LIMNOLOGICAL AND PALEOLIMNOLOGICAL INVESTIGATIONS OF

ENVIRONMENTAL CHANGE IN THREE DISTINCT ECOSYSTEM TYPES,

CANADIAN HIGH ARCTIC

by

BRONWYN ELIZABETH KEATLEY

A thesis submitted to the Department of Biology

in conformity with the requirements for

the degree of Doctor of Philosophy

Queen’s University

Kingston, Ontario, Canada

September 2007

COPYRIGHT © BRONWYN ELIZABETH KEATLEY, 2007

ii

ABSTRACT

The biological remains preserved in the sediments of Arctic lakes and ponds in the

Canadian High Arctic are important indicators of environmental change, especially as long-term

instrumental data are often lacking. Although recent studies have underscored variability amongst

these aquatic ecosystems, data are lacking from several key ecosystems. This thesis addresses

some of these critical knowledge gaps in the Canadian high Arctic, using diatom-based

limnological and paleolimnological techniques.

First, I explore the limnology and diatom ecology along a gradient of bioclimatic zones

on Melville Island in the western High Arctic. Lakes and ponds located in the most lushly

vegetated zone were significantly different from those elsewhere on the island, both in terms of

measured limnological variables and in terms of diatom assemblage composition. Diatom species

distributions from Melville Island can best be explained by differences in pH and related

variables.

Secondly, ponds and lakes located in a High Arctic oasis on northern Ellesmere Island,

recorded significantly higher specific conductivity, nutrients, and dissolved organic carbon than

freshwater bodies from the surrounding polar desert.

In Chapter 5, I provide an examination of long-term environmental change from Melville

Island, a region of the High Arctic for which no paleolimnological data exist. The timing of

diatom shifts in a dated sediment core from a small pond is consistent with the onset of climate

warming in the early 20th century.

Differences in ice cover have often been invoked to explain differences in the timing and

magnitude of diatom shifts in the Arctic, but this hypothesis has not been explicitly tested. In

Chapter 6, I compare two adjacent lakes with similar physical characteristics but different ice

cover regimes from northern Ellesmere Island. I provide strong evidence that extended ice cover

dampens diatom community responses to environmental change.

iii

In the final chapters, I determine that marine-derived nutrients significantly affect the

limnology of ponds on Cape Vera, Devon Island, and are related to the degree of seabird

influence. Although a portion of diatom species distributions can be linked to seabird influence,

the most abundant taxa show little relation to the nutrient gradient. In a paleolimnological

context, diatoms do not appear to provide a robust indicator of seabirds in High Arctic ponds.

iv

CO-AUTHORSHIP This thesis conforms to the Manuscript Format as outline by the School of Graduate Studies and Research. Each chapter has been written in the form of the journal in which it was submitted and contains its own literature cited. Marianne Douglas and John Smol, my thesis co-supervisors, are co-authors on all chapters. All fieldwork was conducted by myself, John Smol and Marianne Douglas, with the assistance of others noted in the acknowledgements section of each chapter. As several of the chapters in this thesis are referred to as “Keatley et al. 2007” or “Keatley et al. in press” etc., each “Keatley et al.” reference is followed by square brackets [ ] enclosing the chapter number. Chapter 2 was co-authored by Marianne Douglas and John Smol. I conducted the statistical analyses, drafted all figures and tables, and was the primary author of the manuscript. This chapter has been published separately. [2] Keatley, B.E., M.S.V. Douglas, and J.P. Smol. 2007. Physical and chemical limnological characteristics of lakes and ponds across environmental gradients on Melville Island, Nunavut/N.W.T., High Arctic Canada. Fundamental and Applied Limnology. 168: 355-376. Chapter 3 was co-authored by Marianne Douglas and John Smol. I conducted the diatom-based lab work, diatom counts, statistical analyses, drafted all figures and tables, and was the primary author of the manuscript. This chapter has been submitted for publication and is currently in review. [3] Keatley, B.E., M.S.V. Douglas, and J.P. Smol. In review. Evaluating the role of environmental and spatial variables on diatom species distributions on Melville Island (Canadian high Arctic). Submitted: 7 June 2007. Chapter 4 was co-authored by Marianne Douglas and John Smol. I conducted the statistical analyses, drafted all figures and tables, and was the primary author of the manuscript. This chapter has been accepted for publication and is currently in press. [4] Keatley, B.E., M.S.V. Douglas, and J.P. Smol. 2007. Limnological characteristics of a high Arctic oasis and comparisons across northern Ellesmere Island. Arctic, in press. Chapter 5 was co-authored by Marianne Douglas and John Smol. I conducted the lab work, diatom counts, statistical analyses, drafted all figures and tables, and was the primary author of the manuscript. This chapter has been published separately. [5] Keatley, B.E., M.S.V. Douglas, and J.P. Smol. 2006. Early-20th century environmental change inferred using sub-fossil diatoms from a small pond on Melville Island, N.W.T., Canadian High Arctic. Hydrobiologia. 533:15-26. Chapter 6 was co-authored by Marianne Douglas and John Smol. I conducted the lab work, diatom counts, statistical analyses, drafted all figures and tables, and was the primary author of the manuscript. This chapter has been accepted for publication and is currently in press.

v

[6] Keatley, B.E., M.S.V. Douglas, and J.P. Smol. Prolonged ice cover dampens diatom community responses to recent climatic change in high Arctic lakes. Arctic Antarctic and Alpine Research, in press. Chapter 7 was co-authored by Marianne Douglas, John Smol, Jules Blais (University of Ottawa), and Mark Mallory (Canadian Wildlife Service). I conducted all the lab work related to diatoms, diatom counts, statistical analyses, drafted all figures and tables, and was the primary author of the manuscript. Jules Blais provided δ15N and elemental data from the surface sediments and Mark Mallory provided seabird ecological information. This chapter is in preparation for publication. [7] Keatley, B.E., M.S.V. Douglas, J. Blais, M. Mallory, and J.P. Smol. In preparation. Impacts of seabird-derived nutrients on water quality and diatom species assemblages from Cape Vera, Devon Island, Canadian High Arctic. Chapter 8 was co-authored by Marianne Douglas, John Smol, Neal Michelutti (Queen’s University), Jules Blais (University of Ottawa), and Mark Mallory (Canadian Wildlife Service). I conducted all the diatom-based lab work, diatom counts, all of the statistical analyses, drafted all figures and tables, and was the primary author of the manuscript. Jules Blais provided the elemental and stable isotope data, and the 210Pb dates. Neal Michelutti provided the sedimentary inferred-chlorophyll a data analysed the 210Pb data. This chapter is in preparation for publication. [8] Keatley, B.E., M.S.V. Douglas, N. Michelutti, J. Blais, M. Mallory, and J.P. Smol. In preparation. Tracking seabirds through time: a multi-proxy paleolimnological study in the Canadian High Arctic.

vi

ACKNOWLEDGEMENTS First and foremost, I would like to thank my co-supervisors, John Smol and Marianne Douglas,

for their support and encouragement over the years. John and Marianne, you are not only

outstanding scientists, but fantastic people. I want to thank you for sharing your insights into

science and life in general, but most of all, thank you both for introducing me to the magical place

that is the Canadian High Arctic

I would also like to thank Brian Cumming, Scott Lamoureux, and Shelley Arnott for providing

feedback as members of my PhD committee, and Jules Blais, Linda Kimpe, Mark Mallory,

Shelley Arnott, and Rene Gregory-Eaves for all their help along the way in various capacities as

collaborators, field assistants, and general scientific discussion partners. All of the past and

present PEARL members have made PEARL a lovely place to work, but I would particularly like

to single out Kat Rühland, Neal Michelutti, and John Glew for their years of scientific discussions

and friendships.

Some of the most amazing friendships I have yet formed have occurred because of my tenure in

graduate school, and I would especially like to acknowledge Anita Holtham, Anne Harris, Carrie

Lyons, Claudia Kraft, Dan Selbie, David Chiasson, Dermot Antoniades, Jon Sweetman, Laura

Lawlor and Roger Bull. Even though “a lot of things can go wrong,” you’ve all been there to

help make sure that, in the end, things usually go right. Extra special thanks go to Angela

Strecker for joining me through all those grad school rites of passage and for being such a

fantastic friend.

I would like to thank my family (Sondra & Evans Keatley, Sarah & Jonathan Sewter, Beth

McFarlane, Franklin Keatley, and Josh & Catherine Keatley,) and extra-grad school friends

(especially Bridget Meigs, Jonathan Hill and Mico Devos) for providing moral support and for

helping me to always keep things in perspective. Finally, I would like to thank my partner,

Alexandre Poulain, for his encouragement, love, and support.

vii

TABLE OF CONTENTS Abstract…………………………………………………………………………………………….ii Co-authorship……………………………………………………………………………………..iv Acknowledgements……………………………………………………………………………..…vi Table of Contents……………………………………………………………………………...…vii List of Tables………………………………………………………………………………...…...xii List of Figures…………………………………………………………………………………...xiv Chapter 1: General Introduction and Literature Review………………………………....……1 Environmental Change in the Arctic …………………………………….……………...…………1 Canadian High Arctic limnology…..……………………………...…………………………….…2 Diatoms as environmental indicators…………………………………………………...…………4 Paleolimnology in the Canadian High Arctic………………………………………………..…….5 Impacts of seabird-derived nutrients on Arctic lakes……………………………..……………….7 Summary of thesis objectives…...……...……………………………………………..……………8 References…………………………………………………...……...………...………………..…10 List of Figures…………………………………………………………………………………….17 Figure…………………………………………………………………...…………………...……18 Chapter 2: Physical and chemical limnological characteristics of lakes and ponds across environmental gradients on Melville Island, Nunavut/N.W.T., High Arctic Canada………19 Abstract……………………………………………………………………………………..…….20 Introduction……………………………………………………………………………………….21 Site description……………………………………………………………………………...…….23 Methods……………………………………………………………………………………..…….25 Results and Discussion……………………………………………………………………..…….28 Conclusions……………………………………………………………………………………….41 Acknowledgements……………………………………………………………………...…..…….44 References…………………………………………………………………………………..…….45 List of Figures…………………………………………..………………….………………….….49 Figures…………………………………………………………………….………………..…….50 Tables……………………………………………………………………………………….…….55 Chapter 3: Evaluating the role of environmental and spatial variables on diatom species distributions on Melville Island (Canadian High Arctic)……………………………….…….65 Abstract……………………………………………………………………………………..…….66 Introduction……………………………………………………………………………………….67 Methods……………………………………………………………………………..............…….69 Results and Discussion……………………………………………………………………..…….73 Summary….……………………………………………………………………………………….81 Acknowledgements……………………………………………………………………...…..…….83 References…………………………………………………………………………………..…….83 List of Figures…………………………………………..………………….…………………..…87

