m at station L1 (71°05.94’ N, 139°08.87’ W) in the Beaufort Sea. Data points in figures (a) and (b) are from a CTD cast conducted at 16:22 MDT on August 31, 2009. Chlorophyll a and nutrient data in figure (c) were obtained from samples collected from Niskin bottles on a rosette deployed with the same CTD cast.  30 temperature and irradiance levels were !1.1 °C and 22 %Io, respectively. At our seawater pumping depth (29 m), the temperature and irradiance levels were !0.9 °C and 6 %Io (30.82 µE m-2 s-1), respectively.  Dissolved oxygen reached a maximum at the base of the mixed layer and persisted to ~ 80 m below which it decreased with depth (Figure 2.2a), indicating the predominance of phytoplankton throughout this intermediate layer. Though this layer excluded the MLD, it encompassed the pumping depth (29 m). Transmissivity was lowest at MLD, indicating an accumulation of particles and organisms at the pycnocline (Figure 2.2b). Transmissivity reached a second minimum at 72 m depth, the base of the euphotic zone, which coincided with a peak in fluorescence. Despite these peaks, transmissivity did not vary by more than 1 % (88.125 - 89.125 %), indicating that primary production was generally low throughout the euphotic zone. Indeed, bottle sampling for chl a revealed very low phytoplankton biomass at 22 m (50 ng L-1) and at 50 m (193 ng L-1; Figure 2.2c). In addition, NO3-, PO43-, and SiO42- concentrations were very low at both 22 m (0.13, 0.66, and 1.8 "mol L-1, respectively) and 50 m (0.16, 0.76, and 2.7 "mol L-1, respectively). Below 50 m, macronutrient concentrations increased by an order of magnitude, indicating the presence of the nutricline between the bottle sampling depths of 50 and 85 m (Figure 2.2c). Similarly, chl a concentrations were highest at 50 m, indicating that the depth of the chlorophyll a maximum lay between 22 and 85 m depth. Finally, dissolved Fe concentration at 30 m depth was 0.154 ± 0.022 nmol L-1 (J. Cullen, personal communication).  2.3.3 Plankton community growth and limitation by nitrate or light  By day 8 of the grow out, total chl a concentrations for all treatments (ranging from 97 to 900 ng L-1) were still low but had increased relative to the initial chl a level (Table 2.3). The  31 Table 2.3: Average size-fractionated and total chlorophyll a (ng L-1) for triplicate initials on day 0 and duplicate samples for control and experimental treatments on the final day of an 8-day grow out initiated at station L1 (71°05.94’ N, 139°08.87’ W) in the Beaufort Sea. Values reported equal the mean ± standard error for initials and the range for day 8 duplicates; absence of a range indicates the treatment was only analyzed in singlet. Values in brackets represent the times increase from the control treatment.   initial community at station L1 was dominated by cells < 5 "m, as only ~ 30 % of total chl a exceeded 5 "m. By day 8, the greatest chl a concentrations were observed in the 0.22-1 "m size fraction for all treatments (Table 2.3). The lowest total chl a concentration was observed in the control treatment (Table 2.3); all experimental treatments exhibited total chl a concentrations at least 1.7 times greater than the control treatment, up to an excess of 14 times. Furthermore, the no Fe added 10 %Io treatment, which differs from the control treatment only in its amended NO3- concentration (0.43 versus 10 "mol L-1), showed an order of magnitude higher total chl a by day 8 (Figure 2.3a). Similarly, the treatment with + 1 nM DFB at 10 %Io (and 10 "mol L-1 NO3-) exhibited total chl a more than twice as high as the control, despite a lower inorganic Fe concentration (Table 2.3). These results indicate that the growth of the in situ plankton community was NO3- limited at this time of the year (August/September). Treatment 0.22-1 µm 1-5 µm > 5 µm Total Initial 9.8 ± 1.1 19 ± 1.9 13 ± 0.69 42 ± 3.3 Control 10 %Io, no NO3- 35 ± 6.8 21 ± 1.5 6.7 ± 0.31 63 ± 8.6 no Fe added 50 %Io 440 ± 53 (13) 260 ± 28 (12) 160 ± 7.6 (24) 860 ± 89 (14) no Fe added 10 %Io 160 ± 100 (5) 150 ± 0.71 (7) 81 ± 9.0 (12) 390 ± 92 (6) no Fe added 1 %Io 96 ± 3.3 (3) 50 ± 5.5 (2) 29 ± 0.67 (4) 180 ± 9.5 (3)  + 1 nM Fe 50 %Io 460 ± 22 (13) 270 ± 12 (13) 170 ± 3.8 (25) 900 ± 30 (14)  + 1 nM Fe 10 %Io 320 ± 30 (9) 140 ± 5.7 (7) 89 ± 2.8 (13) 550 ± 33 (9)  + 1 nM Fe 1 %Io 66 ± 1.8 (2) 35 ± 1.7 (2) 8.5 ± 2.8 (1) 110 ± 33 (2)  + 1 nM DFB 50 %Io 75 ± 2.8 (2) 57 ± 4.5 (3) 34 ± 1.9 (5) 170 ± 0.15 (3)  + 1 nM DFB 10 %Io 78 (2) 48 ± 0 (2) 32 ± 0.69 (5) 160 ± 56 (3)  + 1 nM DFB 1 %Io 52 ± 0.092 (1) 32 ± 0.83 (2) 13 ± 0.48 (2) 97 ± 0.26 (2)  32  Figure 2.3: Average chlorophyll a (ng L-1) concentration for treatment duplicates over the course of an 8-day grow out experiment initiated at station L1 (71°05.94’ N, 139°08.87’ W) in the Beaufort Sea, comparing (a) the no Fe added treatment group, (b) 50 %Io treatments, and (c) 10 %Io treatments. Initial parameters of the seawater collected from 29 m depth include a NO3- concentration, chlorophyll a concentration, and irradiance level of 0.2 µmol L-1 NO3-, 40 ng L-1 chl a, and 6 % of the incident irradiance (Io = 513.73 µE m-2 s-1), respectively. Chlorophyll a concentrations represent biomass collected with 0.22 !m filters on days 2, 4, and 6 and are the summation of biomass collected on 0.22, 1, and 5 !m filters on day 8. Error bars represent the range for duplicate bottles.   33  Light availability also played a key role in controlling phytoplankton biomass (Figure 2.3a). Within the no Fe added treatment group, each light level was associated with approximately double the total chl a as the next lowest light level. For example, the no Fe added 50 %Io treatment had a total chlorophyll a concentration of 860 ± 89 ng L-1, more than twice that of the no Fe added 10 %Io treatment, 390 ± 92 ng L-1, which in turn was more than twice that of the no Fe added 1 %Io chl a concentration, 180 ± 9.5 ng L-1 (Table 2.3). A similar trend was observed in the + 1 nM Fe treatment group (Table 2.3). This trend was not observed in the + 1 nM DFB treatment group; the total chl a concentrations in this group were very low and had a relatively small range of 97 - 170 ng L-1 (Table 2.3).  2.3.4 Plankton growth co-limitation by Fe and light  Chlorophyll a data also reveal the potential for co-limitation of phytoplankton growth by Fe and light. By day 8 of the grow out, relatively higher chl a concentrations were associated with the + 1 nM Fe treatment group and the higher light levels (50 and 10 %Io), whereas the relatively lower chl a concentrations were associated with the + 1 nM DFB treatment group and the lowest light level (1 %Io; Table 2.3). Relative to the control treatment, the largest chl a increase occurred in the + 1 nM Fe 50 %Io and no Fe added 50 %Io treatments in the > 5 µm size fraction, with an increase of 25 and 24 times greater, respectively. It was observed that the + 1 nM Fe 50 %Io and no Fe added 50 %Io treatments had virtually identical chl a concentration in all size fractions and in total (900 ± 30 ng L-1 and 860 ± 89 ng L-1 respectively). These two total chl a values were greater than all other treatments by a margin of 310 ng L-1 chl a (Table 2.3), indicating that the in situ dissolved Fe concentration (0.154 nmol L-1) was not limiting at the highest light level (Figure 2.3b).  34  In contrast, the no Fe added 10 %Io and + 1 nM Fe 10 %Io treatments were virtually identical for all size fractions, except in the 0.22-1 µm size fraction and the total, where chl a concentration was much higher for the + 1 nM Fe treatment (Figure 2.2, Table 2.3). The total chl a concentration of the + 1 nM Fe 10 %Io treatment was 1.4 times higher than the chl a concentration of the no Fe added 10 %Io treatment, with a difference 160 ng L-1 (Figure 2.3c), suggesting that the in situ dissolved Fe concentration was limiting at this light level. Surprisingly, at the 1 %Io light level, chl a concentration of the no Fe added treatment was higher than the + 1 nM Fe treatment for all size fractions and the total (Table 2.3). However, given that total chl a concentrations were comparatively low for the no Fe added 1 %Io and + 1 nM Fe 1 %Io treatments (145 ± 35 ng L-1) relative to the higher light levels (mean total chl a at 10 %Io was 470 ± 80 ng L-1; mean total chl a at 50 %Io was 880 ± 20 ng L-1), this is indicative of insufficient light for growth. Two-way ANOVA analysis, which included all three Fe levels and light intensities, also supports this co-limitation hypothesis (Table 2.4). Tukey testing revealed both independent and interactive effects of Fe and light on total chl a concentrations. Additionally, at the higher light levels (50 and 10 %Io) for cells exceeding 1 µm, the DFB addition was associated with significantly lower chl a than the levels observed for the + 1 nM and no Fe treatments, supporting the Fe deficient condition imposed by this DFB treatment.  Table 2.4: Summary of Fe effects, light (%Io) effects, and Fe " %Io interactive effects, determined by two-way ANOVA statistical analyses, on chlorophyll a, net and maximum nutrient drawdown, and Fe:C ratios. All three Fe levels and three light levels were included in these analyses. Only significant effects are included (p < 0.05). Blank spaces indicate no significant effects; “n/a” indicates test was not possible. Size Fraction Chl a Fe:C Max NO3- Max PO43- Max SiO42- 0.22-1 µm Fe, %Io  n/a n/a n/a 1-5 µm Fe, %Io, Fe # %Io %Io n/a n/a n/a >5 µm Fe, %Io, Fe # %Io %Io n/a n/a n/a Total Fe, %Io, Fe # %Io n/a Fe %Io Fe, Fe # %Io  35 2.3.5 Macronutrient drawdown  Over the course of the grow-out, macronutrients (NO3-, PO43-, and SiO42-) were drawn down within all treatments, relative to initial levels. Though net drawdown of nutrients relative to initial macronutrient concentrations was consistently observed, many nutrients were not drawn down to a maximum on day 8 of the experiment (Table 2.5).  As a result, utilization of macronutrients was calculated two ways: net drawdown is equal to the day 8 nutrient concentration subtracted from the initial nutrient concentration, and maximum drawdown is equal to the lowest nutrient concentration throughout the grow-out subtracted from the initial nutrient concentration. Given that very similar trends were observed in net and maximum nutrient drawdowns, we will focus our discussion here on maximum nutrient drawdown.  Maximum NO3- drawdown ranged between 0.27-3.5 "mol L-1 for all treatments (Table 2.5), and in general, the highest drawdown occurred in the + 1 nM Fe addition treatment. Indeed, both two-way ANOVA analyses revealed an independent Fe effect on maximum NO3- drawdown, highlighting significantly higher NO3- drawdown by the + 1 nM Fe treatment group than the no Fe added treatments (Table 2.4). In addition, the no Fe added and + 1 nM DFB treatment groups exhibited opposing patterns in maximum NO3- drawdown as a function of light levels. In the no Fe added treatment group, as expected, light and NO3- drawdown decreased concomitantly, whereas surprisingly, NO3- drawdown decreased with increasing light level in the + 1 nM DFB treatment group. In comparison, the + 1 nM Fe treatment exhibited no pattern, but NO3- drawdown was uniformly high. Differences in maximum PO43- drawdown between Fe levels were negligible, ranging between 0.05-0.11 "mol L-1 (Table 2.5); however, an independent effect of light was observed, with significantly higher maximum drawdown of PO43- by the 50 %Io light level than the 1 %Io light level (Table 2.4). Finally, maximum SiO42-  36 Table 2.5: Average NO3-, PO43-, and SiO42- drawdown (!mol L-1) for triplicate initials on day 0 and duplicate samples on all sampling days of an 8-day grow out initiated at station L1 in the Beaufort Sea. Initial values reported equal the mean ± standard error. Net drawdown values are equal to the day 8 nutrient concentration subtracted from the initial nutrient concentration. Maximum drawdown values are equal to the lowest nutrient concentration observed throughout the experiment subtracted from the initial nutrient concentration; the day of lowest nutrient concentration is reported in brackets. All net and maximum NO3- drawdown values are calculated with the initial + NO3- value, save for the control treatment which is calculated with the initial value.  drawdown, observed on day 2 for all treatments, ranged 0.60-1.1 "mol L-1 (Table 2.5). Interestingly, at 50 %Io the + 1 nM Fe treatment showed higher maximum SiO42- drawdown than the no Fe added treatment, while at 1 %Io higher drawdown was observed in the no Fe added treatment.  Light trends in maximum SiO42- drawdown were also distinct for the no Fe added and + 1 nM Fe treatments. Whereas drawdown increased as light decreased in the no Fe added treatment group, light and SiO42- drawdown decreased in the + 1 nM Fe group. Both two-way ANOVA analyses revealed an Fe effect and an interaction between Fe and light on maximum SiO42- drawdown (Table 2.4). The interactive effect of light and Fe on maximum SiO42-   Nitrate  Phosphate  Silicate Initial Day 0 0.43 ± 0.039  0.76 ± 0.029  1.8 ± 0.13 Initial + NO3- Day 0 10.43 ± 0.039  0.76 ± 0.029  1.8 ± 0.13 net  0.22  0.050  0.10 Control 10 %Io no NO3- maximum 0.27 (6)  0.050 (6-8)  0.60 (2) net  0.81  0.11  0.40 no Fe added 50 %Io maximum 3.1 (6)  0.11 (8)  0.80 (2) net  1.1  0.090  0.30 no Fe added 10 %Io maximum 1.1 (8)  0.090 (8)  0.85 (2) net  0.53  0.070  0.40 no Fe added 1 %Io maximum 0.83 (4-6)  0.090 (6)  0.93 (2) net  0.43  0.10  0.30 + 1 nM Fe 50 %Io maximum 3.0 (6)  0.10 (8)  1.1 (2) net  2.9  0.10  0.30 + 1 nM Fe 10 %Io maximum 2.9 (8)  0.10 (8)  0.80 (2) net  2.3  0.080  0.10 + 1 nM Fe 1 %Io maximum 3.5 (4)  0.080 (8)  0.60 (2) net  1.3  0.070  0.30 + 1 nM DFB 50 %Io maximum 1.3 (8)  0.10 (6)  0.80 (2) net  1.9  0.080  0.20 + 1 nM DFB 10 %Io maximum 1.9 (8)  0.080 (8)  0.80 (2) net  2.6  0.070  0.20 + 1 nM DFB 1 %Io maximum 2.6 (8)  0.080 (6)  0.80 (2)  37 drawdown highlights that low Fe enhanced Si drawdown in the 1 %Io treatments, though these differences are close to that of the analytical error associated with the initial SiO42- concentration (0.93, and 0.6 µmol L-1 SiO42- for the no Fe added and + 1 nM Fe treatments, respectively).  2.3.6 Plankton community composition  The plankton community composition in all treatments on day 8 was investigated using accessory pigments, which were determined with an HPLC and analyzed with CHEMTAX (Tables 2.2 and 2.6). Many samples contained too little biomass for the pigments to be quantified. Only four treatments – no Fe added 10 %Io, no Fe added 1 %Io, + 1 nM Fe 50 %Io, and + 1 nM Fe 10 %Io – yielded results (Table 2.6). Overall, 63-76 % of the phytoplankton composition for all four treatments was dominated by a combination of chlorophytes and prasinophytes, though group % abundances varied between 12-55 %. The treatments at 10 %Io had very similar high mean % abundances for chlorophytes and prasinophytes. The + 1 nM Fe treatments were characterized by less diatoms and more pelagophytes and dinoflagellates than the no Fe added treatments; this treatment group also had a very small amount of cyanobacteria. Light level differences between the two + 1 nM Fe treatments demonstrated that at 10 %Io, prasinophytes decreased and chlorophytes increased in comparison to the 50 %Io light level. Surprisingly, diatoms and prasinophytes were more abundant at 1 %Io in comparison to 10 %Io in the no Fe added treatment group. This no Fe added 1 %Io treatment had at least double the amount of diatoms as the other three treatments that were analyzed with CHEMTAX.  In only one treatment, + 1 nM Fe 10 %Io, the plankton community composition was analyzed by both HPLC and light microscopy. Fortunately, all three day 0 initials were preserved amongst the few surviving light microscopy samples. The initial phytoplankton community composition was dominated by small flagellates and benthic diatoms (c.f. pinnularia), with a few  38 Table 2.6: Mean abundance (%) of eight algal groups on day 8 of an 8-day grow out initiated at station L1 (71°05.94’ N, 139°08.87’ W) in the Beaufort Sea. Values represent the mean ± the range for duplicate treatment bottles; those without an error value were detected in one bottle per treatment. Mean abundance of Prasinophytes Types 1 and 2 were combined. Mean abundances below detection are noted with “bd”. Data was unavailable for those noted with “-”.       Treatment Diatoms Chlorophytes Prymnesiophytes Prasinophytes  Cryptophytes Pelagophytes Dinoflagellates Cyanobacteria Control 10 %Io, no NO3- bd bd bd bd bd bd bd bd no Fe added 50 %Io - - - - - - - - no Fe added 10 %Io 15 ± 2.5 43 ± 2.8 3.7 ± 0.4 33 ± 3.8 2.1 4.6  ± 0 bd bd no Fe added 1 %Io 31 ± 4.1 12 ± 1.8 1.2 ± 1.2 55 ± 1.1 0.0 0.0 ± 0 bd bd + 1 nM Fe 50 %Io 12 ± 5.0 20 ± 12 2.7 43 ± 6.6 4.6 ± 4.0 16 ± 13 5.4 0.26 ± 0.22 + 1 nM Fe 10 %Io 8.0 ± 7.0 37 ± 7.7 2.6 ± 0.1 36 ± 12 0.0 ± 0 13 ± 8.3 6.7 0.57 + 1 nM Fe 1 %Io bd bd bd bd bd bd bd bd + 1 nM DFB 50 %Io bd bd bd bd bd bd bd bd + 1 nM DFB 10 %Io bd bd bd bd bd bd bd bd + 1 nM DFB 1 %Io bd bd bd bd bd bd bd bd  39 Thalassiosira sp. and diatom frustules. Inferring from HPLC analyses, the unidentified flagellates were likely dominated by prasinophytes and chlorophytes. On day 8 of the incubation, duplicates of the control treatment were dominated by small flagellates and some large Thalassiosira and Chaetoceros (c.f. socialis and laciniosus); a few Dictyocha speculum were also seen. The Fe-enriched treatment, + 1 nM Fe 10 %Io, had a much higher density of phytoplankton species on Day 8 in comparison to the control treatment. Agreeing well with the HPLC results (Table 2.6), this sample was mainly composed of small flagellates (prasinophytes and chlorophytes as determined by HPLC analysis), followed by Chaetoceros (c.f. socialis and laciniosus) and then Pseudo-nitzschia delicatissima; some Cylindrotheca closterium, Thalassiosira sp., and a benthic diatom (c.f. pinnularia) were also observed.  Using flow cytometry and Cytowin analyses, we also determined nano- and picoplankton (< 2 µm) cell densities in our treatments. Both cyanobacteria and picoeukaryotes were identified; however, we were unable to accurately determine cyanobacteria abundance as their signal was consistently cut off by the red fluorescence axis in the Cytowin scatter plots. Despite this, a portion of the cyanobacteria community was captured, and yielded day 8 cell densities between 4,000-38,000 cells mL-1 for all treatments. Cell abundances in the 104 cells mL-1 range were associated with the no Fe added and + 1 nM Fe treatments at 50 or 10 %Io, whereas cell abundances in the 103 cells mL-1 range were associated with the + 1 nM DFB and 1 %Io treatments. Initial cyanobacteria abundance ranged from 2,000-54,000 cells mL-1, but averaged 13,000 cells mL-1. These ranges encompass both low cyanobacteria abundances (those in the 103 cells mL-1 range) observed in the Arctic (Not et al. 2005; Lovejoy et al. 2007) and those of subarctic latitudes (104 cells mL-1 range; Suzuki et al. 2002; Not et al. 2005; Tremblay et al. 2009). In conjunction with results obtained by HPLC pigment analysis, it appears that the  40 incomplete flow cytometry cyanobacteria data may have lead to overestimation of cell abundance, though both analyses suggest low abundance of cyanobacteria.  On day 0, flow cytometry determinations yielded an initial picoeukaryote cell density of 2,134 ± 59 cells mL-1 (n = 9, data not shown). Pooling flow cytometry-determined picoeukaryote cell densities and chl a concentrations for treatment duplicates on days 4, 6, and 8 reveals a significant correlation (Figure 2.4a; p < 0.0001, F = 16.24, 52 df) and by examining each day separately, the effect of light on biomass becomes evident. On day 8 (Figure 2.4b), duplicates of two treatments, no Fe added 50 %Io and + 1 nM Fe 50 %Io, were characterized by the highest picoeukaryote cell densities (between 6,500-11,000 cells mL-1) and chl a concentrations (p < 0.0001, F = 13.37, 16 df), yielding a 5 times increase relative to initial cell density. Duplicates of the no Fe added 10 %Io and + 1 nM Fe 10 %Io treatments were distinguished by intermediate cell densities (4,400-5,900 cells mL-1) and chl a concentrations. Within these two light level groupings, Fe level did not play a factor in the relationship of chl a to picoeukaryote population. All other treatments, characterized by either 1 %Io or + 1 nM DFB, exhibited low picoeukaryote cell densities (< 1,800 cells mL-1) and correspondingly low chl a. Day 6 shows a similar but less pronounced trend for the different nutritional states (Figure 2.4c; p < 0.0001, F =7.38, 16 df). Picoeukaryote flow cytometry from day 4, however, reveals that light intensity had not yet affected differing growth responses in different treatments as no significant correlation was found between picoeukaryote cell density and chl a biomass (p = 0.0675, F = 1.96, 16 df; data not shown), thus highlighting the amount of time needed to observe effects of limiting resources on plankton growth.  