viii

Figures…………………………………………………………………….………………..…….88 Tables…………………………………………………………………………………….…….....94 Chapter 4: Limnological characteristics of a high Arctic oasis and comparisons across northern Ellesmere Island……………………………………………………………………..101 Abstract…………………………………………………………………………………….……102 Introduction……………………………………………………………………………..……….103 Methods………………………………………………………………………………..…..…….105 Results and Discussion……………………………………………………………………..…...109 Summary and Conclusions…………………………………………………………………...….117 Acknowledgements……………………………………………………………………...…….…119 References…………………………………………………………………...……………..……120 List of Figures…………………………………………..………………….……………………124 Figures…………………………………………………………………….…………….....……125 Tables……………………………………………………………………………………………129 Chapter 5: Early-20th century environmental changes inferred using sub-fossil diatoms from a small pond on Melville Island, N.W.T., Canadian high Arctic……………….………...…138 Abstract…………………………………………………………………………………….……139 Introduction……………………………………………………………………………..……….140 Study site……………………………………………………………………………..…..……...142 Materials and Methods………………………………………………………………..…..…….143 Results……………………………………………………………………………..…..…….…..144 Discussion……………………………………………………………………..…………….…..146 Conclusions………………………………………………………………….......................…....151 Acknowledgements……………………………………………………………………...…….…152 References…………………………………………………………………...……………..……152 List of Figures…………………………………………..………………….……………………158 Figures…………………………………………………………………….…………….....……159 Tables……………………………………………………………………………………………164 Chapter 6: Prolonged ice cover dampens diatom community responses to recent climatic change in high Arctic lakes………………………………………………………………...….165 Abstract…………………………………………………………………………………….……166 Introduction……………………………………………………………………………..……….167 Site description…...…………………………………………………………………..…..……...169 Methods………………………………………………………………………………..…..…….170 Results……………………………………………………………………………..…..…….…..172 Discussion……………………………………………………………………..…………….…..173 Conclusions………………………………………………………………….......................…....178 Acknowledgements……………………………………………………………………...…….…179 References…………………………………………………………………...……………..……180 List of Figures…………………………………………..………………….……………………185 Figures…………………………………………………………………….…………….....……186 Tables……………………………………………………………………………………………193 Chapter 7: Impacts of seabird-derived nutrients on water quality and diatom species assemblages from Cape Vera, Devon Island, Canadian High Arctic…………………….…196 Abstract…………………………………………………………………………………….……197 Introduction……………………………………………………………………………..……….198 Site description……...………………………………………………………………..…..……...200

ix

Methods………………………………………………………………………………..…..…….201 Results and Discussion…..………………………………………………………..…..…….…..204 Conclusions………………………………………………………………….......................…....210 Acknowledgements……………………………………………………………………...…….…211 References…………………………………………………………………...……………..……212 List of Figures…………………………………………..………………….……………………218 Figures…………………………………………………………………….…………….....……219 Tables……………………………………………………………………………………………226 Chapter 8: Tracking seabird population dynamics using paleolimnology: A case study from Devon Island, Arctic Canada………………………………………………………………….231 Abstract…………………………………………………………………………………….……232 Introduction……………………………………………………………………………..……….233 Site description…………………………………………………………………...…..…..……...235 Methods………………………………………………………………..…..…………………….236 Results……………………………………………………………………………..…..…….…..238 Discussion……………………………………………………………………..…………….…..245 Acknowledgements……………………………………………………………………...…….…250 References…………………………………………………………………...……………..……250 List of Figures…………………………………………..………………….……………………254 Figures…………………………………………………………………….…………….....……255 Tables……………………………………………………………………………………………276 Chapter 9: General Discussion and Conclusions…………………………………………….277 Modern limnology………………………………………………….…………………..….…….277 Diatom ecology………………………………………………………………………....……….279 Paleolimnology…………………………….…………………………………………....……….280 Future directions…………….……………………….………………………………....……….281 References…………….…………………………………………....……………………………283 Appendices…………………………………………………...…………………………………285 Appendix 1………………………………………………………………………………………286 Detailed description of methods used to collect water samples during field work. Appendix 2………………….………………………………………..………………..…….......287 Spatial variable results from Melville Island diatom calibration set. Appendix 3. ………….…………………….………………………………………..……..…....288 Species response scatterplots relative to specific conductivity for diatoms from Melville Island surface sediments. Appendix 4. ………….…………………….………………………………………..……..…....290 Summary statistics and estimated specific conductivity optima of various weighted averaging models for diatom-inferred specific conductivity from surface sediments of Melville Island. Appendix 5. ………………….……………………………………………………...……..…....292 Raw diatom counts from Melville Island surface sediments.

x

Appendix 6. ………….…………………….………………………………………..……..…....298 Trace metal water chemistry data from northern Ellesmere Island. Appendix 7. ………….…………………….………………………………………..……..…....299 Trace metal water chemistry data from the oasis region northern of Lake Hazen, Ellesmere Island. Appendix 8. ………….…………………….………………………………………..……..…....300 Raw diatom counts from core MVAT, Melville Island. Appendix 9. ……….…………………….………………………………………..……..……....302 Raw diatom counts from Skeleton Lake (EP1), Ellesmere Island. Appendix 10………….…………………….………………………………………..……..…....303 Raw diatom counts from EP2, Ellesmere Island. Appendix 11. ……….…………………….………………………………………..……..…......304 210Pb summary data for Skeleton Lake, Ellesmere Island. Appendix 12………….…………………….………………………………………..……..…....305 Total Pb and Hg from Skeleton Lake, Ellesmere Island. Appendix 13………….…………………….………………………………………..……..…....306 Raw diatom counts from Cape Vera, Devon Island surface sediments. Appendix 14. ……….…………………….………………………………………..……..……..308 Raw diatom counts from CV5 core, Cape Vera, Devon Island. Appendix 15………….…………………….………………………………………..……..…....309 Raw diatom counts from CV6 core, Cape Vera, Devon Island. Appendix 16. ………….………………….………………………………………..……..…......310 Raw diatom counts from CV7 core, Cape Vera, Devon Island. Appendix 17. ……….…………………….………………………………………..……..…......311 Raw diatom counts from CV9 core, Cape Vera, Devon Island. Appendix 18………….…………………….………………………………………..……..…....312 Raw diatom counts from CV9a core, Cape Vera, Devon Island. Appendix 19. ……….…………………….………………………………………..…..…..…....313 Raw diatom counts from CV12 core, Cape Vera, Devon Island. Appendix 20………….…………………….………………………………………..……..…....314 Raw diatom counts from CV13 core, Cape Vera, Devon Island. Appendix 21. ………….………………….………………………………………..……..……..315 Raw diatom counts from CV20 core, Cape Vera, Devon Island. Appendix 22. ……….…………………….………………………………………..…………....316 Raw diatom counts from CV22 core, Cape Vera, Devon Island.

xi

Appendix 23………….…………………….………………………………………..……..…....317 Raw diatom counts from CV24 core, Cape Vera, Devon Island.

xii

LIST OF TABLES Chapter 2 Table 1. …………………………………………………………………….……………………55 The chemical and physical parameters for 46 freshwater sites on Melville Island. Table 2. …………………………………………………………………….……………………61 Selected nutrient ratio parameters for 46 freshwater sites on Melville Island. Table 3. …………………………………………………………………….……………………63 Pearson correlation matrix with Bonferroni-adjusted probabilities. Chapter 3 Table 1. …………………………………………………………………….………………..…..94 Summary of selected limnological characteristics for sites from Melville Island. Table 2. …………………………………………………………………….………………….…95 List of diatom species found in Melville Island surface sediment samples in >1% relative abundance from at least three sites, or >10% relative abundance in at least one site. Table 3. ………..…………………………………………………….………………………..…98 Analysis of similarity (ANOSIM) results for surface sedimentary diatom assemblages from Melville Island. Table 4. ………………………………………………………….…………………………..…..99 List of species contributing the most to the dissimilarity (Diss/SD; average dissimilarity/standard deviation) between Zones 4 and 1 on Melville Island, based on an analysis of similarity percentages (SIMPER). Table 5. ………………………………………………………….…………………………..….100 Summary statistics of various weighted averaging models for pH. Chapter 4 Table 1. ………………………………………………………….…………………………..…129 Summary of selected limnological variables for the northern Ellesmere sites and Lake Hazen. Table 2. ………………………………………………………….…………………………..…133 Summary of selected limnological variables for the oasis sites, with abbreviations as describe in Table 1. Table 3. ………………………………………………………….…………………………...…136 Pearson correlation matrix with Bonferroni-adjusted probabilities. Chapter 5 Table 1. ………………………………………………………….…………………………..…164 Present-day physical and chemical characteristics of pond MV-AT were collected on July 24, 2002. Chapter 6 Table 1. ………………………………………………………….…………………………...…193 Selected limnological characteristics for the three study lakes.

xiii

Table 2. ………………………………………………………….…………………………...…194 Diatom taxa (>3% relative abundance) from the top 2.5 cm intervals from the sediment cores of Skeleton Lake and EP2 and the surface sediment (~2 cm) diatoms from EP3. Table 3. ………………………………………………………….…………………………..….195 Common diatom taxa (>3% relative abundance) found in the sediment cores from Skeleton Lake and EP2. Chapter 7 Table 1. ………………………………………………………….…………………………..….226 Selected limnological data for ponds located near Cape Vera, Devon Island. Table 2. ………………………………………………………….…………………………...…228 Pearson’s correlation matrix with Bonferroni-adjusted probabilities. Significant values are denoted in bold (p<0.01) or in underlined italics (p<0.05). Table 3. ………………………………………………………….…………………………...…229 Summary of dominant diatom species (at least 1% relative abundance in at least 3 sites) found in surface sediments of Cape Vera ponds. Table 4. ………………………………………………………….…………………………..…230 List of dominant diatom species arranged in order of their axis 1 species scores based on a redundancy analysis (RDA) constrained solely to δ15N. Chapter 8 Table 1. ………………………………………………………….…………………………..…276 Summary of limnological data from each pond cored at Cape Vera, Devon Island.

xiv

LIST OF FIGURES Chapter 1 Figure 1. ………………………………………………………….………………………….…18 Regional map of the Canadian Arctic Archipelago with the locations of limnological, diatom ecological, and paleolimnological studies conducted by the Smol and Douglas laboratories prior to this thesis. Chapter 2 Figure 1. ………………………………………………………….…………………………...…50 Regional map of the Canadian Arctic Archipelago with the locations of both Melville Island and previous modern limnological studies to which references are made in the text. The inset map A) indicates the location of the Canadian High Arctic within Canada, and the circle on the main map B) indicates Melville Island, the focus of our study. Figure 2. ………………………………………………………….…………………………....…51 Map of Melville Island, with sites differentiated according to previously defined a) bioclimatic zones (after EDLUND 1994), and b) bedrock geology (after HARRISON 1994). On both maps, numbers indicate the following geographical features: 1) Liddon Gulf, 2) Murray Inlet, 3) Purchase Bay, 4) Leopold Glacier, 5) unnamed ice caps, 6) Bridport Inlet, 7) Sabine Bay, 8) Hecla and Griper Bay. 2a) Bioclimatic zone 1 is the most sparsely vegetated region, while zone 4 has the greatest number and abundance of terrestrial vegetation, including woody shrubs (EDLUND 1994). Figure 3. ………………………………………………………….…………………………....…52 Box plots showing variability of selected environmental variables from islands of the Canadian Arctic Archipelago. Solid lines indicate median values, dashed lines indicate mean values, whiskers represent 10th and 90th percentiles, and dots are 5th and 95th percentiles. Data sources are as follows: Prince Patrick Island (ANTONIADES et al. 2003a), Banks Island (LIM et al. 2005), Victoria Island (MICHELUTTI et al. 2002a), Bathurst Island (LIM et al. 2001), Devon Island (LIM & DOUGLAS 2003), Ellef Ringnes Island (ANTONIADES et al. 2003b), Axel Heiberg Island (MICHELUTTI et al. 2002b), Alert (ANTONIADES et al. 2003a). Figure 4. ………………………………………………………….…………………………....…53 Principal components analysis (PCA) biplot of measured environmental variables (arrows) and sampling sites (symbols). The light lines represent variables that were run passively in the PCA. Sampling sites are differentiated into: 4a) dominant type of bedrock geology (after HARRISON 1994), or 4b) bioclimatic zones (after EDLUND 1994). See caption of Figure 2a and 2b for symbol legends. Figure 5. ………………………………………………………….……………………….......…54 Box plots of selected variables showing sites from bioclimatic zone 4, the mostly lushly vegetated zone on Melville Island (n=16) versus all other bioclimatic zones combined (n=30). Solid lines indicate median values, dashed lines indicate mean values, whiskers represent 10th and 90th percentiles, and dots are 5th and 95th percentiles. Chapter 3 Figure 1. ……………………………………………………….…………………………....…88 Map showing the location of (a) Melville Island in relation to Canada, (b) existing diatom calibration sets in the Canadian Arctic, and (c) the 45 lakes and ponds in this study. The numbers given in (b) correspond to the following studies: 1 = Mould Bay, Prince Patrick Island

xv

(Antoniades et al. 2005); 2 = Isachsen, Ellef Ringnes Island (Antoniades et al. 2003b); 3 = Axel Heiberg Island (Michelutti et al. 2006); 4 = Alert, Ellesmere Island (Antoniades et al. 2005); 5 = Cape Herschel, Ellesmere Island (Douglas and Smol 1993, 1995), 6 = Devon Island (Lim 2004); 7 = Cornwallis Island (Michelutti et al. in press); 8 = Bathurst Island (Lim et al. 2001a, b); 9 = Banks Island (Lim et al. 2007). Figure 2. ……………………………………………………….………………………….…......89 Histograms of the common species found in the surface sediments of 45 Melville Island lakes and ponds. Only species present in at least 1% relative abundance in at least 10 sites are shown, arranged in order of DCA axis 1 species scores. The 45 study sites are also arranged in order of DCA axis 1 sample scores. Measured pH and specific conductivity values are plotted to the far right of the figure. Figure 3. ……………………………………………………….…………………………...….....90 Canonical Correspondence Analysis (CCA) of a) sites and environmental variables, and b) diatom species and environmental variables from Melville Island. See Table 2 for species code numbers. Figure 4. ……………………………………………………….………………………….……...91 Relationship between observed and estimated pH values based on the WAtol inv model (n = 90) for a) bootstrapped pH values (r2

boot = 0.432), and b) bootstrapped pH residuals. Figure 5. ……………………………………………………….………………………….……...92 Species response curves of common species (found in at least 1% relative abundance in at least 10 sites) in relation to pH. Chapter 4 Figure 1. ……………………………………………………….………………………….…....125 Location map of northern Ellesmere Island. Inset a) indicates Ellesmere Island within Canada. Inset b) shows the northern sites around Ellesmere Island. The dashed black line denotes the boundary of Quttinirpaaq National Park and the patterned areas within this boundary represent different climate regions based on Thompson (1994). The black star indicates the location of the oasis sites detailed in inset c). Inset c) details the oasis sites just north of Lake Hazen. Figure 2. ……………………………………………………….………………………….…....126 Plots of total phosphorus unfiltered (TPu) versus total nitrogen (TN) in a) the oasis sites, and b) the northern sites. While there is little relationship between TPu and TN in the northern region, there is a clear positive relationship with TPu and TN in the oasis sites, suggesting that different factors control nutrient cycling within the two regions. Figure 3. ……………………………………………………….………………………….…....127 Biplot of a principal components analysis (PCA) of measured limnological variables for all sites. Oasis sites are represented by filled circles and northern sites are represented by open circles. Lake Hazen is kept separate due to its extremely large size and is represented by a star. Axis 1 most closely represents nutrients and related variables, pH, and conductivity, and explains 52.9% of the variance in the dataset. Axis 2 most closely represents a gradient of metals and explains 16% of the variance in the dataset. The dashed lines represent variables that were run passively in the ordination.

xvi

Figure 4. ……………………………………………………….………………………….…....128 Histograms indicating the change in the values of selected limnological variables in 2003 relative to 1963 for a) pH, b) specific conductivity, c) K, and d) SiO2. Site names with * indicate sites for which identification was approximate. Chapter 5 Figure 1. ……………………………………………………….………………………….…....159 Map showing location of sites discussed in this paper. The oval on the inset map shows location of Canadian High Arctic. Sites are as follows: A) MV-AT, Melville Island; B) Isachsen, Ellef Ringnes Island; C) Alert, Ellesmere Island; D) Tuborg Lake, Ellesmere Island; E) Fosheim Peninsula, Ellesmere Island; F) Devon Ice Cap, Devon Island; G) Cape Hershel, Ellesmere Island; H) Agassiz Ice Cap, Ellesmere Island; I) Melville Island ice caps, Melville Island; J) Meighen Ice Cap, Meighen Island; K) Char Lake, Resolute Bay, Cornwallis Island; L) Mould Bay, Prince Patrick Island; M) Bathurst Island; N) Victoria Island. Figure 2. ……………………………………………………….………………………….…....160 Sedimentation rate for dated sections of the MV-AT core, as calculated based on the Constant Rate of Supply (CRS) method described by Appleby (2001) and Binford (1990). Figure 3. ……………………………………………………….………………………….…....161 Diatom profile showing the common diatom species found in pond MV-AT. Individual species with >5% relative abundance in at least one sample were retained for the profile; “other” is a sum of all other diatoms found in each interval. Dates are based on 210Pb dating using a Constant Rate of Supply model. Percent loss-on-ignition (%LOI 550) is expressed as a % of combustion at 550˚C, and is a proxy for organic matter content of the sediment. Percent carbonates (%LOI 1000) is expressed as a percentage of dry weight combusted at 1000˚C. Zones are based on optimal splitting and broken-stick analysis. While the most marked species change occurs at ~5.25 cm, the shift in diatom assemblage appears to have begun earlier (~5.75 cm). Figure 4. ……………………………………………………….………………………….…....162 Detrended Correspondance Analysis (DCA) of diatom species scores versus depth. The DCA axis 1 species scores are scaled in Standard Deviation (S.D.) units, and provide an estimate of species turnover. Figure 5. ……………………………………………………….………………………….…....163 Mean June-July-August temperature data and annual precipitation data from Mould Bay, Prince Patrick Island (see Figure 1 for location) between 1948-1996 (Meteorological Service of Canada 2004). The smoothed line is a LOWESS curve with a span of 0.35. Chapter 6 Figure 1. ……………………………………………………….………………………….…....186 Map showing the locations of the study site (star) and the other sites (numbered) mentioned in the text: 1. Alert, 2. Hazen Plateau, 3. Agassiz Ice Cap, 4. Fosheim Peninsula, 5. Cape Herschel, 6. Isachsen, Ellef Ringnes Island, 7. Char Lake, Cornwallis Island. Figure 2. ……………………………………………………….………………………….…....187 Topographical map detailing the local topography near Skeleton Lake, EP2 and EP3. Figure 3. ……………………………………………………….………………………….….....188 Ice-off dates for Skeleton Lake (solid bars) and Lake EP2 (hatched bars). The data from the 1960s is from Oliver and Corbett (1966), while the 2003 data is from our own field observations.

xvii

These data are also corroborated by archival air photos that show ice persisting on Skeleton Lake after Lake EP2 has become ice-free. Figure 4. ……………………………………………………….………………………….…....189 Principal components analysis (PCA) biplot based on measured water chemistry variables that are considered to influence diatom assemblages. The arrows represent the measured environmental variables, while the open circles represent 52 lakes and ponds across northern Ellesmere Island. Skeleton Lake is represented by the solid square, EP2 is represented by the open square, and EP3 is represented by the solid circle. The proximity of these three sites represents their highly similar water chemistry. Figure 5. ……………………………………………………….………………………….…....190 5a) Age-depth model for Skeleton Lake based on CRS model (Binford, 1990), and b) Total 210Pb activities as estimated by alpha spectroscopy. The dashed line indicates estimated supported 210Pb. Figure 6. ……………………………………………………….………………………….…....191 6a) Diatom profile of Skeleton Lake showing taxa present in at least at least one interval with a relative abundance of >3%. See Table 2 for synonyms for some of the common taxa. Percent loss-on-ignition (%LOI; an estimate of organic matter) and PCA axis 2 sample scores (PCA2; a summary of change occurring in rare taxa) are presented at the right side of the profile. Figure 6b). Photographs of ice cover on Skeleton Lake and Lake EP2, indicating the physical proximity of the two lakes, Blister Hill, and a nearby pingo for reference (photographs taken 7 July 2003). Figure 6c). Diatom profile of EP2 showing only species that are present in at least 3% relative abundance in at least one interval. See Table 2 for synonyms for some of the common taxa. Percent loss-on-ignition (% LOI), total Pb and total Hg (both expressed per gram organic carbon), and PCA axis two sample scores (PCA2) are presented at the right side of the profile. The rise in total Pb and Hg are interpreted to mark the onset of anthropogenic pollution (mid-19th to early 20th century). Figure 7. ……………………………………………………….………………………….…....192 Schematic diagrams illustrating possible diatom responses to changing ice cover conditions in the two study lakes. Although the length of the ice-free season has likely increased in both lakes, it has yet to reach a critical threshold in Skeleton Lake, resulting in a muted diatom response. In Lake EP2, reduction in ice cover has crossed this critical threshold and resulted in marked diatom changes. Chapter 7 Figure 1. ……………………………………………………….………………………….…....219 Map of study location at Cape Vera, Devon Island. Insets showing location of: a) Devon Island within Canada; b) Cape Vera on Devon Island; and c) sites located throughout Cape Vera. The isoclines are not drawn to scale. The gradient indicates the approximate concentration of northern fulmars along the cliffs at Cape Vera, with the most birds occurring within the southern third of the colony. Figure 2. ……………………………………………………….………………………….…....220 Principal Components Analysis (PCA) biplot of measured environmental variables and Cape Vera sites. The “control sites” (CV11, 16, 17, 18, 22, 23, 24) are denoted with open circles while the rest of the ponds are denoted with solid circles.