41  Figure 2.4: Relationships between cell density of picoeukaryotes (cells mL-1) and total chlorophyll a concentration (ng L-1) for (a) all of days 4, 6, and 8, (b) day 8, and (c) day 6 of an 8-day grow out experiment initiated at station L1 (71°05.94’ N, 139°08.87’ W) in the Beaufort Sea. This figure includes duplicate sample data for all light levels within the no Fe added, + 1 nM Fe, and + 1 nM DFB treatment groups. The no Fe added and + 1 nM Fe treatments at 50 %Io are denoted by !, the no Fe added and + 1 nM Fe treatments at 10 %Io are denoted by ", and all other treatments are denoted by #. Linear regressions: a. (y = 8482.5x + 1162.2, R2 = 0.84, t0.0001(2)52 = 16.24, p < 0.0001), b. (y = 9802.8x -137.36, R2 = 0.92, t0.0001(2)16 = 13.37, p < 0.0001), and c. (y = 17,577x + 376.04, R2 = 0.77, t0.0001(2)16 = 7.38, p < 0.0001). Note the smaller axes scales in 4c.   42 2.3.7 Short-term Fe uptake rates  Volumetric !FeDFB ranged between 0.068-3.1 pmol Fe L-1 d-1 across size fractions for all treatments (Table 2.7). The 0.22-1 µm size fraction exhibited the fastest volumetric !FeDFB (0.14-3.1 pmol Fe L-1 d-1), followed by the 1-5 µm size fraction (0.068-0.77 pmol Fe L-1 d-1), and  Table 2.7: Short-term (24 h) volumetric ($10-12 mol Fe L-1 d-1) and carbon-specific ($10-6 mol Fe (mol C)-1 d-1) Fe uptake rates of size-fractionated phytoplankton on day 6 of an 8-day grow out initiated at station L1 (71°05.94’ N, 139°08.87’ W) in the Beaufort Sea. Iron was precomplexed to ligands DFB or EDTA in the ratios 100 nmol L-1 Fe : 105 nmol L-1 DFB with 5 % of total Fe as 55Fe, or 100 nmol L-1 Fe : 1000 nmol L-1 EDTA with 1 % of total Fe as 55Fe. Values are equal to the mean ± the range for duplicate bottles.      !FeDFB     !FeEDTA Treatment   Volumetric       Carbon-specific     Volumetric      Carbon-specific 0.22-1 µm Control 10 %Io, no NO3- 0.17 ± 0.017 0.95 ± 0.39 2.1 ± 0.31 12 ± 5.3 no Fe added 50 %Io 0.80 ± 0.32 0.95 ± 0.38 7.5 ± 1.2 9.0 ± 1.5 no Fe added 10 %Io 0.26 ± 0.034 0.56 ± 0.058 3.9 ± 0.52 8.5 ± 0.88 no Fe added 1 %Io 0.27 ± 0.014 0.86 ± 0.059 3.6 ± 1.1 12 ± 3.3 + 1 nM Fe 50 %Io 0.69 ± 0.18 0.86 ± 0.24 5.5 ± 0.89 6.8 ± 1.2 + 1 nM Fe 10 %Io 0.30 ± 0.11 0.62 ± 0.25 2.7 ± 0.38 5.4 ± 0.44 + 1 nM Fe 1 %Io 1.4 ± 0.76 5.7 ± 3.1 3.1 ± 0.80 12 ± 3.2 + 1 nM DFB 50 %Io 3.1 ± 0.41 11 ± 0.090 5.8 ± 2.3 19 ± 5.8 + 1 nM DFB 10 %Io 0.14 ± 0.0017 0.68 ± 0.024 5.6 ± 1.1 28 ± 7.0 + 1 nM DFB 1 %Io 1.4 ± 0.82 7.2 ± 4.5 3.9 ± 0.0027 20 ± 0.98 1-5 µm Control 10 %Io, no NO3- 0.068 ± 0.0032 0.37 ± 0.14 2.2 ± 0.49 13 ± 6.5 no Fe added 50 %Io 0.39 ± 0.095 0.47 ± 0.12 16 ± 1.6 19 ± 2.0 no Fe added 10 %Io 0.25 ± 0.018 0.55 ± 0.023 6.9 ± 0.080 15 ± 0.26 no Fe added 1 %Io 0.20 ± 0.036 0.63 ± 0.11 3.5 ± 0.66 11 ± 1.9 + 1 nM Fe 50 %Io 0.43 ± 0.11 0.53 ± 0.14 11 ± 0.079 14 ± 0.27 + 1 nM Fe 10 %Io 0.27 ± 0.040 0.55 ± 0.11 5.0 ± 0.055 9.9 ± 0.69 + 1 nM Fe 1 %Io 0.43 ± 0.23 1.7 ± 0.91 1.3 ± 0.078 5.2 ± 0.33 + 1 nM DFB 50 %Io 0.77 ± 0.18 2.6 ± 0.30 10 ± 0.40 37 ± 3.2 + 1 nM DFB 10 %Io 0.28 ± 0.15 1.3 ± 0.66 4.5 ± 0.11 22 ± 0.50 + 1 nM DFB 1 %Io 0.45 ± 0.20 2.4 ± 1.1 1.9 ± 0.11 9.7 ± 1.0 > 5 µm Control 10 %Io, no NO3- 0.12 ± 0.012 0.57 ± 0.15 2.7 ± 0.59 16 ± 7.9 no Fe added 50 %Io 0.17 ± 0.036 0.21 ± 0.042 31 ± 4.3 37 ± 5.0 no Fe added 10 %Io 0.088 ± 0.032 0.19 ± 0.064 18 ± 0.43 40 ± 0.21 no Fe added 1 %Io 0.18 ± 0.019 0.58 ± 0.051 4.9 ± 0.089 16 ± 0.53 + 1 nM Fe 50 %Io 0.20 ± 0.019 0.25 ± 0.026 20 ± 0.22 24 ± 0.57 + 1 nM Fe 10 %Io 0.14 ± 0.041 0.28 ± 0.098 6.9 ± 0.83 13 ± 0.86 + 1 nM Fe 1 %Io 0.19 ± 0.011 0.79 ± 0.043 1.7 ± 0.058 6.8 ± 0.25 + 1 nM DFB 50 %Io 0.60 ± 0.16 2.0 ± 0.32 22 ± 1.0 80 ± 6.3 + 1 nM DFB 10 %Io 0.093 ± 0.013 0.46 ± 0.041 11 ± 0.039 55 ± 2.8 + 1 nM DFB 1 %Io 0.25 ± 0.11 1.3 ± 0.60 2.0 ± 0.35 10 ± 2.2  43 the > 5 µm size fraction (0.088-0.60 pmol Fe L-1 d-1), demonstrating that smaller cells account for the majority of the Fe uptake in situ. Overall, volumetric !FeEDTA for all experimental treatments ranged between 1.3-31 pmol Fe L-1 d-1, approximately one order of magnitude faster than volumetric !FeDFB. The fastest volumetric !FeEDTA were observed in the > 5 µm size fraction, followed by the 1-5 µm and then the 0.22-1 µm size fractions. This result reflects that larger cells had a higher density of plasmalemma-bound Fe transporters, indicating Fe stress.  Within all size fractions, treatments characterized by either low light (1 %Io) and/or Fe deficiency (+ 1 nM DFB) often exhibited the fastest carbon-normalized !FeDFB (Table 2.7), which ranged from 0.19-11 µmol Fe (mol C)-1 d-1. Within each size fraction, the + 1 nM DFB treatment had highest carbon-normalized !FeDFB for each light level, suggesting that Fe deficient cells achieve the fastest in situ Fe uptake rates in the presence of siderophore-Fe complexes. Within the no Fe added and + 1 nM Fe treatment groups, carbon-normalized !FeDFB was often faster for treatments at 1 %Io (Table 2.7). At this lowest light level, rates were faster for the + 1 nM Fe treatment relative to the no Fe added treatment.  Carbon-normalized !FeEDTA were fastest in the > 5 µm size fraction, followed by the 1- 5 µm and then the 0.22-1 µm size fractions. Treatments with + 1 nM DFB consistently ranked amongst the fastest carbon-normalized !FeEDTA at each light level, indicating that these cells were characterized by the highest densities of plasmalemma-bound Fe transporters, most likely an adaptation to Fe limitation. Comparing Fe treatment groups reveals that carbon-normalized !FeEDTA was generally fastest at the lowest light level in the 0.22-1 µm size fraction and generally fastest at the highest light level in cells exceeding 1 µm (Table 2.7), indicating that larger, Fe-stressed cells were better able to acquire Fe under higher light levels. In general, !FeEDTA in the no Fe added treatments were faster than those in the + 1 nM Fe treatments.  44 2.3.8 Fe:C ratios For all treatments and size fractions, Fe:C ratios ranged from 1.6-310 µmol Fe (mol C)-1 (Table 2.8). The highest Fe:C ratios were those of the 0.22-1 µm size fraction, followed by the > 5 µm and then the 1-5 µm size fractions. The Fe:C ratios observed in this investigation follow the same pattern as those reported in a Fe and light co-limitation field study conducted in the subarctic Pacific Ocean (Maldonado et al., 1999); however, some of the values reported here are higher (up to 45 times). Given that these differences are especially pronounced in the 0.22-1 µm size fraction, it is believed that heterotrophic bacteria skewed Fe:C ratios. Within a time span of only 24 h, heterotrophic bacteria readily take up Fe but not inorganic 14C (or the recently-labeled organic 14C), thereby increasing the ratio of internalized Fe to C.  Table 2.8: Size-fractionated Fe:C ratios (µmol Fe (mol C)-1) of phytoplankton on day 6 of an 8-day grow out initiated at station L1 (71°05.94’ N, 139°08.87’ W) in the Beaufort Sea. Iron was precomplexed to DFB in the ratio 100 nmol L-1 Fe : 105 nmol L-1 DFB with 5 % of total Fe as 55Fe. Carbon was added to each bottle to a final concentration of 20 µCi L-1 14C. Values are equal to the mean ± the range for duplicate bottles.  Treatment 0.22-1 µm 1-5 µm >5 µm Control 10 %Io, no NO3- 14 ± 5.4 3.7 ± 0.78 14 ± 1.1 no Fe added 50 %Io 11 ± 8.4 2.7 ± 0.64 2.5 ± 0.40 no Fe added 10 %Io 4.8 ± 0.43 3.2 ± 0.83 2.6 ± 1.5 no Fe added 1 %Io 33 ± 0.55 32 ± 6.3 43 ± 4.9  + 1 nM Fe 50 %Io 8.7 ± 0.54 2.9 ± 0.17 3.0 ± 0.34  + 1 nM Fe 10 %Io 5.4 ± 2.1 1.6 ± 0.19 2.4 ± 0.89  + 1 nM Fe 1 %Io 270 ± 230 79 ± 59 81 ± 8.8  + 1 nM DFB 50 %Io 310 ± 52 52 ± 22 27 ± 8.0  + 1 nM DFB 10 %Io 15 ± 3.8 18 ± 13 5.5 ± 1.6  + 1 nM DFB 1 %Io 200 ± 150 100 ± 47 140 ± 85    45  Across all size fractions, higher Fe:C ratios were associated with low light (1 %Io) treatments, indicating high Fe requirements by light limited cells. Within the 1-5 and > 5 µm size fractions, Fe:C ratios at the 1 %Io light level were at least one order of magnitude greater than the other two light levels in all the treatment groupings. Within the 0.22-1 µm size fraction, the Fe:C ratios at the 1 %Io light level were greater than those at the higher light levels for both the + 1 nM Fe treatment group (by 31-50 times), and the no Fe added treatment group (by 3-7 times). Treatments with added DFB did not follow the same Fe:C ratio trends observed in the no Fe added and + 1 nM Fe treatment groups (Table 2.8). However, Fe:C ratios were generally highest in the + 1 nM DFB treatment group regardless of light level. Additionally, Fe:C in the + 1 nM DFB 50 %Io treatment was surprisingly high, regardless of size fraction. Overall, two-way ANOVA analyses including all three Fe levels noted that the 50 and 10 %Io light levels exhibited significantly smaller Fe:C than the 1 %Io light level for cells exceeding 1 µm (Table 2.4). Independent of the + 1 nM DFB treatment group, two-way ANOVA comparing the no Fe added and + 1 nM Fe treatment groups revealed an Fe effect (p < 0.05, not shown) with significantly higher ratios by the + 1 nM Fe treatment group than the no Fe added treatment group in cells exceeding 5 µm.  2.4 Discussion  2.4.1 The plankton community at L1 in the Beaufort Sea during the summer During late summer at station L1 in the Beaufort Sea, phytoplankton biomass was extremely low (50-190 ng L-1; Figure 2.2) and the phytoplankton community was dominated by picoplankton (< 2 µm). More specifically, we encountered a community rich in small flagellates, mainly chlorophytes and prasinophytes, with ~ 70 % of the total chl a in cells < 5 µm (Table 2.3)  46 and a high abundance of picoeukaryotes. Similar Arctic phytoplankton communities have been previously reported (Sherr et al., 2003; Hill et al., 2005; Not et al., 2005; Lovejoy et al., 2007; Tremblay et al., 2009). Additionally, the presence of diatom frustules in our triplicate initial light microscopy samples is evidence of a previous spring bloom, which likely occurred as ice melt facilitated light penetration and uptake of nitrogen in May-June (Sherr et al., 2003; Hill et al., 2005; Lovejoy et al., 2007; Tremblay et al., 2009). Spring diatom blooms have been observed along Arctic continental shelves and lead to a rapid increase in chlorophyll concentration, terminating due to NO3- exhaustion (Hill et al., 2005). Indeed, our + 10 µmol L-1 NO3- addition induced increased chl a production by day 8 in all treatments relative to the control, indicating in situ NO3- limitation of phytoplankton productivity in the Beaufort Sea during late summer. Yet, this NO3- enrichment was associated with an increase in the abundance of cells < 5 µm (see control vs. no Fe added treatments at 10 %Io; Table 2.3) and not an increase of diatoms. This surprising result was likely due to the low in situ Si concentration (1.8 "mol L-1) in our bottles. Overall, our chl a (Table 2.3, Figure 2.3) and nutrient data (Table 2.5), supported by past field studies (Hill et al., 2005; Tremblay et al., 2006; 2009), suggest that surface water NO3- concentrations in the Beaufort Gyre in late summer were low enough to limit primary productivity.  2.4.2 Primary nitrogen source Despite its low concentration in the western Arctic (2-7 "mol L-1 prior to the spring bloom; Sherr et al., 2003; Tremblay et al., 2006), NO3- has been considered the primary DIN source to phytoplankton in the Beaufort Sea in spring (Sherr et al., 2003; Hill et al., 2005; Yamamoto-Kawai et al., 2006; Tremblay et al., 2009). Though we were unable to measure NH4+  47 concentrations aboard the Amundsen, low in situ NO3- (0.13 "mol L-1) was observed. The Arctic Ocean receives seawater from the subarctic Pacific Ocean, which is depleted in NO3- due to phytoplankton drawdown as it flows northward through the Bering Strait. Low NO3- in these incoming waters is further depleted by denitrification along the broad continental shelves of the Bering and Chukchi seas (Yamamoto-Kawai et al., 2006). Thus it is plausible that primary production in the Beaufort Sea could be characterized by a fluctuating source of DIN, with NO3- dominating in spring and NH4+ dominating in summer following the depletion of NO3- stocks. Strong seasonality is a dominant feature of polar environments, including nitrogen availability (Koike et al., 1986). Antarctic field studies along coastal regions of the Scotia Sea have described an annual flux in primary DIN source, beginning with a springtime bloom of new production, followed by the depletion of NO3- throughout the spring and summer (Olson, 1980; Glibert et al., 1982; Rönner et al., 1983; Koike et al., 1986). Eventually, diatoms die out due to nutrient depletion and dominance of new production is overtaken by regenerated production as the growing season continues into late summer. Under these low nutrient conditions, zooplankton grazing and bacterial degradation of organic matter increase NH4+ concentration in surface waters and picoplankton proliferate (Koike et al., 1986). This trend is also observed in the Arctic; one particular study noted that NO3- uptake rates and f-ratios were very high throughout a spring bloom in the North Water polyna (Tremblay et al., 2006). After mid-May, NO3- uptake rates declined and the primary source of DIN became NH4+ for the remainder of the growing season. In the Beaufort Sea, the low NO3- concentrations and dominance of cells < 5 µm observed in late summer were typical of a community based on NH4+ use.   48 2.4.3 Co-limitation by light and Fe  The chlorophyll a and nutrient data of the grow out experiment indicate that upon an addition of NO3-, in situ Fe levels and light were also co-limiting growth of the indigenous plankton community at L1 during late summer/early fall. Though the + 1 nM Fe and no Fe added treatments acclimated to 50 %Io had almost identical chl a throughout the grow-out experiment (Figure 2.2b), the treatment grown under 10 %Io with a + 1 nM Fe addition resulted in higher final chl a concentrations (Figure 2.2c) and NO3- drawdown (Table 2.5) than that of its corresponding no Fe added treatment. It is expected that when placed under low light, phytoplankton increase the content of light-harvesting photosynthetic units (PSUs), which have an absolute requirement for Fe (Falkowski and LaRoche, 1991). As a result, the cellular requirements for Fe increases, and cells in Fe-rich media grow more favorably as they can synthesize more PSUs, and utilize light more efficiently (Greene et al., 1991; Strzepek and Harrison, 2004). It appears that the L1 phytoplankton at 10 %Io could not modify their physiology to optimize growth with the low dissolved Fe concentration found at 30 m depth (0.154 nmol L-1; J. Cullen, personal communication), whereas plankton with a + 1 nM Fe addition were able to make physiological changes that facilitated faster growth. Indeed, the ability of the cells in the Fe enriched treatment to acquire excess Fe for photosynthesizing optimally at low light is supported by the significantly higher Fe:C ratio in the + 1 nM Fe treatment relative to those in the no Fe added treatment at 1 %Io. Our results thus indicate that the in situ Fe concentration limited growth at light levels # 10 %Io after NO3- was added.  In support of the high Fe:C ratios in the low light treatments, the carbon-normalized !FeDFB in all size fractions were usually faster (up to 9 times) for phytoplankton in the 1 %Io than the 50 and 10 %Io for both the + 1 nM Fe and no Fe added treatment groups. Faster Fe  49 uptake rates are likely a result of a) an increase in Fe requirements to synthesize additional PSUs, b) decreasing cell size, and/or c) the upregulation of Fe transport mechanisms (Sunda and Huntsman, 1997). Diatoms can reduce cell size to adapt to low Fe concentrations (de Baar et al., 2005; Sarthou et al., 2005). In doing so, these organisms increase their surface area:volume ratios which facilitate more efficient Fe uptake as it increases exchanges at the cell surface and decreases diffusion limitation (Morel et al., 1991). Fast !FeDFB are indicative of Fe limitation, as FeDFB is only accessed by a high affinity Fe transport system that is upregulated under Fe deficiency. Thus, the fastest rates of !FeDFB we observed for phytoplankton in the 1 %Io light level indicate that these phytoplankton were experiencing severe Fe stress in late summer, and were upregulating their high affinity Fe transport systems to enhance Fe uptake.  At the saturating concentrations of FeEDTA that we used in complementary Fe uptake experiments, inorganic Fe uptake is limited by the number of Fe transporters at the cell surface. Previous studies have shown that phytoplankton are able to increase the number of Fe transporters up to 20 fold when experiencing Fe stress (Maldonado and Price, 2001). In our study, the fastest !FeEDTA observed for the + 1 nM Fe and no Fe added treatment groups were at 1 %Io in the 0.22-1 µm size fraction, which was associated with the majority of biomass by day 8 of the grow out. These faster rates (up to 2.2 times) further support the !FeDFB-based findings that at this low light level the phytoplankton were Fe stressed, and were also increasing the number of Fe transporters to enhance inorganic Fe uptake. Interestingly, the fastest !Fe (for both !FeDFB and !FeEDTA) were observed in the > 5 µm size fraction, which should be the most Fe limited phytoplankton as a result of their low surface area:volume ratios. Finally, the no Fe added 10 %Io treatment had significantly higher and lower maximum drawdown of SiO42- and NO3-, respectively, than the + 1 nM Fe treatment. This result may also be an indicator of Fe  50 stress in the no Fe added treatment, given that diatoms under Fe limitation have been shown to increase their uptake of Si relative to N and P (Takeda, 1998).  2.4.4 Light limitation  The incident PAR measured by the CTD cast corresponding to our grow out (513.73 µE m-2 s-1) was the highest of all the Io captured at station L1. Indeed, the clear and mild weather conditions at the time of the cast did not last our stay at station L1; overcast and misty weather was also observed. Incident irradiance was measured 6 times throughout the 2 days spent at L1 between daylight hours of 7 AM and 8 PM MDT, ranging 25.27 – 513.73 µE m-2 s-1 and the mean was 239.03 µE m-2 s-1. Assuming these measurements are representative of the range of Io incident on our incubators throughout the 8-day experiment, the average light experienced by the phytoplankton in the 50, 10, and 1 %Io bottles was 119.52 µE m-2 s-1, 23.90 µE m-2 s-1, and 2.39 µE m-2 s-1, respectively.  Chlorophyll a data elucidated the effect of irradiance on Arctic primary production (Table 2.3; Figure 2.3a). Though our simulated irradiance levels did not correspond to those of the MLD (11.6 m; 22 %Io) or pumping depth (29 m; 6 %Io), our lowest light level is similar to the light level at 50 m (2 %Io), the depth at which CTD bottle samples yielded the highest chl a concentration (Figure 2.2c). Together, these parameters can be utilized to predict the possibility of growth light limitation for the plankton community at station L1. CTD data were used to calculate integrated irradiance ($D) received by phytoplankton cycling throughout the mixed layer. First, surface irradiance and six light levels at various depths within the euphotic zone (72 m) were used to calculate the average extinction coefficient for the water column (kd = 0.096 ± 0.0080). Utilizing the range of light levels observed at station L1, estimated MLD, and Io, a  51 range of $D received by the plankton community circulating in the mixed layer at L1 were calculated. Based on the observed Io and a kd of 0.096, the range of $D received by phytoplankton cycling throughout the mixed layer (11.6 m) at L1 was 15.10 – 306.97 µE m-2 s-1. This range suggests that phytoplankton in the mixed layer, as well as those at the depth of maximum fluorescence (~ 72 m), at station L1 would experience low light levels comparable to those in our 10 %Io (23 µE m-2 s-1) and 1 %Io treatments (2.4 µE m-2 s-1). As such, low irradiances in conjunction with the low Fe concentration observed at station L1 (0.154 nmol L-1) likely co-limit growth of indigenous plankton during late summer/early fall.  