xviii

Figure 3. ……………………………………………………….………………………….…....221 Box plots comparing selected measured environmental variables (TPu, TPf, TN, DOC, pH, chla) between Cape Vera ponds (n = 24, this study) and ponds located in the nearby Haughton Crater, Devon Island (n = 22, Lim & Douglas 2003). Figure 4. ……………………………………………………….………………………….…....222 Composite PCA biplot of selected measured environmental variables common to our labs’ previous limnological surveys in the Canadian Arctic Archipelago. References for each study are as follows: Melville Island (Keatley et al. [2]), Mould Bay, Prince Patrick Island (Antoniades et al. 2003a), Banks Island (Lim et al. 2005), Victoria Island (Michelutti et al. 2002a), Isachsen, Ellef Ringnes Island (Antoniades et al. 2003b), Bathurst Island (Lim et al. 2001), Cornwallis Island (Michelutti et al. in press a), Devon Island (Lim & Douglas 2003), Axel Heiberg Island (Michelutti et al. 2002b), northern Ellesmere Island (Keatley et al. [4]), Alert, Ellesmere Island (Antoniades et al. 2003). Figure 5. ……………………………………………………….………………………….…....223 Histogram of dominant diatom species (>1% relative abundance in at least 5 sites) found in ponds from Cape Vera. Both the sites and the species are ordered according to their DCA axis 1 scores. Figure 6. ……………………………………………………….………………………….…....224 Principal components analysis (PCA) biplot of species and sites from the surface sediments of Cape Vera, Devon Island. Figure 7. ……………………………………………………….………………………….…....225 Redundancy analysis (RDA) biplot constrained to the three measured environmental variables that explained significant portions of the diatom species variance (δ15N, specific conductivity, and dissolved organic carbon (DOC)). Biplot 7a) presents the diatom species, and 7b) the sites in relation to the environmental variables. Chapter 8 Figure 1. ……………………………………………………….………………………….…....255 Location of study site. A) Regional map of the Canadian Arctic with inset detailing location of the Canadian Arctic within Canada; B) Enlargement of Devon Island region with star identifying the location of Cape Vera; and C) Location of study ponds at Cape Vera, Devon Island. Figure 2. ……………………………………………………….………………………….…....256 Photograph of a suite of ponds directly below the cliffs at Cape Vera, July 2004. Figures 3 – 11. Plots for each core summarizing overall changes in diatom assemblage composition based on Principal Components Analysis Axis 1 (PCA1) and PCA axis 2 (PCA2) sample scores, changes in diatom assemblage diversity based on Hill’s N2 (N2; Hill 1973), seabird-derived nutrients based on (δ15N (‰)), sedimentary-inferred chlorophyll a concentrations (Chla mg/g dry weight), %N, %C, and C/N ratios. Figure 3. CV5 summary plot. ………….………………………………………..….…....257 Figure 4. CV6 summary plot. ………….………………………………………..….…....258 Figure 5. CV9 summary plot. ………….………………………………………..….…....259 Figure 6. CV9a summary plot. ……….………………………………………..….…......260 Figure 7. CV12 summary plot. ………….……………………………………..…...…....261 Figure 8. CV13 summary plot. ………….…………………………………..….…..........262 Figure 9. CV20 summary plot. ………….…………………………………..….……......263

xix

Figure 10. CV22 summary plot. ………….…………………………………..….…........264 Figure 11. CV24 summary plot. ………….…………………………………..….……....265 Figures 12 – 21. Diatom stratigraphic plots for each core from the Cape Vera ponds. These plots include all species present in at least 3% relative abundance in at least one sample. Figure 12. CV5 diatom stratigraphy. ………….………………………………….….......266 Figure 13. CV6 diatom stratigraphy. ………….…………………………………….…...267 Figure 14. CV7 diatom stratigraphy. ………….…………………………………..……..268 Figure 15. CV9 diatom stratigraphy. ………….…………………………………..…......269 Figure 16. CV9a diatom stratigraphy. ………….………………………………….….....270 Figure 17. CV12 diatom stratigraphy. ………….…………………………………..…....271 Figure 18. CV13 diatom stratigraphy. ………….…………………………………..…....272 Figure 19. CV20 diatom stratigraphy. ………….…………………………………..…....273 Figure 20. CV22 diatom stratigraphy. ………….…………………………………..…....274 Figure 21. CV24 diatom stratigraphy. ………….…………………………………..…....275

CHAPTER 1

GENERAL INTRODUCTION AND LITERATURE REVIEW

ENVIRONMENTAL CHANGE IN THE ARCTIC

The Canadian High Arctic is a region that is especially sensitive to global environmental

changes (ACIA 2004). Due to a number of positive feedback loops (e.g. snow-ice albedo,

permafrost thawing), temperature increases resulting from warming are likely to be amplified in

high-latitude areas. Some of these changes are apparent even over the course of the relatively

short monitoring window. For example, Arctic regions have warmed at a rate that is more than

double the global average over the past century, the extent of Arctic sea ice has diminished by

2.7% per decade since 1978, and the uppermost layers of permafrost have warmed by as much as

3 °C since the 1980s (IPCC 2007). Climate models predict further changes in the future, as

average annual Arctic temperatures are projected to rise by 2 to 4 °C over the next century (ACIA

2005). In terrestrial ecosystems, changes in biogeochemical cycling brought about by warming

are likely to alter carbon source/sink dynamics in Arctic tundra and peatlands (e.g. Mack et al.

2004; Shaver et al. 2006), and these will likely have significant implications for future global

warming. The ramifications of Arctic warming have already resulted in the northward expansion

of several species of fish, birds, and mammals, and wreaked havoc with indigenous peoples’

traditional ways of interacting with their environment (ACIA 2005). As noted by Schindler and

Smol (2006), it will be “hard to overestimate the ecological, social, and economic impacts of such

large-scale ecosystem shifts.”

Unfortunately, the Canadian High Arctic is a difficult area to monitor due to logistical

constraints, and the short duration and poor spatial coverage of the very few instrumental

meteorological records. The Canadian High Arctic is currently serviced by only five weather

stations that have been in place since 1947 (Eureka, Ellesmere Island, and Resolute Bay,

Cornwallis Island), 1948 (Isachsen, Ellef Ringnes Island, and Mould Bay, Prince Patrick Island –

now both automated), and 1950 (Alert, Ellesmere Island). More recently established weather

stations or automatic data collecting devices are also present at Grise Fiord, Ellesmere Island and

Rea Point, Melville Island. Moreover, these meteorological records generally post-date the onset

of warming in many High Arctic regions. Comprehensive historical data for other types of

information critical to understanding the impacts of environmental change on ecosystem

functioning (e.g. physical, chemical, and biological data across regions and ecosystems) are

lacking. Thus, reconstructions that aid in deciphering the natural variability of past

environmental conditions, including information on the interactions of climate, biogeochemistry,

and biota, are key to understanding future environmental changes.

The overarching theme of my thesis is to address some of the critical knowledge gaps

surrounding High Arctic environmental change using limnological and paleolimnological

techniques. Each chapter has been prepared as a separate manuscript and thus each contains

introductory material, literature reviews, and study rationales pertinent to the specific study

objective. In order to keep repetition to a minimum, below I provide some broader perspectives

of High Arctic limnology, the use of diatoms as environmental indicators and paleolimnology in a

High Arctic context.

CANADIAN HIGH ARCTIC LIMNOLOGY

An abundance of lakes and ponds characterize the High Arctic landscape. These

freshwater bodies are often important locations of enhanced biodiversity, as they not only provide

habitat for aquatic organisms, but also perform critical ecosystem services for terrestrial biota.

Arctic lakes and ponds are known to be sensitive to environmental change (Douglas et al. 1994;

Schindler & Smol 2006), and several impacts associated with warming Arctic temperatures on

freshwater systems have already been noted. For example, melting of permafrost has been linked

both to the drainage of thermokarst lakes in subarctic Siberia (Smith et al. 2005) and to increases

in major ion transport from land to water (Kokelj et al. 2005). Elevated temperatures have

resulted in the complete dessication of some High Arctic ponds, previously in existence for

millenia (Smol & Douglas 2007). However, in order to determine the extent of environmental

change, a critical first step is to establish current base-line limnological conditions.

Some of the earliest limnological data for the Canadian High Arctic was based on data

collected during “Operation Hazen”, a Defence Research Board of Canada program based near

Lake Hazen, on northern Ellesmere Island (Oliver & Corbet 1966). However, these important

data fall into the “grey literature” category, and thus have not been widely used. In research

circles, most will recognize the data collected during the International Biological Program (IBP)

from the ultra-oligotrophic Char Lake and the sewage-enriched Meretta Lake, Cornwallis Island,

as some of the earliest widely available data from High Arctic lakes (e.g. Schindler et al. 1974;

Kalff & Welch 1974; Kalff 1975). These two relatively large lakes, however, are somewhat

atypical freshwater systems in the High Arctic. Since that time, other aspects of High Arctic

limnology have been explored, such as the relationships between ultra-violet radiation and

dissolved organic carbon (DOC) (Laurion et al. 1997), the role of lake ice in dissolved organic

matter distributions (Belzile et al. 2002); factors related to primary (Vezina & Vincent 1997;

Markager et al. 1999; Bonilla et al. 2005) and secondary (Rautio & Vincent 2006; Rautio &

Vincent 2007) production, and the geochemistry and physics of complex lake systems (Gibson et

al. 2002; Van Hove et al. 2006).

In a concerted effort to gather baseline environmental data from much of the Canadian

High Arctic, John Smol and Marianne Douglas have led a number of limnological surveys of

lakes and ponds (Fig. 1). Since 1983, they, along with their students, have provided physical and

chemical limnological data from eastern-central Ellesmere Island (Douglas & Smol 1994); Alert,

Ellesmere Island (Antoniades et al. 2003a); Axel Heiberg Island (Michelutti et al. 2002b);

Cornwallis Island (Michelutti et al. in press a); Bathurst Island (Lim et al. 2001a); Devon Island

(Lim & Douglas 2003; Lim 2004); Isachsen, Ellef Ringnes Island (Antoniades et al. 2003b);

Banks Island (Lim et al. 2005); Wynniatt Bay, Victoria Island (Michelutti et al. 2002a); and

Mould Bay, Prince Patrick Island (Antoniades et al. 2003a). Additional limnological surveys

have been conducted on Southampton Island (Mallory et al. 2006), and from 204 lakes and ponds

across the Canadian Arctic (Hamilton et al. 2001). These records indicate that High Arctic lakes

and ponds span large gradients of physical and chemical variables, and underscore the need for a

more complete survey of present-day limnological conditions in regions of the High Arctic that

have not yet been fully explored.

Two of the regions for which limnological data are lacking include the western High

Arctic (currently represented by only two relatively small surveys by Antoniades et al. 2003a and

Lim et al. 2005), and from northern Ellesmere Island (a region that encompasses a warm, High

Arctic oasis). Chapters 2 and 4 of this thesis focus on addressing some of the current gaps in

limnological research by evaluating the variability of ponds and lakes across wide environmental

gradients in the western High Arctic (Chapter 2), and from a High Arctic oasis on northern

Ellesmere Island (Chapter 4).

DIATOMS AS ENVIRONMENTAL INDICATORS

Diatoms, siliceous unicellular algae (class Bacillariophyceae), are particularly useful

environmental indicators because they are ubiquitous, they respond rapidly to changing

conditions, and different species often have distinct optima to given environmental variables

(Douglas & Smol 1999). In complementary studies to those described above for basic

limnological data, the Smol and Douglas labs have provided diatom autecological information

from regions throughout the High Arctic (Fig. 1; Douglas & Smol 1993; Lim et al. 2001b; Lim et

al. 2001c; Lim 2004; Antoniades et al. 2004; Antoniades et al. 2005a; Michelutti et al. 2006; Lim

et al. 2007; Michelutti et al. in press a). Additionally, a diatom biogeographical study at the scale

of the Canadian Arctic Archipelago has provided distributional data for Arctic diatoms (Bouchard

et al. 2004). An historical review of diatom research in High Arctic regions, as well as the first

diatom flora produced for the Canadian High Arctic, has recently been provided by Antoniades et

al. (in press).

Several High Arctic diatom-based calibration sets have generated statistically robust

quantitative relationships between diatom species assemblages and the measured environmental

variables, including pH, specific conductivity, dissolved organic carbon (DOC), and/or nutrients,

that can best explain their distributions. The diatom ecological information gleaned from such

calibration studies is important for interpreting the environmental significance of changes in

diatom communities through time. However, just as limnological data are lacking from the

western High Arctic region, diatom distribtutional data was also limited from the western High

Arctic. Thus, in Chapter 3 I explore the role of large environmental and spatial gradients on

diatom distributions from the western High Arctic.

PALEOLIMNOLOGY IN THE CANADIAN HIGH ARCTIC

In the absence of direct observational or instrumental records, paleoenvironmental

techniques can provide important insights into past environments. Several commonly used

paleoenvironmental approaches, such as dendrochronology, are difficult to apply in High Arctic

regions where woody species are rare (but see Rayback & Henry 2006 for one example). Other

powerful paleoenvironmental techniques, such as ice core analyses, are necessarily limited to the

relatively few locations where ice caps that are suitable for analysis exist (e.g. Ellesmere, Devon,

and Meighen islands).

The abundance of lakes and ponds throughout the Canadian High Arctic represent a

wealth of repositories of past environmental conditions (Wolfe & Smith 2004).

Paleolimnological analyses make use of physical, chemical, and/or biological indicators in lake

sediments to reconstruct historical conditions (Smol 2002). In the Canadian Arctic, for example,

variations in the thickness of sedimentary varve records have been used to infer past climatic

conditions (e.g. Lamoureux & Gilbert 2004; Hambley & Lamoureux 2006), whereas spectrally-

inferred chlorophyll a techniques (Michelutti et al. 2005) have been used to track recent changes

in primary production. A variety of biological indicators have also been successfully employed

in the Canadian Arctic to track environmental changes including fossils of algae (e.g. diatoms and

chrysophytes, Douglas et al. 2004a), and invertebrates (Bennike et al. 2004; Quinlan et al. 2005).

Existing observational data (Serreze et al. 2000; ACIA 2004) and proxy records (e.g.

Paterson et al. 1977; Fisher 1979; O’Brien et al. 1995; Zdanowicz et al. 2000; Grumet et al. 2001;

Kaufman et al. 2004; Smol et al. 2005) indicate that the timing and nature of environmental

changes have not been synchronous across the Arctic. To date, diatom-based paleolimnological

investigations of climatic and environmental change from the High Arctic are largely restricted to

the eastern Canadian Arctic (Douglas et al. 1994; Michelutti et al. 2003; Perren et al. 2003; Wolfe

& Smith 2004; Antoniades et al. 2005b) and to one study from the western High Arctic (Banks

Island, Lim et al. 2007). This regional bias precludes our ability to assess the spatial and

temporal variability of climatic change across the High Arctic. To address this gap and in order to

better compare magnitudes and rates of diatom-inferred environmental change from the western

High Arctic with those from the eastern High Arctic, in Chapter 5 I assess diatom-inferred

environmental change from a small pond on Melville Island.

Furthermore, while diatom-based studies have revealed marked shifts in diatom species

assemblages over the last 150 years, the magnitude and timing of such changes have been

different. Typically, large High Arctic lakes tend to be less sensitive to environmental changes,

and therefore often record a much later onset of diatom assemblage changes and more muted

responses, as compared to High Arctic ponds (e.g. Douglas et al. 1994; Michelutti et al. 2003;

Antoniades et al. 2005b). The persistence of ice cover (Smol 1983; Smol 1988) has been invoked

for more than two decades to explain this phenomenon, yet more conclusive evidence is lacking.

In Chapter 6 I examine the influence of ice cover on diatom community changes through time by

comparing sediment records from two adjacent lakes with differing ice cover regimes.

IMPACTS OF SEABIRD-DERIVED NUTRIENTS ON ARCTIC LAKES

The Canadian Arctic is an important habitat for ~10 million marine birds that may be

under threat from changing environmental conditions (Mallory & Fontaine 2004; Gaston et al.

2005). However, the natural variability of seabird populations is unknown, as even the best

seabird monitoring records extend for less than 30 years. Indicators that could reconstruct seabird

populations, therefore, would represent a highly desirable tool for wildlife managers.

Seabird-derived nutrients (e.g. guano, feathers, eggshells) can provide significant

subsidies to otherwise oligotrophic environments (Polis et al. 2004). While much evidence exists

for seabird-related influences on soils, plants, and secondary consumers in arid, sub-tropical

environments (e.g. Polis & Hurd 1996; Polis et al. 1997; Anderson & Polis 1999; Sánchez-Piñero

& Polis 2000; Polis et al. 2004) and Antarctic regions (e.g. Vincent & Vincent 1982; Lindebloom

1984; Ryan & Watkins 1989; Erskine et al. 1998), little data exist for corresponding Arctic

terrestrial environments (Wainwright et al. 1998) or for freshwater ecosystems (Evenset et al.

2004; Blais et al. 2005). In the Canadian High Arctic, only one study (affiliated with this thesis)

has examined the role of seabird-derived nutrients on freshwater quality (Blais et al. 2005).

The water columns of High Arctic lakes are generally nutrient-poor (Schindler & Smol

2006). Although evidence from Arctic Alaska suggests that phosphorus enrichment results in

phytoplankton responses that are analagous to those of temperate regions (i.e. cyanobacterial

blooms, Hobbie et al. 1999), High Arctic lakes do not appear to respond similarly (Schindler et al.

1974). In High Arctic lakes, these responses may be a result of the synergistic effects of extreme

environmental constraints (Markager et al. 1999) and different factors limiting benthic and

phytoplanktonic production (Bonilla et al. 2005). Nevertheless, paleolimnological studies of the

sewage-enriched Meretta Lake, Cornwallis Island, have previously shown that diatom assemblage

changes can track both the onset (Douglas & Smol 2000) and cessation (Michelutti et al. 2002c)

of nutrient enrichment, although the diatom assemblage responses were markedly different and

muted compared to similar studies in temperate regions (Hall & Smol 1999). Similarly, sewage

input only resulted in diatom species shifts after a significant time lag in Annak Lake, Belcher

Islands, Arctic Canada (Michelutti et al. in press b). In a small Arctic pond, changes in diatom

assemblages, in conjunction with changes in stable nitrogen isotopes were linked to nutrient

enrichment from decaying whale carcasses (Douglas et al. 2004b). Guano input from seabirds is

a far more common source of nutrients to Arctic ponds than either of the two point sources noted

above, and yet the influence of birds on diatom ecology is unknown.

Chapters 7 and 8 examine the effects of seabird-derived nutrient enrichment on the water

quality and diatom species assemblages of several small High Arctic ponds (Chapter 7), and

evaluate the use of diatoms, in conjunction with other paleolimnological proxies (i.e. stable

isotopes of nitrogen δ15N (Robinson 2001), and sedimentary-inferred chlorophyll a (Das et al.

2005; Wolfe et al. 2006)), to track these effects through time as a means to assess changes in the

dynamics of a seabird colony at Cape Vera, Devon Island (Chapter 8).

SUMMARY OF THESIS OBJECTIVES

Recent studies have provided important advances in limnological, diatom autecological,

and paleolimnological research that, together, can provide robust reconstructions of

environmental change in the Canadian High Arctic. However, critical knowledge gaps remain,

including understanding the regional variability of environmental change in different ecosystems

across the Canadian High Arctic. This thesis provides new limnological and paleolimnological

data to address some of these research gaps. Furthermore, this thesis assesses the use of diatoms

as indicators of seabird activity, a novel application of diatom-based paleolimnology, in the

Canadian High Arctic.

Chapters 2 and 4 augment the existing High Arctic database of limnological data by

providing data from Melville Island (western High Arctic), and from northern Ellesmere Island

(including a warm High Arctic oasis). Chapter 3 builds upon Chapter 2 by assessing diatom

distributional patterns across large environmental gradients identified on Melville Island. These

data provide important new insights into limnological variability and diatom distributions across

distinctive ecosystem types in the Canadian High Arctic for which little data had previously

existed.

Chapters 5 and 6 are diatom-based paleolimnological investigations of environmental

change. The objective of Chapter 5 is to assess the nature, magnitude, rate, and direction of

diatom-inferred environmental change in a small pond from Melville Island, western High Arctic,

and to compare these with other areas of the High Arctic. The objective of Chapter 6 is to

examine how ice cover influences diatom assemblage changes through time by comparing two

very similar, small High Arctic lakes that experience differing ice cover regimes.

Chapters 7 and 8 involve assessing the influence of seabird-derived nutrients on small

High Arctic ponds. The objective of Chapter 7 is to determine the impacts of seabird-derived

nutrients on freshwater quality and diatom distributions across a gradient of seabird influence.

The objective of Chapter 8 extends the results of Chapter 7 to an historical context, to determine

whether diatom assemblage shifts are able to track the historical trajectory of a seabird colony at

Cape Vera, Devon Island.

REFERENCES ACIA, 2004. Impacts of a warming Arctic. Cambridge University Press, Cambridge, UK. ACIA, 2005. Arctic climate impact assessment. Cambridge University Press, New York. Anderson, W. B., & G. A. Polis, 1999. Nutrient fluxes from water to land: seabirds affect plant nutrient status on Gulf of California islands. Oecologia 118: 324-332. Antoniades, D., M. S. V. Douglas, & J. P. Smol, 2003a. Comparative physical and chemical limnology of two Canadian High Arctic regions: Alert (Ellesmere Island, NU) and Mould Bay (Prince Patrick Island, NWT). Archiv für Hydrobiologie 158: 485-516. Antoniades, D., M. S. V. Douglas, & J. P. Smol, 2003b. The physical and chemical limnology of 24 ponds and one lake from Isachsen, Ellef Ringnes Island, Canadian High Arctic. International Review of Hydrobiology 88: 519-538. Antoniades, D., M. S. V. Douglas, & J. P. Smol, 2004. Diatom species-environment relationships and inference models from Isachsen, Ellef Ringnes Island, Canadian High Arctic. Hydrobiologia 529: 1-18. Antoniades, D., M. S. V. Douglas, & J. P. Smol, 2005a. Benthic diatom autecology and inference model development from the Canadian High Arctic Archipelago. Journal of Phycology 41: 30-45. Antoniades, D., M. S. V. Douglas, & J. P. Smol, 2005b. Quantitative estimates of recent environmental changes in the Canadian High Arctic inferred from diatoms in lake and pond sediments. Journal of Paleolimnology 33: 349-360. Antoniades, D., P. B. Hamilton, M. S. V. Douglas & J. P. Smol, in press. Freshwater diatoms from the Canadian High Arctic. Iconographica Diatomologica. Belzile, C., J. A. E. Gibson, & W. F. Vincent, 2002. Colored dissolved organic matter and dissolved organic carbon exclusion from lake ice: Implications for irradiance transmission and carbon cycling. Limnology and Oceanography 47: 1283-1293. Bennike, O., K. P. Brodersen, E. Jeppesen & I. R. Walker, 2004. Aquatic invertebrates and high latitude paleolimnology. In R. Pienitz, M. S. V. Douglas & J. P. Smol (eds), Long-term environmental change in Arctic and Antarctic lakes. Springer, Dordrecht: 159-186. Blais, J. M., L. E. Kimpe, D. McMahon, B. E. Keatley, M. L. Mallory, M. S. V. Douglas, & J. P. Smol, 2005. Arctic seabirds transport marine-derived contaminants. Science 309: 445. Bonilla, S., V. Villeneuve, & W. F. Vincent, 2005. Benthic and planktonic algal communities in a High Arctic Lake: Pigment structure and contrasting responses to nutrient enrichment. Journal of Phycology 41: 1120-1130. Bouchard, G., K. Gajewski, & P. B. Hamilton, 2004. Freshwater diatom biogeography in the Canadian Arctic Archipelago. Journal of Biogeography 31: 1955-1973. Das, B., R. D. Vinebrooke, A. Sanchez-Azofeifa, B. Rivard, & A. P. Wolfe, 2005. Inferring sedimentary chlorophyll concentrations with reflectance spectroscopy: a novel approach to