A relationship between temperature and optimal irradiance has been established for both marine and freshwater phytoplankton (Henley, 1992; Rae and Vincent, 1998; Lovejoy et al., 2007). At 0 °C, Micromonas pusilla, an Arctic prasinophyte that blooms in spring and proliferates throughout the growing season (Sherr et al., 2003; Not et al., 2005; Lovejoy et al., 2007), exhibited maximum growth rates around 10 µE m-2 s-1 (Lovejoy et al., 2007). Micromonas pusilla was identified as one of the most abundant cells in the Arctic (Lovejoy et al., 2007), and is suspected of phagotrophy (Gonzalez et al., 1993; Caron et al., 1995), characterizing this organism as well-adapted to the Arctic environment. In late August/early September, 2009 in the Beaufort Sea, high abundances of chlorophytes and prasinophytes were observed (Table 2.6); similarly, picoeukaryote abundance was high in initial samples and increased up to 5 fold at the highest light and Fe levels in the grow-out experiment. We expect that small phytoplankton such as Micromonas pusilla, able to readily use NH4+ and grow optimally at low irradiances and Arctic temperatures, will dominate at station L1 despite the range of Io observed in late summer/early fall.  52 2.4.5 Decoupled seasonalities of Fe, light, and NO3-  The results presented above were obtained from a ship deck incubation experiment initiated in late summer (August/September) in the Beaufort Sea. As such, one of the most predominant outcomes of this investigation is evidence of NO3- limitation at station L1. The cycle of NO3- bioavailability is seasonal in the Arctic Ocean (Sherr et al., 2003; Hill et al., 2005), thus this outcome would have been much less pronounced or nonexistent were this research initiated in spring. Similarly, light levels fluctuate in the Beaufort Sea on an annual cycle and limit primary productivity in the early spring and again in fall due to sea ice cover and low light intensities (Sherr et al., 2003; Hill et al., 2005). Previous field study in the Beaufort Sea observed a prasinophyte bloom with the onset of light in spring (Lovejoy et al., 2007). Thus, had we made our voyage to the Arctic in late spring/summer, the effects of light limitation would not have been observed. Much like the seasonalities of NO3- and light availability in the Beaufort Sea, Fe undergoes an annual cycle. This cycle is primarily regulated by sea ice cover, which acts as a sink for the micronutrient by entraining Fe-containing suspended sediments during the formation of pack ice and accumulating dust deposited from the atmosphere during the winter (Measures, 1999; Isaksson et al., 2003; Krachler et al., 2005). Throughout the spring and summer, sea ice becomes a source as ice melts, supplying Fe to surface waters. However, this supply of Fe to arctic waters does not guarantee sufficient concentrations for plankton growth, as demonstrated by the low 0.154 nmol L-1 Fe concentration at 30 m depth associated with 75 % ice cover at station L1. Furthermore, sea ice cover is patchy and variable from year to year due to wind, temperature, and climate patterns (Stroeve et al., 2007), affecting the location and the rate of supply of Fe in the Beaufort Sea. In general, light penetrates surface waters due to receding sea  53 ice in spring, and optimal NO3-, light, and Fe levels support growth of plankton in tandem. Yet as summer approaches, NO3- is depleted to limiting levels and decouples from the ongoing supply of light and Fe as ice melt continues into early fall. Eventually, decreasing light levels associated with late summer/early fall exacerbate the NO3--limited state of the in situ plankton community. Due to the integral role of Fe in photosynthesis (Raven, 1988), Fe demand increases as light availability decreases. Ultimately, uptake of Fe becomes decoupled from its rate of supply, as high demand for Fe in the presence of decreasing irradiance leads to severe Fe-deficiency that cannot be supported by in situ Fe concentrations. This deficiency inhibits late summer/early fall primary productivity, ending the growing season and preventing the utilization of any nutrients that may be mixed into the euphotic zone in fall. Although few and nutritionally insignificant autumn mixing events have been observed in the Beaufort Sea (Carmack et al., 2002; Yang et al., 2004; Forest et al., 2007; Tremblay et al., 2008), it has been determined that the fall storm season (August-October) in the southern Beaufort Sea region is associated with more storms in warmer years (Hudak and Young, 2002). This trend suggests that fall storm and mixing events may highlight the Fe-deficient state of Beaufort Sea plankton in future years. Overall, the decoupling in the seasonalities of NO3-, light, and Fe availabilities affects the chemical and physical environment in the Beaufort Sea, and likely influences dynamic plankton communities and their physiological mechanisms to cope with these fluctuations.  2.4.6 Oceanographic implications  The effect of Fe on cell growth is linked to its role in both nitrogen and carbon assimilation pathways (Rueter and Ades, 1987). In an environment with seasonal NO3- and light such as the Arctic, this micronutrient is essential as a supporter of these processes. With the onset  54 of global warming, the seasonal Fe supply by ice melt will shorten and the potential for Fe deficiency will increase. Additionally, Fe availability is expected to decrease due ocean acidification, which will reduce the effectiveness of biologically-mediated redox reactions integral to Fe uptake from organic complexes (Shi et al., 2010). Strengthened stratification, due to enhanced freshwater input, will reduce the influx of macronutrients into the mixed layer (Toole et al., 2010) and shoaling oxygen minimum zones (OMZs), defined as layers of nitrogen loss, will reduce the amount of NO3- in surface waters through denitrification (Paulmier and Ruiz-Pino, 2009). Oxygen minimum zones are also characterized by a greater dissolved Fe concentration than surrounding waters (Landing and Bruland, 1987), though it is undetermined how the combination of low oxygen and nitrate with increased dissolved Fe will affect primary productivity. The interaction will undoubtedly have a particularly important impact on Chukchi and Beaufort Sea headwaters that flow through the OMZ regions of the West Bering Sea and Gulf of Alaska. The combination of ocean acidification and expanding oxygen minimum zones will likely lead to a shift in community composition, preventing diatom-dominated spring blooms and supporting a less dynamic year-round plankton community (Armbrust, 2009). The effect of this community shift on the efficiency of the biological carbon pump is negative (Michaels and Silver, 1988; Smetacek, 1999; Takahashi et al., 2002; Richardson and Jackson, 2007; Armbrust, 2009), as diatoms are considered to be the star players in carbon export from the surface to the deep ocean due to their large size, heavy silicate wall, and ability to form aggregates with other cells (Michaels and Silver, 1988; Lochte et al., 1993; Buesseler et al., 1998; Bopp et al., 2005; Sarthou et al., 2005). However, the contribution of picoplankton is not insignificant, as these organisms also have the ability to form aggregates thereby increasing their sinking rates and efficiency in trophic transfer through the foodweb (Richardson and Jackson,  55 2007). Domination of Arctic Ocean plankton communities by picoplankton may still support the biological carbon pump when diatoms are not abundant. We suggest that future study of the Arctic Ocean explores the seasonality of Fe availability associated with changes in sea ice. In addition, future investigations should focus on the sinking efficiency of the regenerated plankton community observed in late summer, and the determination of the Fe nutritional status of these organisms in March to November via photosynthetic efficiency (FV/FM). These studies will further elucidate the potential for Fe deficiency and carbon sequestration in the warming north throughout the entire growing season.            56 Chapter 3: Non-reductive Fe transport from an Fe-ligand complex by the model diatom Thalassiosira oceanica and in situ plankton communities along Line P in the Subarctic Pacific Ocean   3.1 Introduction From the conception of phytoplankton nutrition research in the late 1920s until present day, the paradigm of bioavailable iron (Fe) has shifted several times. Early Fe bioavailability studies determined that free Fe(III) was the only Fe fraction accessible to phytoplankton (Hopkins, 1930), which remained the tenet of bioavailable Fe until improved analytical techniques allowed Fe uptake experiments with well-defined Fe speciation. Later investigations reported that inorganically-complexed Fe was accessible to phytoplankton (Anderson and Morel, 1982). Once more, this was widely accepted as the complete picture of bioavailable Fe; however, as studies continued to investigate the biogeochemistry and bioavailability of Fe in oceans, this concept of Fe nutrition became the subject of a paradox. Iron readily precipitates out of solution as oxyhydroxide minerals (Whitfield and Turner, 1987), and over 99 % of the remaining minute dissolved fraction of Fe is bound by organic ligands  (Rue and Bruland, 1997). These ligands have been isolated from seawater (Macrellis et al., 2001) and possess functional groups that bind Fe with an affinity typical of siderophores (Witter et al., 2000; Barbeau et al., 2001), the most tenacious Fe chelators in nature (Neilands, 1995).  Coupled with the extremely low supply of Fe from dust and upwelling (Duce, 1986; Martin and Gordon, 1988; Duce and Tindale, 1991), the presence of excess siderophores regulates inorganic Fe concentrations as low as 0.01-0.1 pmol L-1, well below that needed to sustain phytoplankton relying exclusively on this Fe species (Rue and Bruland, 1995; 1997).  57 With this discovery, it was evident that the organically-bound Fe pool must be accessible to phytoplankton by yet-uncharacterized Fe uptake mechanism(s). Research in the past decade has revealed that when Fe-limited, the model marine diatom species Thalassiosira oceanica and Thalassiosira pseudonana can access organically-bound Fe via a reductive process (Maldonado and Price, 1996; 1999; 2000; Maldonado et al., 2001; 2006). This process is analogous to the reductive Fe uptake pathway of the baker’s yeast Saccharomyces cerevisiae (Dancis et al., 1994), in that a non-specific plasmalemma ferric chelate reductase reduces the ligand-bound Fe(III) to Fe(II) to free the Fe from the organic ligand complex. The reduced free Fe is then reoxidized to Fe(III) by a multi-copper oxidase reoxidizes. The ferric ion is then transported into the cell via a permease, completing this highly Fe-specific transport mechanism (Maldonado et al., 2006). As previously, this newly discovered Fe uptake mechanism was believed to complete our understanding of Fe acquisition mechanisms in marine phytoplankton. However, while demonstrating reductive Fe uptake by Fe-limited T. oceanica and T. pseudonana in the presence of a strong organic ligand, Maldonado et al. (2006) noted that surface area-normalized Fe uptake rates were two times faster by T. oceanica than T. pseudonana, despite having virtually identical oxidation and reduction rates. This result suggested that either T. oceanica, the oceanic strain has a system with higher affinity for Fe than T. pseudonana, the coastal strain, or T. oceanica possesses an additional Fe transport system for siderophore-like complexes that functions independently of a reductase system. To investigate these possibilities, a comparative study was conducted (unpublished, Figure 3.1, Semeniuk 2007). First, Fe transport kinetics of Fe-limited T. oceanica and T. pseudonana were examined with the siderophore desferrioxamine B (DFB), a commercially available siderophore routinely used in Fe uptake experiments. Over an increasing  58  Figure 3.1: Short-term uptake rates of Fe(III) ($10-20 mol Fe µm-2 h-1) bound to ferrioxamine B (DFB) in a 1:2 molar ratio as a function of Fe concentration. Data points are a) duplicates of Fe-limited (10 nmol L-1 Fe) T. oceanica cultures and b) an Fe-limited (40 nmol L-1 Fe) T. pseudonana culture. The average growth rate of duplicate 2 L cultures of T. oceanica was 0.98 ± 0.060 doublings d-1 and average cell density was 14 ± 2 $ 104 cells mL-1 when harvested. The growth rate of a 2 L culture of T. pseudonana was 1.7 doublings d-1 and cell density was 33 $ 104 cells mL-1 when harvested.  59 concentration gradient of 55Fe bound with DFB in a 1:2 molar ratio, the extremely tenacious DFB outcompeted the cell surface transporters for Fe; this was demonstrated by the almost complete inhibition of Fe uptake by T. pseudonana at higher FeDFB concentrations (Figure 3.1), preventing the diatoms from reaching a saturated rate of Fe transport (unpublished, Semeniuk 2007). In contrast, T. oceanica was capable of maintaining a saturated Fe uptake rate at extremely high concentrations of FeDFB (Figure 3.1). The lack of transport inhibition at such high DFB concentrations, which are expected to outcompete with cell surface transporters for free Fe(III) (Emery, 1986), suggested that T. oceanica possesses a non-reductive Fe uptake mechanism. As a result, a second investigation was carried out with gallium. Unlike Fe(III), Ga(III) has no stable divalent state (Emery, 1986); however, it can displace Fe(III) from siderophores due to its kinetic lability and thermodynamic stability (Emery and Hoffer, 1980). As such, inhibition of Fe uptake following a Ga addition would be indicative of non-reductive uptake mechanism. Indeed, an Fe uptake experiment was conducted with equimolar 55FeDFB and GaDFB, and resulted in 28 % inhibition of 55Fe uptake rate in T. oceanica. The results of these experiments suggest the presence of a non-reductive Fe uptake system in T. oceanica. In bacteria and yeast, two non-reductive uptake mechanisms have been found: whole-siderophore uptake and ligand exchange. The yeast Saccharomyces cerevisiae has been shown to transport the entire Fe-siderophore complex (Lesuisse and Labbe, 1989; Lesuisse et al., 1998; 2001). In S. cerevisiae, a high-affinity permease, regulated by a SIT1 (Siderophore Iron Transport) gene, transports ferrioxamine siderophores into the cell where the complex is dissociated; this non-reductive Fe uptake mechanism is induced when the organism becomes Fe- limited. In freshwater bacteria Aeromonas hydrophilia (Stintzi et al., 2000), Fe is transported via ligand exchange (Figure 3.2). With this mechanism, a desferrated siderophore bound by a  60  Figure 3.2: Ligand exchange Fe uptake mechanism by Aeromonas hydrophilia: a) a ferrated donor siderophore (square) approaches a desferrated siderophore (oval) bound by a receptor protein in the cell membrane (in grey), b) the donor siderophore and membrane-bound siderophore undergo an exchange of Fe(III), c) the exchange of Fe(III) to the membrane-bound siderophore induces a conformational change in the receptor protein, allowing the the now-ferrated, membrane-bound siderophore to move into the periplasmic space, d) the receptor protein then binds the now-desferrated donor siderophore to the cell surface, returns to its initial conformation, and the cycle begins again. Figure based on Stintzi et al., 2000.  receptor protein in the cell membrane undergoes an exchange of Fe(III) with a donor ferrated siderophore that comes within proximity of the receptor protein in the substrate. The exchange of Fe(III) to the membrane-bound siderophore induces a conformational change in the receptor protein, facilitating movement of the now-ferrated, membrane-bound siderophore into the periplasmic space, after which the now-desferrated donor siderophore is bound to the receptor protein for further Fe exchange with ferrated siderophores in solution (Figure 3.2; Stintzi et al., 2000). This process may occur with different siderophore molecules in solution.  61 The objective of this study was to investigate whether a non-reductive Fe uptake (NRU) mechanism exists in the model diatom T. oceanica. To do this, we methylated the terminal amine group on the siderophore DFB with 14C methyl iodide to create a novel 14DFB compound. Following this process, we sought to detect the internalization of the 55Fe14DFB complex by T. oceanica. We also investigated the presence of NRU in field populations along Line P in the subarctic Pacific. This transect is characterized by a range of Fe concentrations, from the Fe- replete coastal waters off Vancouver Island to the Fe-deficient waters of the open ocean (LaRoche et al., 1996). During this field aspect of the investigation, we sought to answer two fundamental questions. First, do in situ marine plankton communities utilize non-reductive Fe uptake mechanisms? Secondly, does utilization of this mechanism differ between Fe-limited and Fe-replete areas? The lengthy methods development process, as well as the laboratory and fieldwork evidence in support of this mechanism, are discussed below.  3.2 Materials and methods 3.2.1 Culturing techniques Thalassiosira oceanica (CCMP 1003) was obtained from the Bigelow Laboratory for Ocean Sciences (West Boothbay Harbor, ME, USA) and grown in the artificial ocean water medium Aquil. As described in Price et al. (1989), synthetic ocean water (SOW) was enriched with 10 "mol L-1 PO43- and 100 "mol L-1 SiO32- prior to removal of metal contamination via a column containing Chelex ion exchange resin (Bio-Rad). Following the chelexing step, the media (pH 8.2) was sterilized and then enriched with filter-sterilized 300 "mol L-1 chelexed NO3- and vitamins (Price et al., 1989). Trace elements Mn, Zn, Co, Mo, Se, and Cu were also added bound to 100 µmol L-1 EDTA at the concentration specified in Maldonado et al. (2006). Finally,  62 the Fe-limiting media was enriched with 1.28 nmol L-1 (pFe 22) or 13.7 nmol L-1 (pFe 21) Fe, following complexation with EDTA (FeEDTA 1: 1.05). Trace metal speciation in our medium was calculated using MINEQL, a chemical equilibrium calculating software (Westall et al. 1976). Once all additions had been made, the Aquil was allowed to equilibrate for at least 24 h before use and was stored in trace-metal-clean polycarbonate bottles.  Thalassiosira oceanica was grown at 19 ± 1 oC in trace-metal-clean 28 mL polycarbonate tubes and continuously illuminated with cool-white fluorescent lights (150 "mol quanta m-2 s-1; Sylvania, Mississauga, Canada). Axenic semi-continuous batch cultures were maintained in exponential growth phase. Growth acclimation was achieved when growth rates of successive transfers varied by less than 10 % (Brand et al., 1981). Daily biomass measurements were determined via in vivo chl a fluorescence using a Water Properties Turner 10 AU fluorometer. Growth rates (d-1) were determined from linear regressions of ln in vivo chl a fluorescence versus time. Cell size and density were determined with a Coulter Z2 Particle Count and Size Analyzer (Beckman Coulter Inc., Brea, CA, USA). In preparation for the short-term uptake experiments outlined below, 1 L trace metal clean polycarbonate bottles were inoculated with acclimated, Fe- limited cultures from the 28 mL tubes. All manipulations of culture tubes and experimental bottles were done in a laminar flow hood with sterile, trace metal clean techniques.  3.2.2 Laboratory short-term uptake experiments Preparation of 55Fe-ligand complexes was done as per Maldonado and Price (1999). Once 2-8 hours had elapsed to ensure complete complexation of Fe with the ligand (DFB or EDTA), the mixture was subsequently added to 250 mL of Aquil media per treatment and allowed to equilibrate chemically overnight. The following day, duplicate 1 mL initials were sampled from  63 each bottle to determine their specific activities. Three 250 mL aliquots of a 1 L Fe-limited culture of T. oceanica in mid-exponential growth phase were then filtered by gentle vacuum on trace metal clean 2 µm polycarbonate Poretics® filters of 47 mm diameter. The filters were rinsed with 10 mL chelexed SOW before going dry, then immediately resuspended in the 250 mL treatment bottles. Shortly thereafter, each bottle was sampled for ~ 1 mL which was fixed with Lugol’s solution and placed aside for cell counting with a Coulter Z2 Particle Count and Size Analyzer (Beckman Coulter, Inc.) within 24 hours of the uptake experiment. Cell density of experimental cultures typically ranged between 210,000-280,000 cells mL-1. Mean cell diameter of Fe-limited T. oceanica was ~ 4.3 "m. Treatment bottles were first sampled for Fe uptake within 15 minutes of resuspension, after which sampling occurred in 30-60-minute intervals for up to 5 hours. Duplicate 25 mL samplings were taken from each treatment bottle and filtered by gentle vacuum (< 5 PSI) onto 3 µm polycarbonate Poretics® filters of 25 mm diameter. Before going dry, filters were soaked with 5 mL of 0.2 µm filtered Ti(III) citrate EDTA reagent for 5 minutes to dissolve Fe oxides adsorbed to cell surfaces (Hudson and Morel, 1989), then rinsed with 5 mL of SOW. Following this, the filters were transferred to scintillation vials and saturated with 5 mL ScintiSafe Plus 50 % scintillation cocktail (Fisher). 