reconstructing historical changes in the trophic status of mountain lakes. Canadian Journal of Fisheries and Aquatic Sciences 62: 1067-1078. Douglas, M. S. V., & J. P. Smol, 1993. Fresh-Water Diatoms from High Arctic Ponds (Cape Herschel, Ellesmere-Island, NWT). Nova Hedwigia 57: 511-552. Douglas, M. S. V., & J. P. Smol, 1994. Limnology of high arctic ponds (Cape Herschel, Ellesmere Island, N.W.T.). Archiv für Hydrobiologie 131: 410-434. Douglas, M. S. V., & J. P. Smol, 1999. Freshwater diatoms as indicators of environmental change in the High Arctic. In E. F. Stoermer & J. P. Smol (eds), The Diatoms: Applications for the Environmental and Earth Sciences. Cambridge University Press, Cambridge: 227-244. Douglas, M. S. V., & J. P. Smol, 2000. Eutrophication and recovery in the High Arctic: Meretta lake (Cornwallis Island, Nunavut, Canada) revisited. Hydrobiologia 431: 193-204. Douglas, M. S. V., J. P. Smol, & W. Blake Jr, 1994. Marked post-18th century environmental change in high-arctic ecosystems. Science 266: 416-419. Douglas, M. S. V., P. B. Hamilton, R. Pienitz & J. P. Smol, 2004a. Algal indicators of environmental change in arctic and antarctic lakes and ponds. In R. Pienitz, M. S. V. Douglas & J. P. Smol (eds), Long-term environmental change in Arctic and Antarctic lakes. Springer, Dordrecht: 117-158. Douglas, M. S. V., J. P. Smol, J. M. Savelle, & J. M. Blais, 2004b. Prehistoric Inuit whalers affected arctic freshwater ecosystems. Proceedings of the National Academy of Sciences of the United States of America 101: 1613-1617. Erskine, P. D., D. M. Bergstrom, S. Schmidt, G. R. Stewart, C. E. Tweedie, & J. D. Shaw, 1998. Subantarctic Macquarie Island - a model ecosystem for studying animal-derived nitrogen sources using 15N natural abundance. Oecologia 117: 187-193. Evenset, A., G. N. Christensen, T. Skotvold, E. Fjeld, M. Schlabach, E. Wartena, & D. Gregor, 2004. A comparison of organic contaminants in two high Arctic lake ecosystems, Bjørnøya (Bear Island), Norway. The Science of the Total Environment 318: 125-141. Fisher, D. A., 1979. Comparison of 105 years of oxygen isotope and insoluble impurity profiles from the Devon Island and Camp Century ice cores. Quaternary Research 11: 299-305. Gaston, A. J., H. G. Gilchrist, & M. L. Mallory, 2005. Variation in ice conditions has strong effects on the breeding of marine birds at Prince Leopold Island, Nunavut. Ecography 28: 331-344. Gibson, J. A. E., W. F. Vincent, P. Van Hove, C. Belzile, X. W. Wang, & D. Muir, 2002. Geochemistry of ice-covered, meromictic Lake A in the Canadian High Arctic. Aquatic Geochemistry 8: 97-119. Grumet, N. S., C. P. Wake, P. A. Mayewski, G. A. Zielinski, S. I. Whitlow, R. M. Koerner, D. A. Fisher, & J. M. Woollett, 2001. Variability of sea-ice extent in Baffin Bay over the last millenium. Climatic Change 49: 129-145.

Hall, R. I.& J. P. Smol, 1999. Diatoms as indicators of lake eutrophication. In E. F. Stoermer & J. P. Smol (eds), The Diatoms: Applications for the Environmental and Earth Sciences. Cambridge University Press, New York: 128-168. Hambley, G. W., & S. F. Lamoureux, 2006. Recent summer climate recorded in complex varved sediments, Nicolay Lake, Cornwall Island, Nunavut, Canada. Journal of Paleolimnology 35: 629-640. Hamilton, P. B., K. Gajewski, D. E. Atkinson, & D. R. S. Lean, 2001. Physical and chemical limnology of 204 lakes from the Canadian Arctic Archipelago. Hydrobiologia 457: 133-148. Hobbie, J. E., B. J. Peterson, N. Bettez, L. Deegan, W. J. O'Brien, G.W. Kling, G.W. Kipphut, W. B. Bowden, & A.E. Hershey. 1999. Impact of global change on the biogeochemistry and ecology of an Arctic freshwater system. Polar Research 18: 207-214. IPCC, 2007. Summary for policy-makers. In S. Solomon, D. Qin, M. Manning, Z. Chen, M. Marquis, K. B. Averyt, M. Tignor and H. L. Miller (eds), Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, UK. Kalff, J., 1975. Phytoplankton, phytoplankton growth and biomass cycles in an unpolluted and in a polluted polar lake. Verh. int. Ver. Limnol. 19: 487-495. Kalff, J., & H. E. Welch, 1974. Phytoplankton production in Char Lake, a natural polar lake, and in Meretta Lake, a polluted polar lake, Cornwallis Island, Northwest Territories. Journal of Fisheries Research Board of Canada 31: 621-636. Kaufman, D. S., T. A. Ager, N. J. Anderson, P. M. Anderson, J. T. Andrews, P. J. Bartlein, L. B. Brubaker, L. L. Coats, L. C. Cwynar, M. L. Duvall, A. S. Dyke, M. E. Edwards, W. R. Eisner, K. Gajewski, A. Geirsdóttir, F. S. Hu, A. E. Jennings, M. R. Kaplan, M. W. Kerwin, A. V. Lozhkin, G. M. MacDonald, G. H. Miller, C. J. Mock, W. W. Oswald, B. L. Otto-Bliesner, D. F. Porinchu, K. Rühland, J. P. Smol, E. J. Steig, & B. B. Wolfe, 2004. Holocene thermal maximum in the western Arctic (0-180°W). Quaternary Science Reviews 23: 529-560. Kokelj, S. V., R. E. Jenkins, D. Milburn, C. R. Burn, & N. Snow, 2005. The influence of thermokarst disturbance on the water quality of small upland lakes, Mackenzie Delta Region, Northwest Territories, Canada. Permafrost and Periglacial Processes 16: 343-353. Lamoureux, S. F., & R. Gilbert, 2004. A 750-yr record of autumn snowfall and temperature variability and winter storminess recorded in the varved sediments of Bear Lake, Devon Island, Arctic Canada. Quaternary Research 61: 134-147. Laurion, I., W. F. Vincent, & D. R. S. Lean, 1997. Underwater ultraviolet radiation: Development of spectral models for northern high latitude lakes. Photochemistry and Photobiology 65: 107-114. Lim, D.S.S., 2004. Limnology and diatom paleoecology of diatoms on Banks Island, N.W.T., and Devon Island, Nunavut, Canadian Arctic. University of Toronto. PhD Thesis.

Lim, D. S. S., & M. S. V. Douglas, 2003. Limnological characteristics of 22 lakes and ponds in the Haughton Crater region of Devon Island, Nunavut, Canadian High Arctic. Arctic Antarctic and Alpine Research 35: 509-519. Lim, D. S. S., M. S. V. Douglas, J. P. Smol, & D. A. Lean, 2001a. Physical and chemical limnological characteristics of 38 lakes and ponds on Bathurst Island, Nunavut, Canadian High Arctic. 86:1-22. Lim, D. S. S., M. S. V. Douglas, & J. P. Smol, 2001b. Diatoms and thier relationship to environmental variables from lakes and ponds on Bathurst Island, Nunavut, Canadian High Arctic. Hydrobiologia 450: 215-230. Lim, D. S. S., C. Kwan, & M. S. V. Douglas, 2001c. Periphytic diatom assemblages from Bathurst Island, Nunavut, Canadian High Arctic: An examination of community relationships and habitat preferences. Journal of Phycology 37: 379-392. Lim, D. S. S., M. S. V. Douglas, & J. P. Smol, 2005. Limnology of 46 lakes and ponds on Banks Island, NWT, Canadian Arctic Archipelago. Hydrobiologia 545: 11-32. Lim, D. S. S., J. P. Smol, & M. S. V. Douglas, 2007. Diatom assemblages and their relationships to lakewater nitrogen levels and other limnological variables from 36 lakes and ponds on Banks Island, N.W.T., Canadian Arctic. Hydrobiologia DOI 10.1007/s10750-007-0623-8. Lindebloom, H. J., 1984. The nitrogen pathway in a penguin rookery. Ecology 65: 269-277. Mack, M. C., E. A. G. Schuur, M. S. Bret-Harte, G. R. Shaver, & F. S. Chapin, 2004. Ecosystem carbon storage in arctic tundra reduced by long-term nutrient fertilization. Nature 431: 440-443. Mallory, M. L., & A. J. Fontaine, 2004. Key marine habitat sites for migratory birds in Nunavut and the Northwest Territories. Ottawa, Canadian Wildlife Service Occasional Paper No. 109. Mallory, M. L., A. J. Fontaine, P. A. Smith, M. O. W. Robertson, & H. G. Gilchrist, 2006. Water chemistry of ponds on Southampton Island, Nunavut, Canada: Effects of habitat and ornithogenic inputs. Archiv für Hydrobiologie 166: 411-432. Markager, S., W. F. Vincent, & E. P. Y. Tang, 1999. Carbon fixation by phytoplankton in high Arctic lakes: Implications of low temperature for photosynthesis. Limnology and Oceanography 44: 597-607. Michelutti, N., M. S. V. Douglas, D. R. S. Lean, & J. P. Smol, 2002a. Physical and chemical limnology of 34 ultra-oligotrophic lakes and ponds near Wynniatt Bay, Victoria Island, Arctic Canada. Hydrobiologia 482: 1-13. Michelutti, N., M. S. V. Douglas, D. C. G. Muir, X. Wang, & J. P. Smol, 2002b. Limnological characteristics of 38 lakes and ponds on Axel Heiberg Island, High Arctic Canada. International Review of Hydrobiology 87: 385-399. Michelutti, N., M. S. V. Douglas, & J. P. Smol, 2002c. Tracking recent recovery from eutrophication in a high arctic lake (Meretta Lake, Cornwallis Island, Nunavut, Canada) using fossil diatom assemblages. Journal of Paleolimnology 28: 377-381.