55Fe activities were determined with a Beckman Coulter LS 6500 Multi-Purpose Scintillation Counter. Linear regression was used to calculate the rate of 55Fe accumulation over time (mol Fe L-1 h-1). Iron uptake rates (!Fe) were normalized by the specific activity of 55Fe to calculate total volumetric Fe uptake rates. These volumetric !Fe were then normalized to cell densities and cell surface areas to calculate mol Fe µm-2 h-1.    64 3.2.3 Detecting ligand uptake Inspired by the FeDFB transport kinetics experiments described previously, a method for detecting internalization of FeDFB by phytoplankton was developed to investigate non-reductive Fe update by phytoplankton. This method consisted of using 14C to label DFB, and subsequently detecting its internalization during Fe uptake experiments. Since our laboratory shares the same campus as TRIUMF, Canada’s national particle and nuclear physics lab, we aimed to develop a 14C-labeling protocol for DFB with scientists at TRIUMF.  3.2.4 Developing a 14C-labeling protocol for DFB The process of synthesizing 14C-labeled DFB was defined by two phases: labeling and verification. During the labeling phase, the chemical methodology of labeling DFB with 14C was tested and determined. The verification phase sought to characterize the 14C-labeled DFB compound. The labeling phase began with a preliminary test in December 2008. Short-term 55Fe uptake experiments were conducted to compare two potential intermediate DFB molecules in the novel 14C-labeling process. This experiment sought to determine whether complexation of Fe by these DFB molecules was identical or whether synthesizing the novel compound had altered their ability to bind Fe. The starting, unaltered molecule was desferrioxamine mesylate (Desferal®, Novartis), the commercially available DFB compound routinely used in Fe uptake experiments. The altered ligand was synthesized by replacement of the terminal amine group on DFB with a hydrogen chloride (Figure 3.3). For both treatments, 2 nmol L-1 55FeCl3 (PerkinElmer) + 8 nmol L-1 FeCl3 were complexed with 15 nmol L-1 of the DFB compound and later added to a 250 mL bottle containing Aquil media to equilibrate chemically overnight.  65   Figure 3.3: Chemical structures of a) desferrioxamine mesylate (brand name Desferal®), b) hydrogen chloride-labeled DFB, c) methylate salt-labeled DFB, d) 14C-labeled DFB (14DFB), and e) 14DFB complexing 55Fe.   66 Following the uptake experiment and analysis procedures outlined above with an Fe- limited T. oceanica culture grown in media enriched with 1.28 nmol L-1 Fe, it was determined that the modified DFB compound with the hydrogen chloride functional group facilitated a linear cell surface area-normalized !Fe nearly identical to that of the mesylate salt control treatment (Table 3.1). As a result, it was confirmed that this modified DFB compound would be a suitable intermediate in the 14C-labeling process.  Table 3.1: Short-term Fe uptake rates ($10-22 mol Fe µm-2 h-1) of a single culture of Fe-limited (1.28 nmol L-1 Fe) T. oceanica. Fe was precomplexed in a 10 nmol L-1 Fe : 15 nmol L-1 DFB ratio with each of desferrioxamine mesylate (brand name Desferal®) and hydrogen chloride-labeled DFB. The growth rate of a 1 L culture of T. oceanica was 0.69 doublings d-1, cell density was 29 $ 104 cells mL-1, and mean cell size was 4.8 µm when harvested. Reported values are equal to the mean ± the range for pseudoreplicates.     In the spring of 2010, additional testing was done to determine whether the solvent used during the 14C-labeling process affected later complexation of DFB with Fe. Three experiments determined the short-term 55Fe uptake rates by Fe-limited (13.7 nmol L-1 Fe) T. oceanica for DFB compounds with or without traces of solvent N,N-dimethyl formamide anhydrous 99.8 % (DMF; Table 3.2). First, we compared the desferrioxamine mesylate control treatment with the HCl-labeled DFB to ensure they elicited comparable Fe uptake rates; the latter treatment had no traces of solvent DMF. The second and third experiments included a third treatment of HCl- labeled DFB with DMF traces. For each of the DFB treatments, 1 nmol L-1 of 55Fe was bound to 10 nmol L-1 of the DFB compound. It was observed that the HCl-labeled DFB treatments with or without DMF were associated with very similar !Fe (Table 3.2); however, the treatment with no Treatment !Fe mesylate DFB 4.6 ± 0.28 HCl-labeled DFB 6.8 ± 0.15  67 Table 3.2: Short-term Fe uptake rates ($10-22 mol Fe µm-2 h-1) of Fe-limited (13.7 nmol L-1 Fe) T. oceanica cultures. Fe was precomplexed to DFB in a 1 nmol L-1 Fe : 10 nmol L-1 DFB ratio. Reported values are equal to the mean ± the range for pseudoreplicates. The growth rate of a 0.5 L T. oceanica culture for experiment 1 was 0.95 doublings d-1; cell density was 23 $ 104 cells mL-1, and mean cell size was 4.5 µm. For experiment 2, the growth rate of a 1 L culture was 1.2 doublings d-1, the cell density was 23 $ 104 cells mL-1, and mean cell size was 4.3 µm. Experiment 3 was characterized by a growth rate of a 1 L T. oceanica culture of 1.0 doublings d-1, a cell density of 13 $ 104 cells mL-1, and a mean cell size of 4.2 µm.  !Fe Treatment Experiment 1 Experiment 2 Experiment 3 mesylate DFB 14 ± 0.14 7.7 ± 0.065 5.5 ± 0.36 HCl-labeled DFB, no solvent 25 ± 1.2 18 ± 0.43 81 ± 5.6 HCl-labeled DFB, with solvent traces  23 ± 0.11 63 ± 3.3  DMF achieved a marginally more linear !Fe. As such, HCl-labeled DFB with no DMF was selected as the more appropriate intermediate DFB complex for this investigation.  3.2.5 Labeling DFB with 14C These preliminary experiments led to a method of labeling HCl-labeled DFB with 14C methyl iodide (50-62 mCi mmol-1, Quotient Bioresearch), including the evaporation of DMF traces. In the early summer of 2010, 2 mg HCl-labeled DFB were immersed in 10 µL of 10 N Seastar NaOH in a small glass vial in a fume hood and immediately stirred with a stirbar and stirplate. In the same fume hood, acetone and dry ice were placed in a beaker containing a break- seal ampoule of 5 mCi 14C methyl iodide. This facilitated condensation of the isotopic gas to prevent escape upon the breakage of the ampoule; additionally, 0.5 mL of DMF was pipetted into the top reservoir of the break-seal ampoule, which was then shattered, allowing the DMF to react with the 14C gas. This mixture was immediately added to the DFB vial with a disposable glass pipette, leaving the inert Cu metal in the ampoule. The vial was sealed with a crimped aluminum lid and placed inside a larger beaker with near-boiling water on a hotplate. At 80 oC,  68 the 14C-DFB reaction took approximately 1 hour, as indicated by a change in solution from colourless to yellow. Following this, 10 µL of 12 N Seastar HCl were added to neutralize the NaOH, and methanol was added to facilitate the evaporation of DMF. The vial was left stirring in the fume hood for 1 week, after which a sticky, yellow powder of 14C-labeled DFB (14DFB) remained. The 14DFB was dissolved in 1 mL deionized water; assuming 100 % reaction, this yielded a calculated concentration of 3.5 mmol L-1 in our 14DFB stock solution at a pH of ~ 4.5.  3.2.6 Laboratory testing and use of 14DFB To verify the concentration and specific activity calculated for the 14DFB solution, a chrome azurol S (CAS) assay was performed following the procedure outlined in Schwyn and Neilands (1987). Six experimental solutions were analyzed with a Cary 100 BIO UV-Visible Spectrophotometer; these mixtures contained 0.5 mL of the CAS shuttle solution and 0.5 mL of a unique siderophore concentration per treatment. Five (cold) siderophore solutions were prepared, spanning a DFB concentration range of 0.4-1.5 % 10-5 mol L-1. One 14DFB solution was tested, with a calculated concentration of 0.7 % 10-5 mol L-1. In addition, a control CAS reference solution was prepared with 0.5 mL of the CAS shuttle solution and 0.5 mL of deionized water. Triplicate assays revealed that absorption of 14DFB was less than that of the CAS reference solution (Figure 3.4; not all cold siderophore solutions shown). As explained in section 3.3.1.1, we believe these negative results were obtained due to contamination of the 14DFB stock solution with Fe. As a result, we utilized the ferrioxamine B hydrolysis protocol outlined in Bickel et al. (1960) as an attempt to remove the putative Fe contamination from our 14DFB. One modification was made to the protocol; instead of adding ether to an acidified subsample of the contaminated 14DFB and performing a continuous extraction, a subsample of contaminated 14DFB was run  69   Figure 3.4: CAS assay results, including (a) the absorbance spectrum for various CAS-*DFB (DFB or 14DFB) solutions, an (unchelexed) CAS solution, and water, and (b) absorbance as a function of *DFB concentration. 14DFB data in both figures represent an average from duplicate assays.  70 through a Waters Sep-Pak® C18 W108201 cartridge in an attempt to elute the Fe with a pH 2 solution. Following the column extraction, another CAS assay was performed and yielded similarly negative results, indicating the failure of the extraction or very low 14DFB concentration. Despite the unverified 14DFB concentration, we proceeded with duplicate short-term 55Fe14DFB uptake experiments to determine whether 14DFB uptake occurs in T. oceanica. During each uptake experiment, two treatments were tested in duplicate: 55Fe14DFB at room temperature and 55Fe14DFB at near-freezing temperatures. The cold control was utilized to inhibit biological activity without denaturing cell walls; it was expected that negligible concentrations of 55Fe14DFB would be internalized at 0 oC relative to the experimental treatment. To account for the assumed Fe contamination of 14DFB, a modified methodology was developed in preparing the uptake media. We assumed that Fe had contaminated our DFB stock solution (3.5 mmol L-1) and subsequently our secondary dilution (3.5 % 10-5 mol L-1) of 14DFB in a ratio of 1 mol Fe : 2 mol DFB. To compensate for this, 5 nmol L-1 of 55Fe was complexed with 10 nmol L-1 DFB and 10 nmol L-1 14DFB (the latter putatively bound to ~ 5 nmol L-1 Fe from contamination) per treatment. Immediately following complexation, an addition of Seastar HCl was made to lower the pH to 1.7. At this pH, the presumed Fe contamination dissociated from 14DFB and the four compounds intermixed. After 2 hours, 1.5 mL of SOW were added to facilitate equilibration of the mixture at a higher pH without precipitating Fe out of solution; at this stage the solution changed from clear to yellow, indicating the complexation of Fe bound to DFB. Following another 2 hour period, the mixture was added to trace metal clean 125 mL polycarbonate bottles containing Aquil media and allowed to equilibrate chemically overnight. The following day, 125-mL aliquots of an Fe-limited (1.28 nmol L-1 Fe), 500 mL culture of T. oceanica in mid-  71 exponential growth phase were filtered and resuspended as described above. Media and cell density initials were also acquired as previously described. Immediately after resuspension, one bottle was placed in an ice bath. Temperature immediately outside of the bottle was monitored throughout the experiment and remained just above 0 oC. Uptake rates for both Fe and DFB were determined as above, although sampling volumes were smaller (10 mL) due to higher radioactivity of the isotopes per bottle.  3.2.7 Field sites Shortly following laboratory testing of 14DFB, a field investigation of NRU was conducted aboard the CCGS John P. Tully at four stations along Line P in the subarctic Pacific Ocean from late August to early September 2010. Sampling this transect allowed for an investigation of NRU by phytoplankton inhabiting areas characterized by a range of Fe nutritional states, from an Fe-sufficient coastal station (P4) to the Fe-deficient Ocean Station PAPA (P26; OSP). At each station, P4 (48o39.00’N, 126o40.02’W), P16 (49o16.92’N, 134o39.96’W), P20 (49o33.86’N, 135o40.00’W), and P26 (50o00.10’N, 144o59.96’W), a short- term 55Fe14DFB uptake experiment was conducted. In addition, GF/F and size-fractionated chlorophyll a (chl a), nutrients, heterotrophic and autotrophic bacteria flow cytometry, and light microscope phytoplankton identification samples were collected. Finally, a rosette equipped with Niskin bottles (OceanTest Equipment), a Sea-Bird 911plus CTD probe, a Biospherical QSP-400 PAR sensor, and a Wet Labs C-Star transmissometer was cast at each station. All incident irradiance (Io) data reported here were obtained between 9 AM and 5 PM on the day the seawater collection for the 55Fe14DFB uptake experiments took place. Data from triplicate casts at each station were utilized to estimate mixed layer depths (MLD).  72 3.2.8 Biological and chemical parameters At each station, seawater was pumped from 10 m depth with a trace metal clean Teflon® peristaltic pump and tubing system suspended by a Kevlar® line (Johnson et al., 2005). Seawater was transferred to trace metal clean bottles for station biological and chemical parameters, as well as to 9-2 L polycarbonate bottles for the 55Fe14DFB uptake experiments. Along Line P, total chl a was determined from ~ 250 mL of filtered seawater which was filtered with GF/F filters by gentle vacuum (< 100 mm Hg) and subsequently frozen at &80 °C. Additionally, to determine size-fractionated chl a, ~ 500 mL of seawater were collected and filtered on 0.22, 1, and 5 µm Poretics® polycarbonate membranes separated by Millipore nylon mesh drain disks. While on the ship, batches of individually-preserved frozen filters were treated with 90 % acetone to extract chl a. Following a 24-hour acetone extraction in the dark at &20 °C, concentrations of chl a were determined with a Water Properties Turner 10 AU fluorometer. Filtered seawater samples were collected for macronutrients (nitrate, NO3-, phosphate, PO43-, and silicate, SiO42-) and analyzed based on Barwell-Clarke and Whitney (1996). Dissolved Fe concentrations were determined as described in Johnson et al. (2005). Light microscopy phytoplankton identification samples were collected in 250 mL amber bottles and preserved with 0.2 % gluteraldehyde. Once we returned to The University of British Columbia (UBC), these samples were enumerated with a Zeiss Axiovert 10 microscope. Triplicate samples for heterotrophic and autotrophic bacteria flow cytometry were dispensed into cryovials containing a 10 % formaldehyde, 0.5 % gluteraldehyde and 40 % phosphate-buffered fixative solution based on Marie et. al (2005). These 1-mL samples were immediately stored in the dark for 15 minutes, after which they were flash frozen in liquid nitrogen and stored at &80 °C. Upon arrival to UBC, a FACSCalibur flow cytometer (Becton-Dickinson) equipped with a  73 15-mW, 488-nm laser was used to count and characterize heterotrophic and autotrophic bacteria. After thawing to above-freezing temperatures, 50 "L were taken for heterotrophic bacteria analysis and the remainder was analyzed for autotrophic bacteria. An addition of FluoroSpheres carboxylate modified yellow-green 1-"m beads (Molecular Probes; ~ 2 % 10-5 mL-1 final concentration) was added to both subsamples as an internal reference. Heterotrophic bacteria flow cytometry subsamples were then diluted 10-fold in 0.02 "m pore-size filtered TE buffer (10 mmol L-1 Tris, 1 nmol L-1 EDTA, pH 8.0) to avoid coincidence and stained with SYBR Green-I (Invitrogen). Heterotrophic and autotrophic bacteria samples were then incubated and analyzed as per Marie et al. (1999) and Martinez-Martinez et al. (2006). The list files generated by the instrument were then analyzed with Cytowin software (Vaulot, 1989), which distinguishes organism populations based on their natural fluorescence and scatter of instrument lasers. The discrimination of autotrophic bacteria was done via scatter plots of orange and red fluorescence; heterotrophic bacteria were quantified via scatter plots of side scatter signal vs. SYBR Green-I fluorescence. Cell density was calculated from scatter plot cell counts, instrument flow rate, and analysis time.  3.2.9 Short-term 55Fe14DFB uptakes along Line P At each station, we investigated NRU using 55FeCl3 (4.91 ! 103 Ci mol-1, PerkinElmer) and our 14DFB. As described above, seawater was pumped from 10 m depth into 9 trace metal clean 1 L polycarbonate bottles. Three treatments were tested in triplicate: a) 55FeDFB (the control group utilizing desferrioxamine mesylate), b) 55Fe14DFB, and c) 55Fe14DFB + Ga14DFB (Table 3.3). Preparation of these treatments was done as mentioned above for the laboratory 14DFB uptakes. In a small Teflon® vial dedicated to the control treatment, 10 nmol L-1 DFB were  74 added to 5 nmol L-1 Fe with 20 % of total Fe as 55Fe (Table 3.3). The 55Fe14DFB treatment was identical to that of the control, plus an addition of 10 nmol L-1 14DFB, which was putatively bound to 5 nmol L-1 Fe from contamination (Table 3.3). This maintained an Fe : DFB molar ratio of 1 : 2 as in the control. Finally, the 55Fe14DFB + Ga14DFB treatment was identical to that of the 55Fe14DFB treatment, plus an addition of 5 nmol L-1 Ga (Table 3.3). As Ga(III) cannot be reduced, it was expected that this treatment would yield reduced !55Fe and virtually identical !14DFB relative to the 55FeDFB and 55Fe14DFB treatments. Immediately following the additions of Fe, 55Fe, and DFB (and 14DFB and Ga where applicable) to the small Teflon® vials, an addition of Seastar HCl was made to lower pH to 1.7. At this pH, the presumed Fe contamination dissociated from the 14DFB, and the compounds intermixed. After 2 hours had passed, 1 mL of filtered seawater was added to each mixture to facilitate equilibrium at a higher pH. Following another 2 hours, these mixtures were added to the treatment bottles. All enrichments and manipulations were carried out within a laminar flow hood with trace metal clean techniques. Following addition of isotopes, bottles were tightly capped, sealed with Parafilm®, and returned to on-deck incubators which were kept at in situ temperature (14 ± 1.5 °C) and at 10 %Io.   Table 3.3: Control and experimental treatments utilized in short-term 55Fe14DFB uptake rate experiments along Line P. Mixtures of Fe and DFB (and Ga if applicable) are outlined for each treatment; components marked with an X were not added to the respective treatment. Each treatment contained a 1 : 2 molar ratio of Fe : DFB.   Treatments Fe 55Fe DFB 14DFB Ga Total Concentrations 55FeDFB 4 nM 1 nM 10 nM   5 nM Fe 10 nM DFB 55Fe14DFB 4 nM 1 nM 10 nM 10 nM (+ 5 nM Fe contamination) 10 nM Fe 20 nM DFB 55Fe14DFB + Ga14DFB 4 nM 1 nM 10 nM 10 nM (+ 5 nM Fe contamination) 5 nM 10 nM Fe, 5 nM Ga 20 nM DFB  75 At station P4, bottles were sampled following 8, 16, and 24 h of incubation; all other stations’ bottles were sampled at 8 h. All filtrations carried out at 8 h utilized a ~ 500 mL aliquot from each sample bottle, whereas the 16 and 24 h samplings at P4 each utilized ~ 250 mL. During each sampling, seawater was filtered onto 0.22, 1, and 5 "m pore size Poretics® polycarbonate filters of 47 mm diameter, separated by nylon drain disks. Before going dry, filters were soaked with 5 mL of 0.2 "m filtered Ti(III) citrate EDTA reagent for 5 minutes to dissolve Fe oxides adsorbed to cell surfaces (Hudson and Morel, 1989), then rinsed with 5 mL of SOW. Following this, the filters were transferred to scintillation vials and saturated with ScintiSafe Plus 50 % scintillation cocktail (Fisher). 55Fe and 14C activities were determined with a Beckman Coulter LS 6500 Multi-Purpose Scintillation Counter upon return to UBC two weeks later. Size-fractionated carbon-normalized !Fe (mol Fe (mol C)-1 d-1) and !DFB (mol DFB (mol C) -1 d-1) for all stations were calculated by dividing particulate Fe or DFB (mol L-1) by carbon concentrations (mol C L-1) and time elapsed (h). Size-fractionated chl a (µg L-1) concentrations for each station were utilized to derive size-fractionated phytoplankton C (mol C L-1) with a C:chl a ratio of 100 g g-1 for the Fe-limited station P26 and 50 g g-1 for stations P4, P16, and P20. These values have previously been determined to be representative of C:chl a ratios in the subarctic Pacific (Booth, 1988; Welschmeyer et al., 1991; Booth et al., 1993; Kirchman et al., 1993; Welschmeyer et al., 1993; Boyd et al., 1995). Additionally, in the 0.22-1 µm size-fraction, heterotrophic bacteria counts (cell mL-1) were utilized to derive heterotrophic bacterial C (mol C L-1) using a conversion factor of 25 fg C cell-1 (Lee and Fuhrman, 1987; Maldonado and Price, 1999). The total C in the 0.22-1 µm size fraction was then calculated as the sum of the estimated heterotrophic bacterial carbon and that of the picophytoplankton, which was derived from the chl a in the 0.22-1 µm size-fraction and the C:chl a conversion factor. One-  76 way ANOVAs and subsequent Tukey tests were carried out to assess the effects of treatments or stations (the latter as a proxy of phytoplankton Fe nutritional status) on C-normalized !Fe or !DFB. This statistical testing was carried out separately for each time interval at station P4 (8, 16, and 24 h) and for each size fraction.  3.3 Results and discussion 3.3.1 Laboratory results 3.3.1.1 Fe contamination of 14DFB As a consequence of utilizing 5 mCi gaseous 14C methyl iodide to label DFB, it was necessary to carry out the labeling procedure in a (metal) fume hood in our designated radiation lab. Unfortunately, this crucial safety measure sacrificed our preference to conduct the procedure in a trace metal clean environment and as a result, putative Fe contamination compromised the integrity of this investigation. Indeed, the CAS assay, which is dependent upon desferrated DFB, was unable to verify the calculated concentration of 14DFB. Furthermore, an attempted extraction conducted mere days prior to Line P was unable to reverse the contamination. Luckily, pH manipulations salvaged the ability to bind 14DFB with 55Fe for our field study, thus we can still draw some conclusions from this investigation as to the potential of NRU in cultured and in situ plankton.  3.3.1.2 Non-reductive Fe uptake by T. oceanica Preliminary laboratory experiments were conducted prior to fieldwork to investigate whether 55Fe14DFB uptake would be observed in T. oceanica.  Linear uptake rates for both 55Fe and 14DFB were expected in the experimental treatment (55Fe14DFB), with that of 55Fe exceeding  77 that of 14DFB, as no inhibitor had been used to impede the reductive Fe uptake mechanism. In the cold control treatment (55Fe14DFB at near-freezing temperatures), biological activity was hampered, thus negligible concentrations of 55Fe14DFB would be internalized by the cells relative to the experimental treatment; this result would increase our confidence as to the successful labeling of DFB with 14C. Due to the low sampling volumes, 14DFB activities were below background (< 100 dpm) in one of two duplicate experiments, obscuring uptake rates and statistical differences; thus only one set of results is presented and discussed. Uptake of both Fe and DFB as a function of time by T. oceanica was linear (!Fe  = 5.47x [10-21 mol Fe µm-2 % h-1], R2 = 0.98; !DFB = 7.85x [10-21 mol DFB µm-2 % h-1], R2 = 0.93; Figure 3.5), and yielded rates comparable to an analogous laboratory experiment conducted with cold methyl-labeled DFB, which yielded a !Fe of 4.1 ± 1.2 % 10-21 mol Fe µm-2 h-1. The absolute rate of DFB uptake (mol DFB µm-2 h-1) was 1.4 times faster than that of Fe; applying the Student’s t test to assess the equality of these two slopes (from Zar, 1999) reveals that this difference is significant (t0.05(2)10 = 4.9, p < 0.05, 5 df). Uptake rates of Fe and DFB in the cold control treatment were negligible (!Fe = 0.32x [10-21 mol Fe µm-2 % h-1], R2 = 0.04, and !DFB = 3.44x [10-21 mol DFB µm-2 % h-1], R2 = 0.55, respectively), leading to significantly different uptake rates between their corresponding cold control and experimental treatments (for Fe: t0.05(2)10 = 5.8, p < 0.05, 5 df; for DFB: t0.05(2)10 = 4.8, p < 0.05, 5 df). These experiments suggest that T. oceanica has the ability to internalize FeDFB as a whole siderophore complex. However, faster rates of uptake for Fe than for DFB were expected. The faster rates of DFB uptake may be an artifact of our inability to accurately estimate the concentration of the radiolabeled DFB, and thus its specific activity. To make our results conclusive, future experiments are needed using a purified radiolabeled DFB.  78  Figure 3.5: Short-term Fe and DFB uptake rates ($10-21 mol µm-2 h-1) of a Fe-limited (1.28 nmol L-1 Fe) culture of T. oceanica. Uptake was determined for both an experimental treatment and a cold-control treatment, which was subjected to near-freezing temperatures. For both treatments, Fe was precomplexed to DFB in a 5 nmol L-1 Fe : 20 nmol L-1 DFB ratio, with all of added Fe as 55Fe and 50% of added DFB as 14DFB. Additionally, it was estimated that ~ 5 nmol L-1 Fe contamination was present in the 14DFB solution, yielding an estimated final Fe : DFB ratio of 10 nmol L-1: 20 nmol L-1. The growth rate of a 1 L culture was 0.46 doublings d-1, cell density was 28 $ 104 cells mL-1, and cell size was 4.3 µm when harvested. All treatments exhibited significantly different slopes. Linear regressions: Fe uptake (y = 5.47x, R2 = 0.98), DFB uptake (y = 7.85x, R2 = 0.93), Fe cold control (y = 0.32x, R2 = 0.04), and DFB cold control (y = 3.44x, R2 = 0.55). Testing the equality of slopes: Fe uptake vs. DFB uptake (t0.05(2)10 = 4.9, p < 0.05), Fe uptake vs. Fe cold control (t0.05(2)10 = 5.8, p < 0.05), DFB uptake vs. DFB cold control (t0.05(2)10 = 4.8, p < 0.05).  3.3.2 Field results 3.3.2.1 Station parameters Along Line P from the eastern coastline to the open ocean in the subarctic Pacific, the local weather deteriorated. Conditions at station P4 were clear and balmy, whereas stations P16 and P20 were misty but calm, and station P26 was windy with rain. Estimated Io values reflect  79 these weather conditions on the associated dates of pumping (Table 3.4). Although the CTD and rosette system was unable to assess parameters at 0 m depth, logarithmic regression was applied to the shallowest PAR readings (< 15 m) at each station and extrapolated to the surface. Malfunctioning of the PAR sensor during deep CTD casts prevented the determination of the base of the euphotic zone at each station. Density data from triplicate CTD casts at each station were used to estimate MLD using a density difference criteria of 0.125 from the surface (Table 3.4; after Levitus, 1982). At station P4, MLD was approximately half as deep as that of the other three stations (Table 3.4). All stations’ MLD were relatively shallow and indicative of late summer shoaling (Varela and Harrison, 1999). At each station, MLD salinity varied less than 1 % from surface salinity; all stations’ salinity were 32-33 throughout the shallowest 100 m. Seawater temperatures at pumping depth (10 m; Table 3.4) were indicative of summer warming; the variability observed between stations demonstrated natural seasonal fluctuations (Varela and Harrison, 1999).  Table 3.4: Coordinates, estimated incident irradiance (Io; µE m-2 s-1), mixed layer depth (MLD; m), seawater temperature at pumping depth (°C), dissolved Fe (nmol L-1), NO3- (%mol L-1), PO43- (%mol L-1), and SiO42- (%mol L-1) at four stations along Line P in the subarctic Pacific Ocean. Irradiances, MLD, and temperature were determined from CTD data. Samples for nutrients were collected at the same time and depth (10 m) as pumping for NRU experiments. Station incident irradiances were estimated via logarithmic regression of deep cast CTD data; R2 values from these analyses are also reported. Stations’ MLD are equal to the mean ± standard error for triplicate CTD casts. Dissolved Fe concentrations were determined by Keith Johnson as described in Johnson et al. (2005). Station P4 macronutrient concentrations are equal to the mean ± the range for duplicate samples. Station P20 and P26 macronutrient are equal to the mean ± standard error for triplicate samples.  Station Latitude/Longitude Io (R2) MLD °C Fe NO3- PO43- SiO42- P4 48o39.00’N, 126o40.02’W 265 (0.97) 12.8 ± 0.967 11.8 0.24 6.5 ± 0.13 0.87 ± 0.14 15 ± 0.47 P16 49o16.92’N, 134o39.96’W 200 (0.99) 25.7 ± 2.08 15.2 0.03 2.3 0.57 5.2 P20 49o33.86’N, 135o40.00’W 197 (0.99) 22.2 ± 4.32 15.6 0.03 2.1 ± 0.30 0.52 ± 0.057 4.0 ± 0.42 P26 50o00.10’N, 144o59.96’W 105 (0.98) 20.0 ± 2.62 13.6 0.02 8.1 ± 0.48 0.88 ± 0.048 13 ± 0.65  80 3.3.2.2 Plankton growth and nutrient drawdown During our voyage westward, biological, chemical, and physical parameters reflected a changing environment from Fe-sufficiency at station P4 to Fe-deficiency at station P26. Seawater obtained from 10 m depth at each station revealed an order of magnitude greater total chl a concentration at P4 than all other stations (Table 3.5). Size-fractionated chl a samples reveal that the 0.22-1 "m size fraction contributed ~ 4-7 % to total chl a at all stations (Table 3.5). Cells in the 1-5 "m size fraction made up 25 % of total chl a at station P4 and ~ 46-55 % at all other stations. Conversely, cells > 5 "m constituted 68 % of total chl a at station P4 and ~ 43- 49 % at all other stations. It is noteworthy that at station P4, seawater pumping occurred in the chlorophyll maximum, very close to the base of the mixed layer, whereas sampling for all other stations occurred in the middle of the mixed layer. Despite this, our chl a values are typical of this region (Maldonado et al., 1999; Varela and Harrison, 1999). Macronutrient levels (NO3-, PO43-, and SiO42-) were virtually identical and relatively high at stations P4 and P26 (Table 3.4). Whereas high nutrient levels are indicative of coastal upwelling and sediment and riverine input at P4 (Johnson et al., 1997), at P26 they are indicative of severe Fe-deficiency to the extent that nutrient uptake cannot proceed (LaRoche et al., 1996; Whitney and Freeland, 1999). Furthermore, both chl a (Table 3.5) and Fe (Table 3.4) concentrations were an order of magnitude greater at P4 than at all other stations, which had equally lower chl a and Fe content. At stations P16 and P20, macronutrient levels were virtually identical and drawn down to low values (Table 3.4), implying that phytoplankton at these stations are not as Fe-stressed as those at P26 (LaRoche et al., 1996).   81 Table 3.5: Total chlorophyll a (µg L-1), size-fractionated chl a (% contribution to total chl a), relative abundance of diatoms, coccolithophores, and dinoflagellates (%), and cell density of picoeukaryotes and heterophic bacteria (cells mL-1) of four locations along Line P in the subarctic Pacific Ocean. Total and size-fractionated chl a, light microscopy, and flow cytometry samples were collected at the same time and depth (10 m) as pumping for NRU experiments. Station P20 and P26 chlorophyll a concentrations and flow cytometry cell densities are equal to the mean ± standard error for triplicate samples. Station P4 and P16 flow cytometry cell densities are equal to the mean ± the range for duplicate samples.   Station chl a 0.22-1 µm 1-5 µm > 5 µm Diatoms Coccolithophores Dinoflagellates Picoeukaryotes Heterotrophic Bacteria P4 2.445 7 % 25 % 68 % 49 % 22 % 29 % 15,222  ± 937 2,217 ± 110 P16 0.358 5 % 46 % 49 % 37 % 1 % 61 % 6,880 ± 189 1,561 ± 5 P20 0.381 ± 0.053 5 % 55 % 40 % 50 % 6 % 44 % 8,415  ± 100 1,123 ± 20 P26 0.482 ± 0.020 4 % 53 % 43 % 53 % 7 % 40 % 12,924  ± 62 897 ± 4     82 3.3.2.3 Plankton community composition and nutritional status Light microscopy revealed high proportions of diatoms at all stations, however the composition of diatom species differentiated coastal station P4 from the three open ocean stations. Thalassiosira sp., Pseudo-nitzschia delicatissima, and Cylindrotheca closterium made up ~ 50 % of the community at both P20 and P26, and 37 % of the population at P16. Station P4 was characterized by a similar abundance of diatoms (49 % abundance); however, the diatom species distribution was more diverse, including species of Chaetoceros (c.f. socialis and borealis). Dinoflagellates and coccolithophores also differentiated the coastal and open ocean environments. Dinoflagellates, including Cryptophyceae, Chlorophyceae, and Prorocentrum sp. (c.f. gracile) made up 29 % of the community at P4, whereas coccolithophores comprised 22 %. In contrast, at stations P16, P20, and P26, dinoflagellates Cryptophyceae and Chlorophyceae accounted for 61 %, 44 %, and 40 % of the population, respectively, while coccolithophores were very low in numbers (1 %, 6 %, and 7 %, respectively).  Flow cytometry and Cytowin analyses determined autotrophic and heterotrophic cell densities at each station. All of cyanobacteria, picoeukaryotes, and heterotrophic bacteria were identified; however, cyanobacteria abundance could not be accurately determined as the signal was consistently cut off by the red fluorescence axis in the Cytowin scatter plots. Despite this, a portion of the cyanobacteria community was captured, and yielded cell densities between 1,800- 5,700 cells mL-1 for stations P16-P26, and a cell density of ~ 37,000 cells mL-1 for station P4. Though inaccurate, these cell densities suggest that cyanobacteria abundance was much higher in nutrient-replete coastal waters than in the more oceanic stations. Picoeukaryote abundance was much less variable across stations. At stations P16 and P20, cell density was 6,880 ± 189 cells mL-1 and 8,415 ± 100 cells mL-1, respectively, half the abundances observed at P4 and P26  83 (Table 3.5). The coastal station had an abundance of 15,222 ± 937 cells mL-1, whereas the most western station was characterized by 12,924 ± 62 cells mL-1 picoeukaryotes. Heterotrophic bacteria abundance followed the same pattern as chl a concentration and the total range in cell density was 1,320 cells mL-1 (Table 3.5). At P4, heterotrophic bacteria abundance was 2,217 ± 110 cells mL-1, whereas P26 was characterized by 897 ± 4 cells mL-1 (Table 3.5).  Nutritional status of the plankton communities at each station, as indicated by nutrient and chl a data, was confirmed by plankton community characterization. First, the notably higher abundance of coccolithophores at station P4 than stations P16, P20, and P26 indicates a more turbulent environment at the coastal station, characteristic of a region with high nutrient input (Margalef, 1978). Indeed, this is supported by the high macronutrients and Fe concentrations determined for P4, which is associated with the largest fraction of cells exceeding 5 !m at all four stations. At stations P16 and P20, low macronutrients, Fe and chl a levels are complemented by a large proportion of cells in the 1-5 !m size fraction and low abundance of picoeukaryotes and heterotrophic bacteria. Together, these characteristics indicate that the plankton communities at these stations are indeed Fe-limited but not deficient, as bacteria and dinoflagellate abundances are higher than that of P26, and nutrient levels are uniformly low. Finally, station P26 had the lowest abundance of heterotrophic bacteria, reflected in the smallest proportion of cells in the 0.22-1 !m size fraction of all stations. This result can be attributed to low Fe concentration, as heterotrophic bacteria are characterized by high Fe demand (Tortell et al., 1996). Additionally, this station boasts the highest proportion of diatoms, yet unlike station P4, is not characterized by a high proportion of cells > 5 !m. Thus it appears that the high proportion of diatoms at station P26 is likely supported by Fe storage in Pseudo-nitzschia sp. during  84 sporadic Fe events (Marchetti et al., 2009), and by survival of small diatoms with a low Fe demand (Morel et al., 1991; Sunda, 1997).  3.3.2.4 Non-reductive Fe uptake by indigenous plankton communities along Line P Field research in the subarctic Pacific along the Line P transect was vital to our study as it allowed us to explore the use of NRU by in situ phytoplankton. It was hypothesized that the rates of Fe uptake by the 55FeDFB and 55Fe14DFB treatments would be equal and relatively higher to those of the 55Fe14DFB + Ga14DFB treatment (see section 3.2.9). Conversely, rates of DFB uptake were expected to be equal for the 55Fe14DFB and 55Fe14DFB + Ga14DFB treatments, as Ga(III) would not hinder uptake of DFB. It was anticipated that all uptake rates of 55Fe would exceed those of 14DFB, as no inhibitor had been used to impede the reductive Fe uptake mechanism. At station P4, C-normalized uptake rates of Fe and DFB ("Fe and "DFB) by the 55FeDFB and 55Fe14DFB treatments were equal at all sampling intervals (8, 16, and 24 h), demonstrating the linearity of these uptake rates from 0 to 24 h (Table 3.6). As a result, all further comparisons among stations focus on uptake rates obtained at 8 h. First, rates of Fe uptake are discussed, as these results act as a control to ensure that the short-term uptake experiment methodology utilized in this investigation was successful. Overall, the Fe uptake rates observed in this investigation were similar to those measured previously along Line P using FeDFB (Maldonado and Price, 1999). Iron uptake rates (8 h) by the 55FeDFB and 55Fe14DFB treatments were often similar at each station. Only two comparisons were significantly different, both within the 1-5 µm size fraction at P20 and P26 (Table 3.7). More specifically, the "Fe of the 55Fe14DFB treatment was  85 Table 3.6: Short-term (8, 16, and 24 h) C-normalized Fe (!10-6 mol Fe (mol C)-1 d-1) and DFB (!10-3 mol DFB (mol C)-1 d-1) uptake rates by size- fractionated phytoplankton at station P4 along Line P in the subarctic Pacific Ocean. Seawater was obtained before sunrise from 10 m depth. During the precomplexation step for each treatment, 1 nmol L-1 55Fe and 4 nmol L-1 Fe were complexed with 10 DFB nmol L-1; 10 nmol L-1 14DFB was also added to the experimental treatments. Additionally, it was estimated that ~ 5 nmol L-1 Fe contamination was present in the 14DFB solution, maintaining a Fe : DFB molar ratio of 1 : 2 for each treatment. Gallium was added in a 1 : 1 molar ratio to Fe where applicable. Rates are ± standard error for triplicate measurements. Significantly different (p < 0.05) uptake rates within each size-fraction for each sampling event are marked with different letters (a,b); rates without letters were not significantly different.   !Fe ("10-6 mol Fe (mol C)-1 d-1) !DFB ("10-3 mol DFB (mol C)-1 d-1) Station Treatment 0.22-1 µm  1-5 µm  > 5 µm 0.22-1 µm 1-5 µm  > 5 µm P4 – 8 h 55FeDFB (control) 0.54 ± 0.010 b  0.13 ± 0.0074 b   0.17 ± 0.0068 b  –    –    –  55Fe14DFB 0.64 ± 0.42 b  0.47 ± 0.096 b   0.23 ± 0.014 b 2.7 ± 0.66  0.14 ± 0.025  0.13 ± 0.062  55Fe14DFB + Ga14DFB 75 ± 24 a  18 ± 1.8 a  8.0 ± 0.17 a 3.1 ± 1.4  0.097 ± 0.020  0.10 ± 0.017 P4 – 16 h 55FeDFB (control) 0.68 ± 0.059 b  0.21 ± 0.043 b  0.19 ± 0.0084 b  –    –    –  55Fe14DFB 0.90 ± 0.34 b  0.49 ± 0.075 b  0.46 ± 0.11 b 2.8 ± 0.38  0.14 ± 0.013  0.16 ± 0.052  55Fe14DFB + Ga14DFB 86 ± 14 a  16 ± 3.3 a  10 ± 1.8 a 2.6 ± 0.42  0.13 ± 0.021  0.18 ± 0.033 P4 – 24 h 55FeDFB (control) 0.67 ± 0.091 b  0.18 ± 0.012 b  0.18 ± 0.015 b  –    –    –  55Fe14DFB 1.2 ± 0.33 b  0.44 ± 0.013 b  0.35 ± 0.069 b 2.8 ± 0.18 b  0.14 ± 0.0071 b  0.11 ± 0.017  55Fe14DFB + Ga14DFB 78 ± 11 a  11 ± 0.88 a  6.6 ± 1.1 a 11 ± 2.6 a  0.41 ± 0.043 a  0.38 ± 0.17     86 typically less than 2 times greater than that of the 55FeDFB, and did not exceed 4 times. This indicates that the estimate of putative Fe contamination of 14DFB (5 nmol L-1) was somewhat inaccurate. Given that the average difference in !Fe between the 55FeDFB and 55Fe14DFB treatments was 2 times, this indicates that the putative Fe contamination in solution (see section 3.2.6) may have been closer to 2.5 nmol L-1. Due to the similarity between the rates of Fe uptake in the 55FeDFB and 55Fe14DFB treatments, only the 55Fe14DFB treatment is compared with the 55Fe14DFB + Ga14DFB treatment in the discussion below. At all sampling intervals, the 55Fe14DFB + Ga14DFB treatment exhibited !Fe that were consistently faster (by 1.1-117 fold) than those in the 55Fe14DFB treatment; one-way ANOVA analysis often emphasized this result (Table 3.7). As Ga(III) cannot be reduced but can displace Fe(III) from siderophores (Emery and Hoffer, 1980), this treatment was expected to yield lower !Fe relative to those in the 55Fe14DFB treatment due to competitive inhibition. However, the observed results may indicate that the treatment with the Ga(III) addition resulted in lower unchelated DFB concentrations (5 nmol L-1 vs. 10 nmol L-1, Table 3.3), and thus less desferrated DFB was available to compete for free Fe(III) with cell surface transporters associated with the reductive Fe transport system. As a result, slightly faster !Fe would be observed in our Ga addition treatment.  However, we believe that this effect would account for the small differences between the 55Fe14DFB + Ga14DFB and the 55Fe14DFB treatments in the open ocean stations (up to 1.1-5 fold), but not for the large ones observed in P4 (35-117 fold).  