Michelutti, N., M. S. V. Douglas, & J. P. Smol, 2003. Diatom response to recent climatic change in a high arctic lake (Char Lake, Cornwallis Island, Nunavut). Global and Planetary Change 38: 257-271. Michelutti, N., A. P. Wolfe, R. D. Vinebrooke, B. Rivard, & J. P. Briner, 2005. Recent primary production increases in arctic lakes. Geophysical Research Letters 32: L19715-doi: 10.1029/2005GL023693. Michelutti, N., M. S. V. Douglas, & J. P. Smol, 2006. Ecological characteristics of modern diatom assemblages from Axel Heiberg Island (High Arctic Canada) and their application to paleolimnological inference models. Canadian Journal of Botany 84: 1695-1713. Michelutti, N., M. S. V. Douglas, & J. P. Smol, in press a. Evaluating diatom community composition in the absence of marked limnological gradients in the high Arctic: A surface sediment calibration set from Cornwallis Island (Nunavut, Canada). Polar Biology. Michelutti, N., M. H. Hermanson, J. P. Smol, P. J. Dillon, & M. S. V. Douglas, in press b. Delayed response of diatom assemblage changes to sewage inputs in an Arctic lake. Aquatic Sciences. O'Brien, S. R., P. A. Mayewski, L. D. Meeker, D. A. Meese, M. S. Twickler, & S. I. Whitlow, 1995. Complexity of Holocene climate as reconstructed from a Greenland ice core. Science 270: 1962-1964. Oliver, D. R., & P. S. Corbet, 1966. Aquatic habitats in a high arctic locality: the Hazen Camp study area, Ellesmere Island, N.W.T. Department of National Defence, Ottawa. 115 pp. Paterson, W. S. B., R. M. Koerner, D. A. Fisher, S. Johnsen, H. B. Clausen, W. Dansgaard, P. Bucher, & H. Oeschger, 1977. An oxygen-isotope climate record from the Devon Island ice cap, arctic Canada. Nature 266: 508-511. Perren, B. B., R. S. Bradley, & P. Francus, 2003. Rapid lacustrine response to recent High Arctic warming: A diatom record from Sawtooth Lake, Ellesmere Island, Nunavut. Arctic Antarctic and Alpine Research 35: 271-278. Polis, G. A., & S. D. Hurd, 1996. Linking marine and terrestrial food webs: allochthonous input from the ocean supports high secondary productivity on small islands and coastal land communities. The American Naturalist 147: 396-423. Polis, G. A., W. B. Anderson, & R. D. Holt, 1997. Toward an integration of landscape and food web ecology: the dynamics of spatially subsidized food webs. Annual Reviews of Ecological Systems 28: 289-316. Polis, G. A., M. E. Power & G. R. Huxel, 2004. Food Webs at the Landscape Level. University of Chicago Press, Chicago. Quinlan, R., M. S. V. Douglas, & J. P. Smol, 2005. Food web changes in arctic ecosystems related to climate warming. Global Change Biology 11: 1381-1386. Rautio, M., & W. F. Vincent, 2006. Benthic and pelagic food resources for zooplankton in shallow high-latitude lakes and ponds. Freshwater Biology 51: 1038-1052.

Rautio, M., & W. F. Vincent, 2007. Isotopic analysis of the sources of organic carbon for zooplankton in shallow subarctic and arctic waters. Ecography 30: 77-87. Rayback, S. A., & G. H. R. Henry, 2006. Reconstruction of summer temperature for a Canadian High Arctic site from retrospective analysis of the dwarf shrub, Cassiope tetragona. Arctic Antarctic and Alpine Research 38: 228-238. Robinson, D., 2001. δ15N as an integrator of the nitrogen cycle. Trends in Ecology and Evolution 16: 153-162. Ryan, P. G., & B. P. Watkins, 1989. The influence of physical factors and ornithogenic products on plant and arthropod abundance at an inland nunatak group in Antarctica. Polar Biology 10: 151-160. Sánchez-Piñero, F., & G. A. Polis, 2000. Bottom-up dynamics of allochthonous input: direct and indirect effects of seabirds on islands. Ecology 81: 3117-3132. Schindler, D. W., J. Kalff, H. E. Welch, G. J. Brunskil, H. Kling, & N. Kritsch, 1974. Eutrophication in High Arctic - Meretta Lake, Cornwallis-Island (75 Degrees N Lat). Journal of the Fisheries Research Board of Canada 31: 647-662. Schindler, D. W., & J. P. Smol, 2006. Cumulative effects of climate warming and other human activities on freshwaters of Arctic and subarctic North America. Ambio 35: 160-168. Serreze, M. C., J. E. Walsh, F. S. Chapin III, T. Osterkamp, M. Dyurgerov, V. Romanovsky, W. C. Oechel, J. Morison, T. Zhang, & R. G. Barry, 2000. Observational evidence of recent change in the northern high-latitude environment. Climatic Change 46: 159-207. Shaver, G. R., A. E. Giblin, K. J. Nadelhoffer, K. K. Thieler, M. R. Downs, J. A. Laundre, & E. B. Rastetter, 2006. Carbon turnover in Alaskan tundra soils: effects of organic matter quality, temperature, moisture and fertilizer. Journal of Ecology 94: 740-753. Smith, L. C., Y. Sheng, G. M. MacDonald, & L. D. Hinzman, 2005. Disappearing Arctic lakes. Science 308: 1429. Smol, J. P., 1983. Paleophycology of a high arctic lake Cape Herschel, Ellesmere Island. Canadian Journal of Botany 61: 2195-2204. Smol, J. P., 1988. Paleoclimate proxy data from freshwater arctic diatoms. Verh. int. Ver. Limnol. 23: 837-844. Smol, J. P., 2002. Pollution of lakes and rivers: a paleoenvironmental perspective. Arnold, London. Smol, J. P., & M. S. V. Douglas, 2007. Crossing the final ecological threshold in high Arctic ponds. Proceedings of the National Academy of Sciences of the United States of America 104: 12395-12397. Smol, J. P., A. P. Wolfe, H. J. B. Birks, M. S. V. Douglas, V. J. Jones, A. Korhola, R. Pienitz, K. Ruhland, S. Sorvari, D. Antoniades, S. J. Brooks, M. A. Fallu, M. Hughes, B. E. Keatley, T. E. Laing, N. Michelutti, L. Nazarova, M. Nyman, A. M. Paterson, B. Perren, R. Quinlan, M. Rautio,

E. Saulnier-Talbot, S. Siitoneni, N. Solovieva, & J. Weckstrom, 2005. Climate-driven regime shifts in the biological communities of arctic lakes. Proceedings of the National Academy of Sciences of the United States of America 102: 4397-4402. Van Hove, P., C. Belzile, J. A. E. Gibson, & W. F. Vincent, 2006. Coupled landscape-lake evolution in High Arctic Canada. Canadian Journal of Earth Sciences 43: 533-546. Vezina, S., & W. F. Vincent, 1997. Arctic cyanobacteria and limnological properties of their environment: Bylot Island, Northwest Territories, Canada (73 degrees N, 80 degrees W). Polar Biology 17: 523-534. Vincent, W. F., & C. L. Vincent, 1982. Nutritional state of the plankton in Antarctic coastal lakes and the inshore Ross Sea. Polar Biology 1: 159-165. Wainwright, S. C., J. C. Haney, C. Kerr, A. N. Golovkin, & M. V. Flint, 1998. Utilization of nitrogen derived from seabird guano by terrestrial and marine plants at St. Paul, Pribilof Islands, Bering Sea, Alaska. Marine Biology 131: 63-71. Wolfe, A. P., & I. R. Smith, 2004. Paleolimnology of the Middle and High Canadian Arctic. In R. Pienitz, M. S. V. Douglas & J. P. Smol (eds), Long-term Environmental Change in Arctic and Antarctic Lakes. Kluwer Academic Publishers, Dordrecht. 241-268. Wolfe, A. P., R. D. Vinebrooke, N. Michelutti, B. Rivard, & B. Das, 2006. Experimental calibration of lake-sediment spectral reflectance to chlorophyll a concentrations: methodology and paleolimnological validation. Journal of Paleolimnology 36: 91-100. Zdanowicz, C. M., G. A. Zielinski, C. P. Wake, D. A. Fisher, & R. M. Koerner, 2000. A Holocene record of atmospheric dust deposition on the Penny Ice Cap, Baffin Island, Canada. Quaternary Research 53: 62-69.

Figure captions Figure 1. Regional map of the Canadian Arctic Archipelago with the locations of limnological, diatom ecological, and paleolimnological studies conducted by the Smol and Douglas laboratories prior to this thesis. The numbers correspond to the following studies: 1 = Mould Bay, Prince Patrick Island (Antoniades et al. 2003a, 2005a, 2005b); 2 = Isachsen, Ellef Ringnes Island (Antoniades et al. 2003b, 2004, 2005b); 3 = Axel Heiberg Island (Michelutti et al. 2002b, 2006); 4 = Alert, Ellesmere Island (Antoniades et al. 2003a, 2005b); 5 = Cape Herschel, Ellesmere Island (Douglas & Smol 1993, 1995; Douglas et al. 1994), 6 = Devon Island (Lim et al. 2003; Lim 2004); 7 = Cornwallis Island (Douglas & Smol 2000; Michelutti et al. 2002c, 2003, in press a); 8 = Bathurst Island (Lim et al. 2001a, 2001b, 2001c); 9 = Victoria Island (Michelutti et al. 2002a); 10 = Banks Island (Lim et al. 2005, 2007). The letters correspond to regions studied in this thesis: A) Melville Island (Chapters 2, 3, and 5); B) northern Ellesmere Island (Chapters 4 and 6); and C) Cape Vera, Devon Island (Chapters 7 and 8).

80° 80°

75° 75°

70° 70°

65° 65°120°

120°

100°

100°

80°

80°

60°

60°

0 200 400km

1

3

4

6

5

8

2

9

710

A

B

C

Figure 1.

CHAPTER 2

PHYSICAL AND CHEMICAL LIMNOLOGICAL CHARACTERISTICS OF LAKES AND PONDS ACROSS

ENVIRONMENTAL GRADIENTS ON MELVILLE ISLAND, NUNAVUT/N.W.T., HIGH ARCTIC CANADA

BRONWYN E. KEATLEY, MARIANNE S.V. DOUGLAS, AND JOHN P. SMOL

This chapter has been published separately: Keatley, B.E., M.S.V. Douglas, and J.P. Smol. 2007. Physical and chemical limnological characteristics of lakes and ponds across environmental gradients on Melville Island, Nunavut/N.W.T., High Arctic Canada. Fundamental and Applied Limnology. 168: 355-376.

Abstract

Physical and chemical limnological variables were measured from 40 ponds and 6 lakes

across Melville Island, Nunavut/N.W.T., Canadian high Arctic, an environmentally sensitive

region where very limited limnological data were available. Mean values of most variables were

mid-range when compared to other high Arctic limnological surveys, yet the ranges of most

measured variables were amongst the largest yet encountered in Canadian high Arctic regional

surveys. The first two axes of a Principal Components Analysis explained 55.2% of the variation

in the environmental data. Variables most strongly associated with axis one were pH, dissolved

organic carbon, total dissolved nitrogen, specific conductivity and related variables, while axis

two represented gradients of other nutrients and trace metals. High elevation sites near

permanent ice caps recorded the lowest specific conductivity and Ca2+ values yet reported in high

Arctic systems. High phosphorus values (>20 μg/L) in some of the Melville Island sites are

likely indicative of re-suspended sediments, rather than eutrophic conditions. Total nitrogen to

total phosphorus ratios suggest that ~50% of the sites are P limited, while 33% are N limited,

supporting previous research which suggests N limitation is more commonly encountered in

Arctic than in temperate freshwater ecosystems. Finally, when freshwater sites on Melville

Island were grouped according to predefined bioclimatic zones, only the most lushly vegetated

zone appeared to affect limnological conditions, with these sites having higher mean total

dissolved nitrogen, pH, and specific conductivity.

Introduction

High-latitude regions are known to be sensitive to environmental change. For example,

the region north of 60° latitude has warmed by 0.04°C/year compared to 0.025°/year for lower

latitudes over the last 40 years (ACIA 2004). Data from boreholes in Arctic areas, including

Melville Island, suggest that ground surface temperatures have warmed between 1-2°C since the

late 18th century (MAJOROWICZ et al. 2005), while multiproxy records across the Arctic suggest

that the 20th century experienced the warmest temperatures in about 400 years (OVERPECK et al.