We are unable to explain the fast rates of !Fe in the presence of Ga at P4.  Future experiments should guarantee identical excess free DFB concentrations in the treatments with and without Ga. The decreasing differential in !Fe between the 55Fe14DFB + Ga14DFB and the 55Fe14DFB treatments from station P4 to P26 (Table 3.7) generally agrees with the increasing Fe limited  87 Table 3.7: Short-term (8-h) C-normalized Fe (!10-6 mol Fe (mol C) -1 d-1) and DFB (!10-3 mol DFB (mol C) -1 d-1) uptake rates by size-fractionated phytoplankton at stations P4, P16, P20, and P26 along Line P in the subarctic Pacific Ocean. Seawater was obtained before sunrise from 10 m depth at each station. During the precomplexation step for each treatment, 1 nmol L-1 55Fe and 4 nmol L-1 Fe were complexed with 10 DFB nmol L-1; 10 nmol L-1 14DFB was also added to the experimental treatments. Additionally, it was estimated that ~ 5 nmol L-1 Fe contamination was present in the 14DFB solution, maintaining a Fe : DFB molar ratio of 1 : 2 for each treatment. Gallium was added in a 1 : 1 molar ratio to Fe where applicable. Rates are ± standard error for triplicate measurements. Significantly different (p < 0.05) uptake rates within each size-fraction for each sampling event are marked with different letters (a,b,c); rates without letters were not significantly different.   !Fe ("10-6 mol Fe (mol C)-1 d-1) !DFB ("10-3 mol DFB (mol C)-1 d-1) Station Treatment 0.22-1 µm  1-5 µm  > 5 µm 0.22-1 µm 1-5 µm  > 5 µm P4 – 8 h 55FeDFB (control) 0.54 ± 0.010 b  0.13 ± 0.0074 b   0.17 ± 0.0068 b  –    –    –  55Fe14DFB 0.64 ± 0.42 b  0.47 ± 0.096 b   0.23 ± 0.014 b 2.7 ± 0.66  0.14 ± 0.025  0.13 ± 0.062  55Fe14DFB + Ga14DFB 75 ± 24 a  18 ± 1.8 a  8.0 ± 0.17 a 3.1 ± 1.4  0.097 ± 0.020  0.10 ± 0.017 P16 – 8 h 55FeDFB (control) 2.3 ± 0.087 b  0.50 ± 0.055 b  0.31 ± 0.024  –    –    –  55Fe14DFB 8.3 ± 2.9 b  0.62 ± 0.075 b  0.15 4.5 ± 0.98  0.28 ± 0.073  0.48 ± 0.0082  55Fe14DFB + Ga14DFB 32 ± 1.7 a  1.4 ± 0.14 a  0.58 ± 0.25 9.8 ± 2.8  0.31 ± 0.072  0.52 ± 0.023 P20 – 8 h 55FeDFB (control) 2.0 ± 0.17 b  0.34 ± 0.0055 c  0.33 ± 0.022  –    –    –  55Fe14DFB 3.0 ± 0.87 b  0.87 ± 0.11 b  0.42 ± 0.17 5.0 ± 0.44  0.081 ± 0.0046  0.36 ± 0.065  55Fe14DFB + Ga14DFB 16 ± 1.4 a  1.4 ± 0.061 a  0.49 ± 0.055 5.0 ± 0.34  0.079 ± 0.0071  0.36 ± 0.075 P26 – 8 h 55FeDFB (control) 9.8 ± 0.52 b  0.87 ± 0.064 b  0.60 ± 0.098  –    –    –  55Fe14DFB 13 ± 2.2 ab  1.9 ± 0.073 a  0.50 ± 0.091 11 ± 1.5  0.25 ± 0.010  0.53 ± 0.10  55Fe14DFB + Ga14DFB 21 ± 3.1 a  2.2 ± 0.20 a  0.87 ± 0.16 15 ± 1.3  0.27 ± 0.11  0.64 ± 0.091     88 status of plankton communities along Line P from east to west. Conversely, the differential in !Fe between 55Fe14DFB + Ga14DFB and 55Fe14DFB was less pronounced for cells exceeding 5 µm (Table 3.7). In particular, the rates of !Fe for 55FeDFB, 55Fe14DFB and 55Fe14DFB + Ga14DFB were indistinguishable for plankton > 5 µm at stations P16, P20, and P26. These combined results, as supported by comparable !Fe for the treatments with and without Ga, may suggest the importance of the NRU mechanism for the most Fe limited phytoplankton, found in the most western stations (the most Fe deficient waters), and the larger size fractions (resulting from low surface area to volume ratios). Internalization of DFB by the in situ plankton communities along Line P was also observed, supporting the potential for this non-reductive Fe uptake mechanism (Table 3.7). Furthermore, !DFB by the 55Fe14DFB and 55Fe14DFB + Ga14DFB treatments were typically equal; this result increased our confidence in the success of the DFB-labeling method as it further suggests that uptake of DFB was mediated by the NRU mechanism, and was therefore unaffected by the Ga addition. Despite these favourable results, !DFB were orders of magnitude faster than !Fe, in contrast to the anticipated slower rates. It is possible that this result was due to internalization of excess, desferrated 14DFB to be used as an organic C and/or N source. In support of this, the fastest !DFB were observed for the 0.22-1 µm size fraction, where heterotrophic bacteria are abundant. In addition, methylation of DFB may have produced a photolabile complex, promoting its UV oxidation while incubating on deck and ultimately releasing 14C to be taken up by the plankton community. Alternatively, the methylation of DFB may have resulted in a lipophilic compound that is easily transported across the cell membrane. However, this latter explanation is not supported by the culture experiments, where similar rates (1.4 fold difference) of 55Fe and 14DFB uptake were measured in the presence of 55Fe14DFB.  The  89 Table 3.8: Summary of statistically significant differences (p < 0.05) between stations’ size-fractionated C- normalized 8-h !Fe for each treatment, determined by one-way ANOVA. Statistical differences are denoted by different letters (A,B,C).  !Fe 55FeDFB (control) 55Fe14DFB 55Fe14DFB + Ga14DFB Station 0.22-1 µm 1-5 µm > 5 µm 0.22-1 µm 1-5 µm > 5 µm 0.22-1 µm 1-5 µm > 5 µm P4 B C B B B A A A A P16 B B B AB B A AB B B P20 B BC B B B A B B B P26 A A A A A A AB B B  fast rates of 14DFB uptake in the field may highlight the complex and fast biochemical cycling of small organic C compounds in the sea.  3.3.2.5 Trends in non-reductive Fe uptake along Line P The Fe-limited station P26 was generally characterized by the fastest (p < 0.05) !Fe for all plankton in each of the 55FeDFB and 55Fe14DFB treatments (Table 3.8). Additionally, !DFB was generally fastest at station P26, though differences between stations were less pronounced (Table 3.9). These trends are likely due to the Fe deficient nutritional status of the plankton community at this station, as rates of uptake of Fe bound to DFB are faster for severely Fe-limited than mildly Fe-stressed cells (Maldonado and Price, 1999). In addition, under chronic Fe limitation, plankton optimize their physiology to obtain Fe via less energy efficient mechanisms (Maldonado and Price, 1999; 2000; Lesuisse et al., 2001; Maldonado and Price, 2001; Maldonado et al., 2006); therefore, this result may support the potential for NRU in Fe-limited phytoplankton.   90 Table 3.9: Summary of statistically significant differences (p < 0.05) between stations’ size-fractionated C- normalized 8-h !DFB for each treatment, determined by one-way ANOVA. Statistical differences are denoted by different letters (A,B,C).        3.3.3 Oceanographic implications and future direction The results reported here may represent the first evidence of NRU by phytoplankton. Indeed, the development of novel methods and field investigation were met with many challenges: despite the internalization of 14DFB observed by phytoplankton cultures and in situ communities along Line P, there exists a division of evidence for and against the presence of NRU in phytoplankton. The results of our experimental treatments along Line P suggest that although the 14DFB compound was not purified or desferrated before the cruise, the pH manipulation of our complex prior to inoculating field treatments was successful and allowed the proper dual labeling with 55Fe. Indeed, the control (55FeDFB) and 55Fe14DFB treatments facilitated similar C-normalized !Fe. Unfortunately, absolute rates of DFB uptake exceeded those of Fe, implying error in our calculations or unknown biological and chemical reactivities of the 14DFB compound in surface waters. In general, we were able to explain some patterns of variations in !Fe and !DFB across treatments, size fractions, or stations by environmental characteristics at each station.  However, many unanswered questions still remain. !DFB 55Fe14DFB 55Fe14DFB + Ga14DFB Station 0.22-1 µm 1-5 µm > 5 µm 0.22-1 µm 1-5 µm > 5 µm P4 B AB B B A C P16 B A A AB A AB P20 B B AB B A BC P26 A AB A A A A  91 Certainly, many aspects of this investigation can be improved and honed, and we urge oceanographers to continue this research. A number of recommendations in the pursuit of evidence of NRU by phytoplankton can be made. First, labeling DFB with non-radioactive methyl iodide would allow the characterization of a cold analog of 14DFB. These results could then be used to characterize the radioactive compound 14DFB and validate the success of the 14C- labeling method. For example, one could utilize electrochemical methods to determine the stability constant of DFB for Fe in this modified complex, as well as mass spectrometry to verify its chemical structure. Second, it would be ideal to distinguish between the two non-reductive Fe uptake mechanisms, ligand exchange and direct siderophore internalization.  To rule out ligand exchange, chromium, Cr(III), which is chemically inert but structurally similar to Fe(III), could be used. When complexed with siderophores, Cr(III) is inert to ligand substitution and cannot be taken up via ligand exchange. Third, genetic evidence for a non-reductive Fe uptake pathway by phytoplankton could be sought out. As it has already been determined for S. cerevisiae (Lesuisse et al., 2001), identification of any gene in the SIT (Siderophore Iron Transport) family would confirm the ability of phytoplankton to acquire Fe via NRU. The development of an Expressed Sequence Tags (EST) database for T. oceanica under both Fe-limiting and Fe-sufficient conditions will be invaluable in this regard (Maheswari et al., 2005).   92 Chapter 4: Conclusion The research presented in this dissertation addresses two relevant questions in the realm of Fe bioavailability to phytoplankton. First, the role of Fe in co-limiting the growth of plankton in the Beaufort Sea in conjunction with light and NO3- in late summer was tested. It was discovered that the plankton community at an open ocean station in the Beaufort Sea was NO3- limited during our study in September 2009. However, after a NO3- addition, the cells were shown to be co-limited by Fe and light at light levels " 10 % surface irradiance. Our experimental results are likely relevant at certain times of the year when decoupling of seasonal biogeochemical cycles of these resources may be observed. Second, a novel physiological investigation provided preliminary evidence that the model diatom T. oceanica, as well as in situ plankton along Line P, acquire Fe bound to siderophores via a non-reductive uptake mechanism. This ferrated whole-siderophore transport mechanism, like the well-characterized reductive Fe uptake mechanism in T. oceanica (Maldonado and Price, 2000; 2001; Maldonado et al., 2006) is also found in baker’s yeast S. cerevisiae (Lesuisse et al., 1998; 2001). Together, these investigations also contribute to our broader understanding of phytoplankton Fe nutrition and acquisition. Looking forward, the research presented here will contribute to the growing and all- important field of modeling changes in global primary productivity and carbon export in the face of global warming.  4.1 Fe and Arctic primary productivity The seasonality of NO3- and light in the high north is well documented, as are the effects of these resources on primary productivity (Sherr et al., 2003; Hill et al., 2005). In contrast, sources of Fe to the Beaufort Sea are not well characterized, nor is the seasonal supply rate of  93 this micronutrient. Furthermore, despite the important stake of Fe in the physiological processes of nitrogen assimilation and photosynthesis (Raven, 1988), an interactive effect of Fe with NO3- and/or light on primary productivity had not been investigated in the Arctic Ocean. This dissertation addresses the shortcomings of previous research on Fe in the Beaufort Sea. Intrigued by the small supply of existing Arctic Fe data, we investigated the potential for Fe limitation of plankton growth in late summer/early fall in the Beaufort Sea to shed light on the seasonality of Fe as a limiting resource in this region. Indeed, the unique seasonality of Fe sources in the Canada Basin, predominantly controlled by changes in sea ice (Measures, 1999; Isaksson et al., 2003; Krachler et al., 2005), is hypothesized to create a dynamic interaction among light, NO3- and Fe as a determinant of primary productivity. In early spring, NO3-, light and Fe facilitate growth of in situ plankton communities in tandem. However, spring drawdown of NO3- to limiting levels decouples this resource from light and Fe, becoming the first stressor by limiting plankton growth in early summer. This effect is later exacerbated by light which decreases with the onset of the late summer/early fall, and increases the phytoplankton demand for Fe. Regardless of the presence of sea ice, decreasing light availability eventually forces the decoupling of high Fe demand from Fe supply, inhibiting growth of in situ plankton. Understanding the seasonality of Fe, light, and NO3- availability in the Arctic Ocean, as well as their interaction will be crucial to predicting future changes in Arctic primary productivity.  4.2 Fe and phytoplankton survival The driving force behind elucidating Fe acquisition by phytoplankton originated when Rue and Bruland’s (1997) field study in the equatorial Pacific Ocean demonstrated that the ubiquitous presence of putative siderophore-like organic complexes in seawater regulates an  94 equilibrium concentration of inorganic Fe in surface waters (~ 10 fmol L-1) well below that which limits diffusive supply of inorganic Fe to the cell surface (4 pmol L-1). Research inspired by this puzzling result hypothesized that phytoplankton utilize a reductive Fe uptake mechanism when Fe limited and examined this possibility in the model diatoms T. oceanica and T. pseudonana, via a series of short-term physiological experiments (Maldonado and Price, 2001; Maldonado et al., 2006). These tests elucidated a reductive mechanism for Fe acquisition from dissolved organic complexes, which was similar to that of other microbes (Maldonado and Price, 2001; Maldonado et al., 2006). However, this work left one outstanding anomaly: uptake of Fe from organically bound complexes (FeDFB) by the oceanic model diatom T. oceanica was two times faster than that by the coastal species T. pseudonana, despite having virtually identical oxidation and reduction rates in the reductive Fe uptake mechanism (Maldonado et al., 2006). The results presented in this dissertation partly address this disparity. Though uptake of 14DFB was observed in cultures and in plankton communities along Line P in the subarctic Pacific Ocean, the results of these experiments are not conclusive due to the inability to a) verify its concentration, b) purify the 14DFB compound, and c) to desferrate the Fe14DFB complex. Despite this, the development of novel methods and suggestions for improvement are valuable to the future of this investigation. The verification of the 14C-labeling DFB method and the purification of the radiolabeled DFB are essential to obtain conclusive evidence of this NRU mechanism in phytoplankton. Interestingly, genetic evidence for non-reductive Fe uptake mechanisms may already exist. A study investigating the response of diatom Phaeodactylum tricornutum to Fe deficiency revealed that the encoding gene for a protein of unknown function, ISIP1, was the most highly expressed under Fe deficiency (Allen et al., 2008). An orthologous gene has been described in Dunaliella salina, a green algae, that encodes a protein believed to  95 enhance binding and uptake of ferric ions through the plasma membrane. Furthermore, this ISIP1 protein is absent in T. pseudonana, supporting the results of the short-term physiological experiments with T. oceanica and T. pseudonana that inspired this investigation (Maldonado et al., 2006). Determining the uptake mechanism that is encoded by this gene may prove to be an essential part of elucidating a non-reductive Fe uptake mechanism by phytoplankton.  4.3 Fe and global warming As our oceans respond to climate change, it is imperative to understand the dynamic relationship between phytoplankton and Fe to predict the fate of phytoplankton communities and the biological carbon pump. Indeed, marine primary productivity and subsequent CO2 draw down to the deep ocean play an integral part in the global C cycle (Falkowski et al., 1998; Field et al., 1998) as biologically mediated export of organic carbon to the deep ocean represents a long-term sink for atmospheric CO2 (Karl et al., 1997). As a cofactor in many metabolic pathways in plankton (Raven, 1988), Fe has a crucial stake in the global C cycle due to its direct influence on nutrition and proliferation of these microorganisms. Many models have been developed to assess the fate of phytoplankton and primary productivity in a warming world (Lancelot et al., 2000; Moore et al., 2002; Bopp et al., 2005; Sarthou et al., 2005; Alvain et al., 2008; Galbraith et al., 2010). The strength of climate models can only increase as controls of regional primary productivity continue to be elucidated. The work presented in this dissertation contributes to this endeavor by demonstrating the potential importance of Fe in controlling primary productivity in the Beaufort Sea, and identifying a potential mechanism by which phytoplankton may access this essential micronutrient in the subarctic Pacific Ocean.  96  Broadly, climate change is already affecting the expansion of oligotrophic gyres and strengthening of stratification, thereby limiting mixing and upwelling of nutrients (Polovina et al., 2008). Furthermore, enhanced precipitation (Dai et al., 1997) may reduce the aeolian flux of Fe, thereby stressing these physiological processes that depend on this micronutrient. As a result, the ability to acquire Fe via a myriad of uptake mechanisms will be the best survival strategy of phytoplankton species. It is imperative that continued oceanographic research elucidates the physiological capabilities of Fe-stressed phytoplankton to acquire Fe. In the high north, reduced sea ice (Rothrock et al., 2003; Belchansky et al., 2005) will facilitate a positive feedback system by which increased absorption of solar radiation by seawater will increase temperatures of the surface waters (Perovich et al., 2007) and atmosphere (Morales-Maqueda et al., 1999), promoting further ice melt and thus stratification. This cycle will considerably reduce the most important supply of Fe to surface waters, sea ice. The decreased supply of Fe will be exacerbated by decreased bioavailability of Fe from organic complexes due to ocean acidification (Shi et al., 2010). Furthermore, expanding oxygen minimum zones (Paulmier and Ruiz-Pino, 2009) will increase both dissolved Fe concentrations and denitrification in the source waters of the Chukchi and Beaufort Seas. This will ultimately result in less NO3-, but changes in Fe inventories in the source waters to the Arctic are unpredictable as the fate of elevated Fe concentrations that reach surface waters is unknown. The changing availabilities of these essential resources will have a considerable effect on Arctic primary productivity, thus continued research and monitoring is necessary to better predict, and hopefully circumvent, these changes.   97 Bibliography  Allen, A., LaRoche, J., Maheswari, U., Lommer, M., Schauer, N., Lopez, P., Finazzi, G., Fernie, A., Bowler, C., 2008. Whole-cell response of the pennate diatom Phaeodactylum tricornutum to iron starvation. P Natl Acad Sci USA 105, 10438–10443.  Alvain, S., Moulin, C., Dandonneau, Y., Loisel, H., 2008. Seasonal distribution and succession of dominant phytoplankton groups in the global ocean: A satellite view. Global Biogeochem Cy 22, GB3001.  Anderson, M., Morel, F., 1982. The influence of aqueous iron chemistry on the uptake of iron by the coastal diatom Thalassiosira weissflogii. Limnol Oceanogr 27, 789–813.  Armbrust, E., 2009. The life of diatoms in the world's oceans. Nature 459, 185–192.  Barbeau, K., Rue, E., Bruland, K., Butler, A., 2001. Photochemical cycling of iron in the surface ocean mediated by microbial iron(III)-binding ligands. Nature 413, 409–413.  Barwell-Clarke, J., Whitney, F., 1996. Institute of Ocean Sciences Nutrient Method and Analysis ( No. 182), Canadian Technical Report of Hydrography and Ocean Science. Canadian Technical Report of Hydrography and Ocean Sciences.  Belchansky, G., Douglas, D., Platonov, N., 2005. Spatial and temporal variations in the age structure of Arctic sea ice. Geophys Res Lett 32, L18504.  Bickel, H., Fechtig, B., Hall, G., Keller-Schierlein, W., 1960. Metabolic products of actinomycetes. XXIV. Isolation and synthesis of 1-amino-5-(hydroxylamino) pentane, an essential hydrolysis product of ferrioxamine and ferrimycin. Helv Chim Acta 43, 901–904.  Booth, B., 1988. Size classes and major taxonomic groups of phytoplankton at two locations in the subarctic Pacific Ocean in May and August, 1984. Mar Biol 97, 275–286.  Booth, B., Lewin, J., Postel, J., 1993. Temporal variation in the structure of autotrophic and heterotrophic communities in the subarctic Pacific. Prog Oceanogr 32, 57–99.  Bopp, L., Aumont, O., Cadule, P., Alvain, S., Gehlen, M., 2005. Response of diatoms distribution to global warming and potential implications: A global model study. Geophys Res Lett 32, L19606.  Boyd, P., Whitney, F., Harrison, P., Wong, C., 1995. The NE subarctic Pacific in winter: II. Biological rate processes. Mar Ecol-Prog Ser 128, 25–34.  Brand, L., 1991. Minimum iron requirements of marine-phytoplankton and the implications for the biogeochemical control of new production. Limnol Oceanogr 36, 1756–1771.   98 Brand, L., Guillard, R., Murphy, L., 1981. A method for the rapid and precise determination of acclimated phytoplankton reproduction rates. J Plankton Res 3, 193–201.  Brand, L., Sunda, W., Guillard, R., 1983. Limitation of marine phytoplankton reproductive rates by zinc, manganese, and iron. Limnol Oceanogr 28, 1182–1198.  Buesseler, K., Ball, L., Andrews, J., Benitez-Nelson, C., Belastock, R., Chai, F., Chao, Y., 1998. Upper ocean export of particulate organic carbon in the Arabian Sea derived from thorium- 234. Deep-Sea Res Pt II 45, 2461–2487.  Carmack, E., Macdonald, R., 2002. Oceanography of the Canadian Shelf of the Beaufort Sea: A setting for marine life. Arctic 55, 29–45.  Caron, D., Dam, H., Kremer, P., Lessard, E., Madin, L., Malone, T., Napp, J., Peele, E., Roman, M., Youngbluth, M., 1995. The contribution of microorganisms to particulate carbon and nitrogen in surface waters of the Sargasso Sea near Bermuda. Deep-Sea Res Pt I 42, 943– 972.  Carson, K., Holliday, S., Glenn, A., Dilworth, M., 1992. Siderophore and organic acid production in root nodule bacteria. Arch Microbiol 157, 264–271.  Cavender-Bares, K., Mann, E., Chisholm, S., Ondrusek, M., Bidigare, R., 1999. Differential response of equatorial Pacific phytoplankton to iron fertilization. Limnol Oceanogr 44, 237– 246.  Cárdenas, J., Rivas, J., Paneque, A., Losada, M., 1972. Effect of iron supply on the activities of the nitrate-reducing system from Chlorella. Arch Microbiol 81, 260–263.  Coale, K., Johnson, K., Fitzwater, S., Gordon, R., Tanner, S., Chavez, F., Ferioli, L., Sakamoto, C., Rogers, P., Millero, F., Steinberg, P., Nightingale, P., Cooper, D., Cochlan, W., Landry, M., Constantinou, J., Rollwagen, G., Trasvina, A., Kudela, R., 1996. A massive phytoplankton bloom induced by an ecosystem-scale iron fertilization experiment in the equatorial Pacific Ocean. Nature 383, 495–501.  Cooper, L., 1935. Iron in the sea and in marine plankton. P Roy Soc Lond B Bio 118, 419–438.  Dai, A., Fung, I., Del Genio, A., 1997. Surface observed global land precipitation variations during 1900–88. J Climate 10, 2943–2962.  Dancis, A., Haile, D., Yuan, D., Klausner, R., 1994. The Saccharomyces cerevisiae copper transport protein (Ctr1p): Biochemical characterization, regulation by copper, and physiologic role in copper uptake. J Biol Chem 269, 25660–25667.  de Baar, H., Boyd, P., Coale, K., Landry, M., Tsuda, A., Assmy, P., Bakker, D., Bozec, Y., Barber, R., Brzezinski, M., Buesseler, K., Boyé, M., Croot, P., Gervais, F., Gorbunov, M., Harrison, P., Hiscock, W., Laan, P., Lancelot, C., Law, C., Levasseur, M., Marchetti, A.,  99 Millero, F., Nishioka, J., Nojiri, Y., van Oijen, T., Riebesell, U., Rijkenberg, M., Saito, H., Takeda, S., Timmermans, K., Veldhuis, M., Waite, A., Wong, C., 2005. Synthesis of iron fertilization experiments: From the iron age in the age of enlightenment. J Geophys Res- Oceans 110, C09S16.  de Baar, H., Boyd, P., 2000. The role of iron in plankton ecology and carbon dioxide transfer of the global oceans, in: The Changing Ocean Carbon Cycle: A Midterm Synthesis of the Joint Global. University Press, Cambridge, UK, pp. 61–140.  De Jong, J., Das, Den, J., Bathmann, U., Stoll, M., Kattner, G., Nolting, R., de Baar, H., 1998. Dissolved iron at subnanomolar levels in the Southern Ocean as determined by ship-board analysis. Analytica Chimica Acta 377, 112–124.  Duce, R., 1986. The impact of atmospheric nitrogen, phosphorus, and iron species on marine biological productivity, in: The Role of Air-Sea Exchange in Geochemical Cycling. Springer, Hingham, Mass., pp. 497–529.  Duce, R., Tindale, N., 1991. Atmospheric transport of iron and its deposition in the ocean. Limnol Oceanogr 36, 1715–1726.  Emery, T., 1986. Exchange of iron by gallium in siderophores. Biochemistry 25, 4629–4633.  Emery, T., Hoffer, P., 1980. Siderophore-mediated mechanism of gallium uptake demonstrated in the microorganism Ustilago sphaerogena. J Nucl Med 21, 935–939.  Falkowski, P., Barber, R., Smetacek, V., 1998. Biogeochemical controls and feedbacks on ocean primary production. Science 281, 200–206.  Falkowski, P., LaRoche, J., 1991. Molecular biology in studies of ocean processes. Int Rev Cytol, International Review of Cytology 128, 261–303.  Field, C., Behrenfeld, M., Randerson, J., Falkowski, P., 1998. Primary production of the biosphere: Integrating terrestrial and oceanic components. Science 281, 237–240.  Forest, A., Sampei, M., Hattori, H., Makabe, R., Sasaki, H., Fukuchi, M., Wassmann, P., Fortier, L., 2007. Particulate organic carbon fluxes on the slope of the Mackenzie Shelf (Beaufort Sea): Physical and biological forcing of shelf-basin exchanges. J Marine Syst 68, 39–54.  Fujiki, T., Matsumoto, K., Honda, M., Kawakami, H., Watanabe, S., 2009. Phytoplankton composition in the subarctic North Pacific during autumn 2005. J Plankton Res 31, 179–191.  Galbraith, E., Gnanadesikan, A., Dunne, J., Hiscock, M., 2010. Regional impacts of iron-light colimitation in a global biogeochemical model. Biogeosciences 7, 1043–1064.  Geider, R., La Roche, J., 1994. The role of iron in phytoplankton photosynthesis, and the potential for iron-limitation of primary productivity in the sea. Photosynth Res 39, 275–301.  100 Giesbrecht, T., Sutherland, N., Johnson, W., Cullen, J., 2010. The distribution of dissolved iron and aluminum in the Beaufort Sea. Goldschmidt Conference Abstracts.  Glibert, P., Biggs, D., McCarthy, J., 1982. Utilization of ammonium and nitrate during austral summer in the Scotia Sea. Deep-Sea Res 29, 837–850.  Glover, H., 1977. Effects of iron deficiency on Isochrysis galbana (Chrysophyceae) and Phaeodactylum tricornutum (Bacillariophyceae). J Phycol 13, 208–212.  Goldberg, E., 1952. Iron assimilation by marine diatoms. Biol Bull 102, 243–248.  Gonzalez, J., Sherr, B., Sherr, E., 1993. Digestive enzyme activity as a quantitative measure of protistan grazing: the acid lysozyme assay for bacterivory. Mar Ecol-Prog Ser 100, 197–206.  Gran, H., 1931. On the conditions for the production of plankton in the sea. Rapp Proc-verb Reun Cons Int Explor Mer 75, 37–46.  Gran, H., 1933. Studies on the biology and chemistry of the Gulf of Maine II Distribution of phytoplankton in August 1932. Biol Bull 64, 159–182.  Greene, R., Geider, R., Falkowski, P., 1991. Effect of iron limitation on photosynthesis in a marine diatom. Limnol Oceanogr 36, 1772–1782.  Guerinot, M., Yi, Y., 1994. Iron: nutritious, noxious, and not readily available. Plant Physiol 104, 815–820.  Guéguen, C., Guo, L., Yamamoto-Kawai, M., Tanaka, N., 2007. Colored dissolved organic matter dynamics across the shelf-basin interfaces in the western Arctic Ocean. J Geophys Res-Oceans 112, C05038.  Harrison, G., Morel, F., 1986. Response of the marine diatom Thalassiosira weissflogii to iron stress. Limnol Oceanogr 31, 989–997.  Hart, T., 1934. On the phytoplankton of the south-west Atlantic and the Bellingshausen Sea, 1929-31, Discovery Reports Vol VIII. University Press.  Harvey, H., Cooper, L., Lebour, M., Russell, F., 1935. Plankton production and its control. J Mar Biol Ass 20, 407–441.  Hashihama, F., Horimoto, N., Kanda, J., Furuya, K., Ishimaru, T., Saino, T., 2008. Temporal variation in phytoplankton composition related to water mass properties in the central part of Sagami Bay. J Oceanogr 64, 23–37.  Henley, W., 1992. Growth and photosynthesis of Ulva rotundata (Chlorophyta) as a function of temperature and square wave irradiance in indoor culture. J Phycol 28, 625–634.   101 Hill, V., Cota, G., Stockwell, D., 2005. Spring and summer phytoplankton communities in the Chukchi and Eastern Beaufort Seas. Deep-Sea Res Pt II 52, 3369–3385.  Hopkins, E., 1930. Iron-ion concentration in relation to growth and other biological processes. Bot Gaz 89, 209–240.  Hopkins, E., Wann, F., 1926. Relation of hydrogen-ion concentration to growth of Chlorella and to the availability of iron. Bot Gaz 81, 353–376.  Hopkins, E., Wann, F., 1927. Iron requirement for Chlorella. Bot Gaz 84, 407–427.  Hudak, D., Young, J., 2002. Storm climatology of the Southern Beaufort sea ( No. 2), Atmosphere-Ocean. Meteorological Service of Canada.  Hudson, R., Covault, D., Morel, F., 1992. Investigations of iron coordination and redox reactions in seawater using 59Fe radiometry and ion-pair solvent extraction of amphiphilic iron complexes. Mar Chem 38, 209–235.  Hudson, R., Morel, F., 1989. Distinguishing between extra- and intracellular iron in marine phytoplankton. Limnol Oceanogr 34, 1113–1120.  Isaksson, E., Hermanson, M., Hicks, S., Igarashi, M., Kamiyama, K., Moore, J., Motoyama, H., Muir, D., Pohjola, V., Vaikmäe, R., van de Wal, R., Watanabe, O., 2003. Ice cores from Svalbard-useful archives of past climate and pollution history. Phys Chem Earth 28, 1217– 1228.  Jickells, T., An, Z., Andersen, K., Baker, A., Bergametti, G., Brooks, N., Cao, J., Boyd, P., Duce, R., Hunter, K., Kawahata, H., Kubilay, N., LaRoche, J., Liss, P., Mahowald, N., Prospero, J., Ridgwell, A., Tegen, I., Torres, R., 2005. Global Iron Connections Between Desert Dust, Ocean Biogeochemistry, and Climate. Science 308, 67–71.  Johnson, K., Gordon, R., Coale, K., 1997. What controls dissolved iron concentrations in the world ocean? Mar Chem 57, 137–161.  Johnson, W., Miller, L., Sutherland, N., Wong, C., 2005. Iron transport by mesoscale Haida eddies in the Gulf of Alaska. Deep-Sea Res Pt II 52, 933–953.  Karl, D., Letelier, R., Tupas, L., Dore, J., Christian, J., Hebel, D., 1997. The role of nitrogen fixation in biogeochemical cycling in the subtropical North Pacific Ocean. Nature 388, 533– 538.  Ketchum, B., 1954. Mineral nutrition of phytoplankton. Ann Rev Plant Physio 5, 55–74.  Kim, T., Lee, Y., Kim, G., 2010. Hydrographically mediated patterns of photosynthetic pigments in the East/Japan Sea: Low N:P ratios and cyanobacterial dominance. J Marine Syst 82, 72– 79.  102 Kirchman, D., Keel, R., Simon, M., Welschmeyer, N., 1993. Biomass and production of heterotrophic bacterioplankton in the oceanic subarctic Pacific. Deep-Sea Res Pt I 40, 967– 988.  Koike, I., Holm-Hansen, O., Biggs, D., 1986. Inorganic nitrogen-metabolism by Antarctic phytoplankton with special reference to ammonium cycling. Mar Ecol-Prog Ser 30, 105– 116.  Krachler, M., Zheng, J., Fisher, D., Shotyk, W., 2005. Analytical procedures for improved trace element detection limits in polar ice from Arctic Canada using ICP-SMS. Analytica Chimica Acta 530, 291–298.  Lam, P., Bishop, J., Henning, C., Marcus, M., Waychunas, G., Fung, I., 2006. Wintertime phytoplankton bloom in the subarctic Pacific supported by continental margin iron. Global Biogeochem Cy 20.  Lam, P., Bishop, J., 2008. The continental margin is a key source of iron to the HNLC North Pacific Ocean. Geophys Res Lett 35, L07608.  Lancelot, C., Hannon, E., Becquevort, S., Veth, C., de Baar, H., 2000. Modeling phytoplankton blooms and carbon export production in the Southern Ocean: Dominant controls by light and iron in the Atlantic sector in Austral spring 1992. Deep-Sea Res Pt I 47, 1621–1662.  Landing, W., Bruland, K., 1987. The contrasting biogeochemistry of iron and manganese in the Pacific Ocean. Geochimica et Cosmochimica Acta 51, 29–43.  LaRoche, J., Geider, R., Graziano, L., Murray, H., Lewis, K., 1993. Induction of specific proteins in eukaryotic algae grown under iron-, phosphorus-, or nitrogen-deficient conditions. J Phycol 29, 767–777.  LaRoche, J., Murray, H., Orellana, M., Newton, J., 1995. Flavodoxin expression as an indicator of iron limitation in marine diatoms. J Phycol 31, 520–530.  LaRoche, J., Boyd, P., McKay, R., Geider, R., 1996. Flavodoxin as an in situ marker for iron stress in phytoplankton. Nature 382, 802–805.  Latasa, M., 2007. Improving estimations of phytoplankton class abundances using CHEMTAX. Mar Ecol-Prog Ser 329, 13–21.  Lee, S., Fuhrman, J., 1987. Relationships between biovolume and biomass of naturally derived marine bacterioplankton. Appl Environ Microb 53, 1298–1303.  Lesuisse, E., Simon-Casteras, M., Labbe, P., 1998. Siderophore-mediated iron uptake in Saccharomyces cerevisiae: the SIT1 gene encodes a ferrioxamine B permease that belongs to the major facilitator superfamily. Microbiol-Uk 144, 3455–3462.   103 Lesuisse, E., Blaiseau, P., Dancis, A., Camadro, J., 2001. Siderophore uptake and use by the yeast Saccharomyces cerevisiae. Microbiol-Sgm 147, 289–298.  Lesuisse, E., Labbe, P., 1989. Reductive and non-reductive mechanisms of iron assimilation by the yeast Saccharomyces cerevisiae. J Gen Microbiol 135, 257–263.  Levitus, S., 1982. Climatological atlas of the world ocean, NOAA Professional Paper 13. US Government Printing Office, Rockville, Md.  Lochte, K., Ducklow, H., Fasham, M., Stienen, C., 1993. Plankton succession and carbon cycling at 47°N 20°W during the JGOFS North Atlantic Bloom Experiment. Deep-Sea Res Pt II 40, 91–114.  Lovejoy, C., Vincent, W., Bonilla, S., Roy, S., Martineau, M., Terrado, R., Potvin, M., Massana, R., Pedrós-Alió, C., 2007. Distribution, phylogeny, and growth of cold-adapted picoprasinophytes in Arctic seas. J Phycol 43, 78–89.  Mackey, D., Higgins, H., Mackey, M., Holdsworth, D., 1998. Algal class abundances in the western equatorial Pacific: Estimation from HPLC measurements of chloroplast pigments using CHEMTAX. Deep-Sea Res Pt I 45, 1441–1468.  Mackey, M., Mackey, D., Higgins, H., Wright, S., 1996. CHEMTAX - a program for estimating class abundances from chemical markers: Application to HPLC measurements of phytoplankton. Mar Ecol-Prog Ser 144, 265–283.  Macrellis, H., Trick, C., Rue, E., Smith, G., Bruland, K., 2001. Collection and detection of natural iron-binding ligands from seawater. Mar Chem 76, 175–187.  Maheswari, U., Montsant, A., Goll, J., Krishnasamy, S., Rajyashri, K., Patell, V., Bowler, C., 2005. The diatom EST database. Nucleic Acids Res 33, D344–D347.  Maldonado, M., Boyd, P., Harrison, P., Price, N., 1999. Co-limitation of phytoplankton growth by light and Fe during winter in the NE subarctic Pacific Ocean. Deep-Sea Res Pt II 46, 2475–2485.  Maldonado, M., Boyd, P., LaRoche, J., Strzepek, R., Waite, A., Bowie, A., Croot, P., Frew, R., Price, N., 2001. Iron uptake and physiological response of phytoplankton during a mesoscale Southern Ocean iron enrichment. Limnol Oceanogr 46, 1802–1808.  Maldonado, M., Allen, A., Chong, J., Lin, K., Leus, D., Karpenko, N., Harris, S., 2006. Copper- dependent iron transport in coastal and oceanic diatoms. Limnol Oceanogr 51, 1729–1743.  Maldonado, M., Price, N., 1996. Influence of N substrate on Fe requirements of marine centric diatoms. Mar Ecol-Prog Ser 141, 161–172.    104 Maldonado, M., Price, N., 1999. Utilization of iron bound to strong organic ligands by plankton communities in the subarctic Pacific Ocean. Deep-Sea Res Pt II 46, 2447–2473.  Maldonado, M., Price, N., 2000. Nitrate regulation of Fe reduction and transport by Fe-limited Thalassiosira oceanica. Limnol Oceanogr 45, 814–826.  Maldonado, M., Price, N., 2001. Reduction and transport of organically bound iron by Thalassiosira oceanica (Bacillariophyceae). J Phycol 37, 298–309.  Marchetti, A., Maldonado, M., Lane, E., Harrison, P., 2006. Iron requirements of the pennate diatom Pseudo-nitzschia: Comparison of oceanic (high-nitrate, low-chlorophyll waters) and coastal species. Limnol Oceanogr 51, 2092–2101.  Marchetti, A., Parker, M., Moccia, L., Lin, E., Arrieta, A., Ribalet, F., Murphy, M., Maldonado, M., Armbrust, E., 2009. Ferritin is used for iron storage in bloom-forming marine pennate diatoms. Nature 457, 467–470.  Margalef, R., 1978. Life-forms of phytoplankton as survival alternatives in an unstable environment. Oceanol Acta 1, 493–509.  Marie, D., Brussaard, C., Thyrhaug, R., Bratbak, G., Vaulot, D., 1999. Enumeration of marine viruses in culture and natural samples by flow cytometry. Appl Environ Microb 65, 45–52.  Marie, D., Simon, N., Vaulot, D., 2005. Phytoplankton Cell Counting by Flow Cytometry, in: Algal culturing techniques. Elsevier Academic Press, Burlington, MA, pp. 253–268.  Martin, J., 1990. Glacial-interglacial CO2 change: The iron hypothesis. Paleoceanography 5, 1– 13.  Martin, J., 1992. Iron as a Limiting Factor in Oceanic Productivity, in: Primary Productivity and Biogeochemical Cycles in the Sea. 43, 123–137.  Martin, J., Gordon, R., Fitzwater, S., Broenkow, W., 1989. Vertex: phytoplankton/iron studies in the Gulf of Alaska. Deep-Sea Res 36, 649–680.  Martin, J., Fitzwater, S., Gordon, R., Hunter, C., Tanner, S., 1993. Iron, primary production and carbon-nitrogen flux studies during the JGOFS North Atlantic bloom experiment. Deep-Sea Res Pt II 40, 115–134.  Martin, J., Gordon, R., 1988. Northeast Pacific iron distributions in relation to phytoplankton productivity. Deep-Sea Res 35, 177–196.  Martínez-Martínez, J., Norland, S., Thingstad, T., Schroeder, D., Bratbak, G., Wilson, W., Larsen, A., 2006. Variability in microbial population dynamics between similarly perturbed mesocosms. J Plankton Res 28, 783–791.   105 Measures, C., 1999. The role of entrained sediments in sea ice in the distribution of aluminium and iron in the surface waters of the Arctic Ocean. Mar Chem 68, 59–70.  Meehl, G., Washington, W., Collins, W., Arblaster, J., Hu, A., Buja, L., Strand, W., Teng, H., 2005. How much more global warming and sea level rise? Science 307, 1769–1772.  Michaels, A., Silver, M., 1988. Primary production, sinking fluxes and the microbial food web. Deep-Sea Res 35, 473–490.  Miki, M., Ramaiah, N., Takeda, S., Furuya, K., 2008. Phytoplankton dynamics associated with the monsoon in the Sulu Sea as revealed by pigment signature. J Oceanogr 64, 663–673.  Moore, J., Doney, S., Glover, D., Fung, I., 2002. Iron cycling and nutrient-limitation patterns in surface waters of the World Ocean. Deep-Sea Res Pt II 49, 463–507.  Morales-Maqueda, M., Willmott, A., Darby, M., 1999. A numerical model for interdecadal variability of sea ice cover in the Greenland-Iceland-Norwegian Sea. Clim Dynam 15, 89– 113.  Morel, F., 2008. The co-evolution of phytoplankton and trace element cycles in the oceans. Geobiology 6, 318–324.  Morel, F., Hudson, R., Price, N., 1991. Limitation of productivity by trace metals in the sea. Limnol Oceanogr 36, 1742–1755.  Morel, F., Price, N., 2003. The biogeochemical cycles of trace metals in the oceans. Science 300, 944–947.  Nakayama, Y., Fujita, S., Kuma, K., Shimada, K., 2011. Iron and humic-type fluorescent dissolved organic matter in the Chukchi Sea and Canada Basin of the western Arctic Ocean. J Geophys Res-Oceans 116, C07031.  Neilands, J., 1981a. Iron-absorption and transport in microorganisms. Annu Rev Nutr 1, 27–46.  Neilands, J., 1981b. Microbial iron compounds. Annu Rev Biochem 50, 715–731.  Neilands, J., 1995. Siderophores: Structure and function of microbial iron transport compounds. J Biol Chem 270, 26723–26726.  Nelson, D., Tréguer, P., Brzezinski, M., Leynaert, A., Quéguiner, B., 1995. Production and dissolution of biogenic silica in the ocean: Revised global estimates, comparison with regional data and relationship to biogenic sedimentation. Global Biogeochem Cy 9, 359– 372.     106 Not, F., Massana, R., Latasa, M., Marie, D., Colson, C., Eikrem, W., Pedrós-Alió, C., Vaulot, D., Simon, N., 2005. Late summer community composition and abundance of photosynthetic picoeukaryotes in Norwegian and Barents Seas. Limnol Oceanogr 50, 1677–1686.  Obata, H., Karatani, H., Nakayama, E., 1993. Automated determination of iron in seawater by chelating resin concentration and chemiluminescence detection. Anal Chem 65, 1524–1528.  O'Brien, I., Gibson, F., 1970. The structure of enterochelin and related 2,3-dihydroxy-n- benzoylserine conjugates from Escherichia coli. Bba-Bioenergetics 215, 393–402.  Olson, R., 1980. Nitrate and ammonium uptake in Antarctic waters. Limnol Oceanogr 25, 1064– 1074.  Parekh, P., Follows, M., Boyle, E., 2004. Modeling the global ocean iron cycle. Global Biogeochem Cy 18, GB1002.  Paulmier, A., Ruiz-Pino, D., 2009. Oxygen minimum zones (OMZs) in the modern ocean. Prog Oceanogr 80, 113–128.  Peers, G., Price, N., 2006. Copper-containing plastocyanin used for electron transport by an oceanic diatom. Nature 441, 341–344.  Perovich, D., Light, B., Eicken, H., Jones, K., Runciman, K., Nghiem, S., 2007. Increasing solar heating of the Arctic Ocean and adjacent seas, 1979-2005: Attribution and role in the ice- albedo feedback. Geophys Res Lett 34, L19505.  Peterson, B., Holmes, R., McClelland, J., Vörösmarty, C., Lammers, R., Shiklomanov, A., Shiklomanov, I., Rahmstorf, S., 2002. Increasing river discharge to the Arctic Ocean. Science 298, 2171–2173.  Polovina, J., Howell, E., Abecassis, M., 2008. Ocean's least productive waters are expanding. Geophys Res Lett 35, L03618.  Price, N., Harrison, G., Hering, J., Hudson, R., Nirel, P., Palenik, B., Morel, F., 1989. Preparation and chemistry of the artificial algal culture medium Aquil. Biol Oceanogr 6, 443–461.  Price, N., Andersen, L., Morel, F., 1991. Iron and nitrogen nutrition of equatorial Pacific plankton. Deep-Sea Res 38, 1361–1378.  Prospero, J., 1996. Saharan Dust Transport Over the North Atlantic Ocean and Mediterranean: An Overview, in: The Impact of Desert Dust Across the Mediterranean. Kluwer Academic Publishers, The Netherlands, pp. 133–151.     107 Rae, R., Vincent, W., 1998. Phytoplankton production in subarctic lake and river ecosystems: Development of a photosynthesis-temperature-irradiance model. J Plankton Res 20, 1293– 1312.  Raven, J., 1988. The iron and molybdenum use efficiencies of plant growth with different energy, carbon and nitrogen sources. New Phytol 109, 279–287.  Raven, J., Evans, M., Korb, R., 1999. The role of trace metals in photosynthetic electron transport in O2-evolving organisms. Photosynth Res 60, 111–149.  Raymond, K., Muller, G., Matzanke, B., 1984. Complexation of iron by siderophores: A review of their solution and structural chemistry and biological function. Top Curr Chem 123, 49– 102.  Reid, R., Live, D., Faulkner, D., Butler, A., 1993. A siderophore from a marine bacterium with an exceptional ferric ion affinity constant. Nature 366, 455–458.  Richardson, T., Jackson, G., 2007. Small phytoplankton and carbon export from the surface ocean. Science 315, 838–840.  Rodríguez, F., Varela, M., Zapata, M., 2002. Phytoplankton assemblages in the Gerlache and Bransfield Straits (Antarctic Peninsula) determined by light microscopy and CHEMTAX analysis of HPLC pigment data. Deep-Sea Res Pt II 49, 723–747.  Rothrock, D., Zhang, J., Yu, Y., 2003. The arctic ice thickness anomaly of the 1990s: A consistent view from observations and models. J Geophys Res-Oceans 108, C33083.  Rönner, U., Sörensson, F., Holm-Hansen, O., 1983. Nitrogen assimilation by phytoplankton in the Scotia Sea. Polar Biol 2, 137–147.  Rue, E., Bruland, K., 1995. Complexation of iron(III) by natural organic ligands in the Central North Pacific as determined by a new competitive ligand equilibration/adsorptive cathodic stripping voltammetric method. Mar Chem 50, 117–138.  Rue, E., Bruland, K., 1997. The role of organic complexation on ambient iron chemistry in the equatorial Pacific Ocean and the response of a mesoscale iron addition experiment. Limnol Oceanogr 42, 901–910.  Rueter, J., Ades, D., 1987. The role of iron nutrition in photosynthesis and nitrogen assimilation in Scenedesmus Quadricauda (Chlorophyceae). J Phycol 23, 452–457.  Ryther, J., Kramer, D., 1961. Relative iron requirement of some coastal and offshore plankton algae. Ecology 42, 444–446.  Sarthou, G., Timmermans, K., Blain, S., Tréguer, P., 2005. Growth physiology and fate of diatoms in the ocean: A review. J Sea Res 53, 25–42.  108 Schwyn, B., Neilands, J., 1987. Universal chemical-assay for the detection and determination of siderophores. Anal Biochem 160, 47–56.  Sherr, E., Sherr, B., Wheeler, P., Thompson, K., 2003. Temporal and spatial variation in stocks of autotrophic and heterotrophic microbes in the upper water column of the central Arctic Ocean. Deep-Sea Res Pt I 50, 557–571.  Shi, D., Xu, Y., Hopkinson, B., Morel, F., 2010. Effect of ocean acidification on iron availability to marine phytoplankton. Science 327, 676–679  Schilitzer, R., 2011. Ocean Data View, http://odv.awi.de  Sigel, S., Payne, S., 1982. Effect of iron limitation on growth, siderophore production, and expression of outer membrane proteins of Vibrio cholerae. J Bacteriol 150, 148–155.  Smetacek, V., 1999. Diatoms and the ocean carbon cycle. Protist 150, 25–32.  Smetacek, V., Nicol, S., 2005. Polar ocean ecosystems in a changing world. Nature 437, 362– 368.  Stintzi, A., Barnes, C., Xu, J., Raymond, K., 2000. Microbial iron transport via a siderophore shuttle: A membrane ion transport paradigm. P Natl Acad Sci USA 97, 10691–10696.  Stroeve, J., Serreze, M., Fetterer, F., Arbetter, T., Meier, W., Maslanik, J., Knowles, K., 2005. Tracking the Arctic's shrinking ice cover: Another extreme September minimum in 2004. Geophys Res Lett 32, L04501.  Stroeve, J., Holland, M., Meier, W., Scambos, T., Serreze, M., 2007. Arctic sea ice decline: Faster than forecast. Geophys Res Lett 34, L09501.  Strzepek, R., Harrison, P., 2004. Photosynthetic architecture differs in coastal and oceanic diatoms. Nature 431, 689–692.  Sunda, W., 1997. Control of dissolved iron concentrations in the world ocean: A comment. Mar Chem 57, 169–172.  Sunda, W., Swift, D., Huntsman, S., 1991. Low iron requirement for growth in oceanic phytoplankton. Nature 351, 55–57.  Sunda, W., Huntsman, S., 1997. Interrelated influence of iron, light and cell size on marine phytoplankton growth. Nature 390, 389–392.  Suzuki, K., Minami, C., Liu, H., Saino, T., 2002. Temporal and spatial patterns of chemotaxonomic algal pigments in the subarctic Pacific and the Bering Sea during the early summer of 1999. Deep-Sea Res Pt II 49, 5685–5704.   109 Takahashi, K., Fujitani, N., Yanada, M., 2002. Long term monitoring of particle fluxes in the Bering Sea and the central subarctic Pacific Ocean, 1990–2000. Prog Oceanogr 55, 95–112.  Takeda, S., 1998. Influence of iron availability on nutrient consumption ratio of diatoms in oceanic waters. Nature 393, 774–777.  Toole, J., Timmermans, M., Perovich, D., Krishfield, R., Proshutinsky, A., Richter-Menge, J., 2010. Influences of the ocean surface mixed layer and thermohaline stratification on Arctic Sea ice in the central Canada Basin. J Geophys Res-Oceans 115, C10018.  Tortell, P., Maldonado, M., Price, N., 1996. The role of heterotrophic bacteria in iron-limited ocean ecosystems. Nature 383, 330–332.  Tortell, P., Payne, C., Li, Y., Trimborn, S., Rost, B., Smith, W., Riesselman, C., Dunbar, R., Sedwick, P., DiTullio, G., 2008. CO2 sensitivity of Southern Ocean phytoplankton. Geophys Res Lett 35, L04605.  Tovar-Sanchez, A., Duarte, C., Alonso, J., Lacorte, S., Tauler, R., Galbán-Malagón, C., 2010. Impacts of metals and nutrients released from melting multiyear Arctic sea ice. J Geophys Res-Oceans 115, C07003.  Tremblay, G., Belzile, C., Gosselin, M., Poulin, M., Roy, S., Tremblay, J., 2009. Late summer phytoplankton distribution along a 3500 km transect in Canadian Arctic waters: strong numerical dominance by picoeukaryotes. Aquat Microb Ecol 54, 55–70.  Tremblay, J., Michel, C., Hobson, K., Gosselin, M., Price, N., 2006. Bloom dynamics in early opening waters of the Arctic Ocean. Limnol Oceanogr 51, 900–912.  Tremblay, J., Simpson, K., Martin, J., Miller, L., Gratton, Y., Barber, D., Price, N., 2008. Vertical stability and the annual dynamics of nutrients and chlorophyll fluorescence in the coastal, southeast Beaufort Sea. J. Geophys. Res. 113, C07S90.  Trick, C., 1989. Hydroxamate-siderophore production and utilization by marine eubacteria. Curr Microbiol 18, 375–378.  Varela, D., Harrison, P., 1999. Seasonal variability in nitrogenous nutrition of phytoplankton assemblages in the northeastern subarctic Pacific Ocean. Deep-Sea Res Pt II 46, 2505–2538.  Vaulot, D., 1989. CYTOPC: Processing software for flow cytometric data. Signal Noise 2:8.  Venturi, V., Ottevanger, C., Leong, J., Weisbeek, P., 1993. Identification and characterization of a siderophore regulatory gene (pfrA) of Pseudomonas putida WCS358: homology to the alginate regulatory gene aigQ of Pseudomonas aeruginosa. Mol Microbiol 10, 63–73.  Verstreate, D., Storch, T., Dunham, V., 1980. A comparison of the influence of iron on the growth and nitrate metabolism of Anabaena and Scenedesmus. Physiol Plant 50, 47–51.  110 Vidussi, F., Roy, S., Lovejoy, C., Gammelgaard, M., Thomsen, H., Booth, B., Tremblay, J., Mostajir, B., 2004. Spatial and temporal variability of the phytoplankton community structure in the North Water Polynya, investigated using pigment biomarkers. Can J Fish Aquat Sci 61, 2038–2052.  Wells, M., Price, N., Bruland, K., 1994. Iron limitation and the cyanobacterium Synechococcus in equatorial Pacific waters. Limnol Oceanogr 39, 1481–1486.  Welschmeyer, N., Goericke, R., Strom, S., Peterson, W., 1991. Phytoplankton growth and herbivory in the subarctic Pacific: A chemotaxonomic analysis. Limnol Oceanogr 36, 1631– 1649.  Welschmeyer, N., Strom, S., Goericke, R., DiTullio, G., Belvin, M., Petersen, W., 1993. Primary production in the subarctic Pacific Ocean: Project SUPER. Prog Oceanogr 32, 101–135.  Westall, J.C., Zachary, J.L., Morel, F.M.M., 1976. MINEQL: A Computer Program for the Calculation of Chemical Equilibrium Composition of Aqueous Systems, R.M. Parsons Laboratory, Massachusetts Institute of Technology, Cambridge, MA.  Whitfield, M., Turner, D., 1987. The Role of Particles in Regulating the Composition of Seawater, in: Aquatic Surface Chemistry: Chemical Processes at the Particle-Water Interface. John Wiley and Sons, New York, pp. 457–493.  Whitney, F., Freeland, H., 1999. Variability in upper-ocean water properties in the NE Pacific Ocean. Deep-Sea Res Pt II 46, 2351–2370.  Witter, A., Hutchins, D., Butler, A., Luther, G., III, 2000. Determination of conditional stability constants and kinetic constants for strong model Fe-binding ligands in seawater. Mar Chem 69, 1–17.  Wright, S., Thomas, D., Marchant, H., Higgins, H., Mackey, M., Mackey, D., 1996. Analysis of phytoplankton of the Australian sector of the Southern Ocean: Comparisons of microscopy and size frequency data with interpretations of pigment HPLC data using the “CHEMTAX” matrix factorisation program. Mar Ecol-Prog Ser 144, 285–298.  Wright, S., van den Enden, R., Pearce, I., 2010. Phytoplankton community structure and stocks in the Southern Ocean (30–80°E) determined by CHEMTAX analysis of HPLC pigment signatures. Deep-Sea Res Pt II 57, 758–778.  Yamamoto-Kawai, M., Carmack, E., McLaughlin, F., 2006. Nitrogen balance and Arctic throughflow. Nature 443, 43–43.  Yamamoto-Kawai, M., McLaughlin, F., Carmack, E., Nishino, S., Shimada, K., 2008. Freshwater budget of the Canada Basin, Arctic Ocean, from salinity, #18O, and nutrients. J Geophys Res-Oceans 113, C01007.   111 Yamamoto-Kawai, M., McLaughlin, F., Carmack, E., Nishino, S., Shimada, K., Kurita, N., 2009. Surface freshening of the Canada Basin, 2003-2007: River runoff versus sea ice meltwater. J Geophys Res-Oceans 114, C00A05.  Yang, J., Comiso, J., Walsh, D., Krishfield, R., Honjo, S., 2004. Storm-driven mixing and potential impact on the Arctic Ocean. J Geophys Res-Oceans 109, C04008.  Zapata, M., Rodríguez, F., Garrido, J., 2000. Separation of chlorophylls and carotenoids from marine phytoplankton: A new HPLC method using a reversed phase C8 column and pyridine-containing mobile phases. Mar Ecol-Prog Ser 195, 29–45.  Zar, J., 1999. Biostatistical analysis. Prentice Hall, Upper Saddle River, New Jersey.  Zumft, W., 1974. Ferredoxin: Nitrate Oxidoreductase From Chlorella Purification and Properties, in: Bioenergetics & metabolism of green algae, Volume 2. MSS Information Corporation, pp. 14–26.   112 Appendix The experimental design of the grow out initiated at station L1 in the Beaufort Sea (see Chapter 2) included twelve treatments in duplicate, 24 bottles in total. The two treatments not discussed in Chapter 2 are + 5 nM DFB 10 %Io and + 10 nM DFB 10 %Io. In conjunction with the + 1 nM DFB 10 %Io treatment included in Chapter 2, these three treatments were utilized to ensure that Fe limitation of primary productivity would be observed in the grow out. Ultimately, the + 5 nM DFB 10 %Io and + 10 nM DFB 10 %Io treatments were not included in the text of Chapter 2 for two reasons. First, these treatments did not enhance the discussion of Fe limitation as Fe was shown to be limiting at in situ Fe levels, as well as with an addition of + 1 nM DFB. Second, the effects of the + 5 and 10 nM DFB treatments often yielded marginally different results than that of the + 1 nM DFB treatment; thus did not contribute interesting results. Tables including data associated with the + 5 and 10 nM DFB treatments are included below.   Table A.1: Control and experimental treatments (marked by X) and their corresponding light, iron, and nitrate levels.      Light level   1 %Io 10 %Io 50 %Io + 1 nM Fe  X  X  X no Fe added  X X X  X + 1 nM DFB  X  X  X + 5 nM DFB    X Fe  level + 10 nM DFB    X   no N added +10 µM N no N added +10 µM N no N added +10 µM N   Nitrate level   113 Table A.2: Average size-fractionated and total chlorophyll a (ng L-1) for triplicate initials on day 0 and duplicate samples for control and experimental treatments on the final day of an 8-day grow out initiated at station L1 (71°05.94’ N, 139°08.87’ W) in the Beaufort Sea. Values reported equal the mean ± standard error for initials and the range for day 8 duplicates; absence of a range indicates the treatment was only analyzed in singlet. Values in brackets represent the times increase from the control treatment.      Treatment 0.22-1 µm 1-5 µm > 5 µm Total Initial 9.8 ± 1.1 19 ± 1.9 13 ± 0.69 42 ± 3.3 Control 10 %Io, no NO3- 35 ± 6.8 21 ± 1.5 6.7 ± 0.31 63 ± 8.6 no Fe added 50 %Io 440 ± 53 (13) 260 ± 28 (12) 160 ± 7.6 (24) 860 ± 89 (14) no Fe added 10 %Io 160 ± 100 (5) 150 ± 0.71 (7) 81 ± 9.0 (12) 390 ± 92 (6) no Fe added 1 %Io 96 ± 3.3 (3) 50 ± 5.5 (2) 29 ± 0.67 (4) 180 ± 9.5 (3)  + 1 nM Fe 50 %Io 460 ± 22 (13) 270 ± 12 (13) 170 ± 3.8 (25) 900 ± 30 (14)  + 1 nM Fe 10 %Io 320 ± 30 (9) 140 ± 5.7 (7) 89 ± 2.8 (13) 550 ± 33 (9)  + 1 nM Fe 1 %Io 66 ± 1.8 (2) 35 ± 1.7 (2) 8.5 ± 2.8 (1) 110 ± 33 (2)  + 1 nM DFB 50 %Io 75 ± 2.8 (2) 57 ± 4.5 (3) 34 ± 1.9 (5) 170 ± 0.15 (3)  + 1 nM DFB 10 %Io 78 (2) 48 ± 0 (2) 32 ± 0.69 (5) 160 ± 56 (3)  + 1 nM DFB 1 %Io 52 ± 0.092 (1) 32 ± 0.83 (2) 13 ± 0.48 (2) 97 ± 0.26 (2)  + 5 nM DFB 10 %Io 67 ± 3.8 (2) 43 ± 0.48 (2) 22 ± 0.12 (3) 130 ± 4.2 (2)  + 10 nM DFB 10 %Io 51 ± 5.5 (1) 39 ± 3.1 (2) 22 ± 0.67 (3) 110 ± 1.7 (2)  114 Table A.3: Size-fractionated Fe:C ratios (µmol Fe (mol C)-1) of size-fractionated phytoplankton on day 6 of an 8-day grow out initiated at station L1 (71°05.94’ N, 139°08.87’ W) in the Beaufort Sea. Iron was precomplexed to DFB in the ratio 100 nmol L-1 Fe : 105 nmol L-1 DFB with 5 % of total Fe as 55Fe. Carbon was added to each bottle to a final concentration of 20 µCi L-1 14C. Values are equal to the mean ± the range for duplicate bottles.    Treatment 0.22-1 µm 1-5 µm >5 µm Control 10 %Io, no NO3- 14 ± 5.4 3.7 ± 0.78 14 ± 1.1 no Fe added 50 %Io 11 ± 8.4 2.7 ± 0.64 2.5 ± 0.40 no Fe added 10 %Io 4.8 ± 0.43 3.2 ± 0.83 2.6 ± 1.5 no Fe added 1 %Io 33 ± 0.55 32 ± 6.3 43 ± 4.9  + 1 nM Fe 50 %Io 8.7 ± 0.54 2.9 ± 0.17 3.0 ± 0.34  + 1 nM Fe 10 %Io 5.4 ± 2.1 1.6 ± 0.19 2.4 ± 0.89  + 1 nM Fe 1 %Io 270 ± 230 79 ± 59 81 ± 8.8  + 1 nM DFB 50 %Io 310 ± 52 52 ± 22 27 ± 8.0  + 1 nM DFB 10 %Io 15 ± 3.8 18 ± 13 5.5 ± 1.6  + 1 nM DFB 1 %Io 200 ± 150 100 ± 47 140 ± 85  + 5 nM DFB 10 %Io 47 ± 32 10 ± 5.5 11 ± 6.2  + 10 nM DFB 10 %Io 23 ± 3.1 7.6 ± 4.1 6.1 ± 3.9     115 Table A.4: Short-term (24 h) volumetric (!10-12 mol Fe L-1 d-1) and carbon-specific (!10-6 mol Fe (mol C)-1 d-1) Fe uptake rates of size-fractionated phytoplankton on day 6 of an 8-day grow out initiated at station L1 (71°05.94’ N, 139°08.87’ W) in the Beaufort Sea. Iron was precomplexed to ligands DFB or EDTA in the ratios 100 nmol L-1 Fe : 105 nmol L-1 DFB with 5 % of total Fe as 55Fe, or 100 nmol L-1 Fe : 1000 nmol L-1 EDTA with 1 % of total Fe as 55Fe. Values are equal to the mean ± the range for duplicate bottles.     !FeDFB     !FeEDTA Treatment   Volumetric       Carbon-specific     Volumetric      Carbon-specific 0.22-1 µm Control 10 %Io, no NO3- 0.17 ± 0.017 0.95 ± 0.39 2.1 ± 0.31 12 ± 5.3 no Fe added 50 %Io 0.80 ± 0.32 0.95 ± 0.38 7.5 ± 1.2 9.0 ± 1.5 no Fe added 10 %Io 0.26 ± 0.034 0.56 ± 0.058 3.9 ± 0.52 8.5 ± 0.88 no Fe added 1 %Io 0.27 ± 0.014 0.86 ± 0.059 3.6 ± 1.1 12 ± 3.3 + 1 nM Fe 50 %Io 0.69 ± 0.18 0.86 ± 0.24 5.5 ± 0.89 6.8 ± 1.2 + 1 nM Fe 10 %Io 0.30 ± 0.11 0.62 ± 0.25 2.7 ± 0.38 5.4 ± 0.44 + 1 nM Fe 1 %Io 1.4 ± 0.76 5.7 ± 3.1 3.1 ± 0.80 12 ± 3.2 + 1 nM DFB 50 %Io 3.1 ± 0.41 11 ± 0.090 5.8 ± 2.3 19 ± 5.8 + 1 nM DFB 10 %Io 0.14 ± 0.0017 0.68 ± 0.024 5.6 ± 1.1 28 ± 7.0 + 1 nM DFB 1 %Io 1.4 ± 0.82 7.2 ± 4.5 3.9 ± 0.0027 20 ± 0.98 + 5 nM DFB 10 %Io 0.49 ± 0.25 2.9 ± 1.5 4.0 ± 0.76 24 ± 4.7 + 10 nM DFB 10 %Io 0.11 ± 0.00085 0.69 ± 0.014 5.7 ± 0.64 37 ± 5.2 1-5 µm Control 10 %Io, no NO3- 0.068 ± 0.0032 0.37 ± 0.14 2.2 ± 0.49 13 ± 6.5 no Fe added 50 %Io 0.39 ± 0.095 0.47 ± 0.12 16 ± 1.6 19 ± 2.0 no Fe added 10 %Io 0.25 ± 0.018 0.55 ± 0.023 6.9 ± 0.080 15 ± 0.26 no Fe added 1 %Io 0.20 ± 0.036 0.63 ± 0.11 3.5 ± 0.66 11 ± 1.9 + 1 nM Fe 50 %Io 0.43 ± 0.11 0.53 ± 0.14 11 ± 0.079 14 ± 0.27 + 1 nM Fe 10 %Io 0.27 ± 0.040 0.55 ± 0.11 5.0 ± 0.055 9.9 ± 0.69 + 1 nM Fe 1 %Io 0.43 ± 0.23 1.7 ± 0.91 1.3 ± 0.078 5.2 ± 0.33 + 1 nM DFB 50 %Io 0.77 ± 0.18 2.6 ± 0.30 10 ± 0.40 37 ± 3.2 + 1 nM DFB 10 %Io 0.28 ± 0.15 1.3 ± 0.66 4.5 ± 0.11 22 ± 0.50 + 1 nM DFB 1 %Io 0.45 ± 0.20 2.4 ± 1.1 1.9 ± 0.11 9.7 ± 1.0 + 5 nM DFB 10 %Io 0.17 ± 0.074 0.99 ± 0.43 4.1 ± 0.12 24 ± 0.58 + 10 nM DFB 10 %Io 0.075 ± 0.021 0.49 ± 0.15 3.8 ± 0.61 25 ± 4.7 > 5 µm Control 10 %Io, no NO3- 0.12 ± 0.012 0.57 ± 0.15 2.7 ± 0.59 16 ± 7.9 no Fe added 50 %Io 0.17 ± 0.036 0.21 ± 0.042 31 ± 4.3 37 ± 5.0 no Fe added 10 %Io 0.088 ± 0.032 0.19 ± 0.064 18 ± 0.43 40 ± 0.21 no Fe added 1 %Io 0.18 ± 0.019 0.58 ± 0.051 4.9 ± 0.089 16 ± 0.53 + 1 nM Fe 50 %Io 0.20 ± 0.019 0.25 ± 0.026 20 ± 0.22 24 ± 0.57 + 1 nM Fe 10 %Io 0.14 ± 0.041 0.28 ± 0.098 6.9 ± 0.83 13 ± 0.86 + 1 nM Fe 1 %Io 0.19 ± 0.011 0.79 ± 0.043 1.7 ± 0.058 6.8 ± 0.25 + 1 nM DFB 50 %Io 0.60 ± 0.16 2.0 ± 0.32 22 ± 1.0 80 ± 6.3 + 1 nM DFB 10 %Io 0.093 ± 0.013 0.46 ± 0.041 11 ± 0.039 55 ± 2.8 + 1 nM DFB 1 %Io 0.25 ± 0.11 1.3 ± 0.60 2.0 ± 0.35 10 ± 2.2 + 5 nM DFB 10 %Io 0.18 ± 0.068 1.1 ± 0.40 7.8 ± 0.60 47 ± 3.3 + 10 nM DFB 10 %Io 0.070 ± 0.024 0.47 ± 0.17 5.6 ± 0.58 37 ± 4.9  116  Table A.5: Average NO3-, PO43-, and SiO42- drawdown ("mol L-1) for triplicate initials on day 0 and duplicate samples on all sampling days of an 8-day grow out initiated at station L1 (71°05.94’ N, 139°08.87’ W) in the Beaufort Sea. Initial values reported equal the mean ± standard error. Net drawdown values are equal to the day 8 nutrient concentration subtracted from the initial nutrient concentration. Maximum drawdown values are equal to the lowest nutrient concentration observed throughout the experiment subtracted from the initial nutrient concentration; the day of lowest nutrient concentration is reported in brackets. All net and maximum NO3- drawdown values are calculated with the Initial + NO3- value, save for the control treatment which is calculated with the initial value.     Nitrate  Phosphate  Silicate Initial Day 0 0.43 ± 0.039  0.76 ± 0.029  1.8 ± 0.13 Initial + NO3- Day 0 10.43 ± 0.039  0.76 ± 0.029  1.8 ± 0.13 net  0.22  0.050  0.10 Control 10 %Io no NO3- maximum 0.27 (6)  0.050 (6-8)  0.60 (2) net  0.81  0.11  0.40 no Fe added 50 %Io maximum 3.1 (6)  0.11 (8)  0.80 (2) net  1.1  0.090  0.30 no Fe added 10 %Io maximum 1.1 (8)  0.090 (8)  0.85 (2) net  0.53  0.070  0.40 no Fe added 1 %Io maximum 0.83 (4-6)  0.090 (6)  0.93 (2) net  0.43  0.10  0.30 + 1 nM Fe 50 %Io maximum 3.0 (6)  0.10 (8)  1.1 (2) net  2.9  0.10  0.30 + 1 nM Fe 10 %Io maximum 2.9 (8)  0.10 (8)  0.80 (2) net  2.3  0.080  0.10 + 1 nM Fe 1 %Io maximum 3.5 (4)  0.080 (8)  0.60 (2) net  1.3  0.070  0.30 + 1 nM DFB 50 %Io maximum 1.3 (8)  0.10 (6)  0.80 (2) net  1.9  0.080  0.20 + 1 nM DFB 10 %Io maximum 1.9 (8)  0.080 (8)  0.80 (2) net  2.6  0.070  0.20 + 1 nM DFB 1 %Io maximum 2.6 (8)  0.080 (6)  0.80 (2) net  3.1  0.080  0.40 + 5 nM DFB 10 %Io maximum 3.1 (8)  0.080 (6-8)  0.70 (2) net  2.9  0.070  0.20 + 10 nM DFB 10 %Io maximum 2.9 (8)  0.090 (6)  0.70 (2)

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0053264/manifest

Comment

Related Items