1997). These environmental changes have manifested themselves in a variety of ways, including

declining snow cover, thawing permafrost, melting glaciers and sea ice, as well as a series of

limnological changes (e.g. ROUSE et al. 1997, MAGNUSON et al. 2000, ACIA 2004, ANTONIADES

et al. 2005, HINZMAN et al. 2005, SOLOVIEVA et al. 2005).

Limnological changes are pronounced in high Arctic lakes and ponds which are known to

be highly susceptible to environmental change (e.g. DOUGLAS et al. 1994). Indeed, recent

paleolimnological studies from circumpolar regions have highlighted dramatic shifts in algal and

invertebrate community structure, which are consistent with recent climatic warming (SMOL et al.

2005), and increased biological production (MICHELUTTI et al. 2005). Undoubtedly, a warming

climate is already having, and will continue to have, major repercussions on limnological

conditions in Arctic regions. In order to identify whether lakes and ponds are affected by

environmental change, however, it is necessary to document and understand present-day

conditions. Limnological data are scarce in high Arctic regions, as regular monitoring of lakes

and ponds are often hampered by logistical and financial constraints. Nevertheless, these modern

limnological data are important for general monitoring, and are necessary for assessing the

ecological distributions of biological indicators (e.g. KORHOLA et al. 2005), thereby allowing for

scientifically defensible reconstructions of past environmental conditions using paleolimnological

techniques. For example, these data, together with paleoindicators such as diatoms and

cladocerans preserved in lake and pond sediments, can be used as integrative bioindicators of

environmental quality in remote regions such as the Arctic.

Although many Arctic regions remain unexplored, there has been a marked increase in

Arctic limnological studies over the past 10 to 20 years. In the Canadian Arctic Archipelago,

physical and chemical characteristics have been documented for lakes and ponds on Ellesmere

Island (DOUGLAS & SMOL 1994, HAMILTON et al. 2000, ANTONIADES et al. 2003a), Axel

Heiberg Island (MICHELUTTI et al. 2002b), Banks Island (LIM et al. 2005), Bathurst Island (LIM

et al. 2001), Devon Island (LIM & DOUGLAS 2003), Ellef Ringnes Island (ANTONIADES et al.

2003b), Prince Patrick Island (ANTONIADES et al. 2003a), and Victoria Island (MICHELUTTI et al.

2002a). In addition, limnological data from 204 lakes and ponds across the Arctic have also been

published (HAMILTON et al. 2001). These studies indicate that extensive limnological variability

results from the large diversity in geographical and physical settings of high Arctic lakes and

ponds. For example, freshwater bodies throughout the Arctic range from extremely nutrient poor

(e.g. MICHELUTTI et al. 2002a) to eutrophic (e.g. LIM et al. 2005), from extremely dilute to highly

saline (e.g. MICHELUTTI et al. 2002b), and from very acidic, with pH <4 (MICHELUTTI et al.

2002b), to highly alkaline (pH>9, HAMILTON et al. 2001). These wide-ranging conditions

underscore the need for a more complete survey of present-day limnological conditions in

different regions. Despite this recent work, gaps in limnological knowledge persist in many

regions, especially in the western High Arctic, hampering our ability to confidently recognize and

anticipate the effects of climatic changes.

Melville Island is a large island located in the western-central Canadian High Arctic (Fig.

1). The goals of this study are to provide baseline environmental data on the physical and

chemical characteristics of 40 ponds and 6 lakes from across Melville Island, and to determine

which environmental factors most influence the chemical and physical properties of these lakes

and ponds. Our aim is to place the present-day limnology of Melville Island into context with

respect to limnological studies previously completed on high Arctic islands of Canada. Given

that Melville Island encompasses five bioclimatic zones (EDLUND 1994), we also attempt to

distinguish limnological differences between sites from these pre-defined bioclimatic zones.

In order to achieve the above aims, we have divided the analysis of our data into two

inter-related sections. First, as Melville Island represents a large (i.e. 42 149 km2), diverse, but

previously unstudied reference area for polar limnology, we begin by a descriptive analysis of the

physical and chemical properties of the 46 study sites. In this section, we compare our data to

those collected from other Arctic islands, using identical sampling techniques, thus facilitating

comparisons. Second, using principal components analysis (PCA) and other statistical techniques,

we determine which environmental variables best explain the differences in the measured

limnological characteristics.

Site Description

Melville Island (Fig. 1) is the fourth largest island of the Queen Elizabeth Islands

(consisting of Arctic islands north of the Parry Channel) in High Arctic Canada, and straddles the

border between the territory of Nunavut and the Northwest Territories (N.W.T.). As the only

weather records from Melville Island come from the automated data collector at Rea Point and

are only available since 1975, meteorological data is better estimated from the nearest “long-

term” record from the weather station at Mould Bay, Prince Patrick Island (Fig. 1, 76°13’N,

119°19’W), located ~250 km from the centre of Melville Island, where average February and July

temperatures are -34.0°C and 4.0°C respectively, and the mean annual precipitation is 111.0 mm

(METEOROLOGICAL SERVICE OF CANADA 2002). Sea ice reaches its minimum extent around 10

September and by approximately mid-November sea ice cover is generally 100% around Melville

Island (CANADIAN ICE SERVICE 2002). However, the break-up begins first on the southern

coastal areas (McClintock Channel and Viscount Melville Sound) in late July/August, ultimately

reaching between 1% to 60% ice cover, and continues northeast and northwest. The northern

coastal regions typically retain at least 90% ice cover throughout the year (CANADIAN ICE

SERVICE 2002). Prevailing wind patterns are difficult to ascertain due to a lack of data; however,

the unpredictability of wind patterns in Canadian Arctic regions has been noted as an impact of

climate change (ACIA 2004).

Five bioclimatic zones have been previously described for Melville Island (Fig. 2a),

which are delineated mainly by surficial geology and summer climate (EDLUND 1994).

Bioclimatic zone 0 represents completely unvegetated areas and occurs at high elevations. Zone

1 supports low botanical diversity consisting of vascular plants with less than 5% ground cover.

Zone 2 is dominated by herbaceous species and grasses. Zone 3 is dominated by woody plants

and shrubs (Dryas integrifolia, Salix arctica), with a low diversity of sedges (Carex aquatilis var.

stans and Eriophorum sp.) and emergent and aquatic species including Pleuropogon sabinii and

Ranunculus hyperboreus. Zone 4, the most diverse and abundantly vegetated zone (at least 25%

ground cover), is dominated by the same species as found in Zone 3 with the addition of legumes

(Oxytropis sp. and Astragalus alpinus), Rosacea, Asteraceae, Taraxacum sp., and several sedges

and cotton grasses (EDLUND 1994).

The surficial geology throughout Melville Island is largely composed of weathered

bedrock, which is highly variable within regions. In general, the geology of eastern Melville

Island is comprised of sandstones, mudstones, limestones, shales, and evaporites (BARNETT et al.

1977); the western part of the island is dominated by sandstones and carbonates, and the Dundas

Peninsula contains glacial deposits, including some calcareous tills of southern provenance

(TOZER & THORSTEINSSON 1964, HODGSON 1989). Melville Island is the only island in the

western Arctic that currently retains small permanent ice caps (Fig. 2).

Four main physiographic regions have been described on Melville Island: 1) the southern

half of the Dundas Peninsula is characterized by plateaus on horizontal Palaeozoic rocks; 2) the

western peninsula between Murray Inlet and Purchase Bay is characterized by ridges and plateaus

on Tertiary faults; 3) the central and eastern parts of the island are comprised of ridges and

plateaus on folded Palaeozoic rocks; and 4) the Sabine and Sproule peninsulas in the north parts

of the Melville Island are represented by plateaus developed on Upper Paleozoic to early Tertiary

rocks (TOZER & THORSTEINSSON 1964).

There are no permanent settlements on Melville Island, although stone circles found near

Bridport Inlet and Liddon Gulf suggest occupation by Dorset and Pre-Dorset peoples, and

evidence of hunting camps suggests that the Thule visited Melville Island between ca. 500 and

3000 years ago (CHRISTIE 1994).

Although little is known of the limnology of this large area, the anticipation of an Arctic

pipeline project sparked a number of environmental resource surveys, including some preliminary

aquatic analyses, in the 1970s (reviewed in STEWART & BERNIER 1982). Indian and Northern

Affairs Canada also conducted aquatic sampling of 7 lakes from Melville Island in 1981

(STEWART & BERNIER 1982). These surveys provided only limited water chemistry data for two

lakes and concluded that there are few freshwater fish on Melville Island (STEWART & BERNIER

1982).

Methods

Based on our criteria that lakes are > 2 m deep and do not freeze to the bottom in the

winter, while ponds are < 2 m deep and do freeze completely, our 46 sites comprised 6 lakes and

40 ponds. Due to the danger of weak ice on some sites, we could only estimate depth, and thus

our descriptions are largely restricted to classification of a site as either a lake or a pond. A

subset of 14 ponds was located within a radius of ~5 km of Winter Harbour (Fig. 2) and these

were sampled by foot. The remaining 32 sites were sampled via helicopter. All sites were given

alphabetical names preceded by the prefix MV to designate Melville Island (MVA, MVB, etc.).

The sites were chosen to encompass the widest possible environmental gradients (e.g. size,

proximity to the coast, vegetation, geology). We sampled all the lakes and ponds within a short

time frame between 15-24 July 2002, using identical sampling techniques as other limnological

surveys carried out by our labs (e.g. LIM et al. 2001, MICHELUTTI et al. 2002a, MICHELUTTI et al.

2002b, LIM & DOUGLAS 2003, ANTONIADES et al. 2003a, ANTONIADES et al. 2003b, LIM et al.

2005), thereby facilitating comparisons between our previous studies. Detailed methods regarding

field water chemistry sampling are given in Appendix 1.

Latitude, longitude, and elevation measurements were taken using either a handheld

global positioning unit or the helicopter altimeter. Water samples were collected using either

Nalgene® plastic bottles (major ions and trace metals) or 125 mL glass bottles (for phosphorus

and nutrients) following guidelines outlined in Environment Canada (1994). Temperature (temp),

pH, and specific conductivity (cond) were measured from all sites in the field using handheld

thermometers, calibrated handheld Hanna pHep pH meters, and a YSI model 33 conductivity

meter, respectively. All other analyses were performed at the National Water Research Institute

(NWRI) in Burlington, Ontario (Environment Canada), using protocols described in Environment

Canada (1994).

Water was sampled for major cations including calcium (Ca2+), magnesium (Mg2+),

sodium (Na+), and potassium (K+). The major anions measured included chloride (Cl-), and

sulphate (SO42-). Minor ions measured included barium (Ba), lithium (Li), and strontium (Sr).

Metals measured were aluminum (Al), beryllium (Be), cadmium (Cd), chromium (Cr), cobalt

(Co), copper (Cu), iron (Fe), lead (Pb), manganese (Mn), molybdenum (Mo), nickel (Ni),

vanadium (V), zinc (Zn), and silver (Ag). The nutrients and related variables measured included

dissolved silica (SiO2), total phosphorus unfiltered (TP), total phosphorus filtered (TPF), soluble

reactive phosphorus (SRP), nitrogen fractions (nitrate-nitrogen (NO3-N), nitrate-nitrite-nitrogen

(NO3NO2-N), ammonia-nitrogen (NH3-N), total Kjeldahl nitrogen (filtered, TKN), total dissolved

nitrogen (TdN), and particulate nitrogen (PON)), carbon (dissolved inorganic carbon (DIC),

dissolved organic carbon (DOC), particulate organic carbon (POC)), and chlorophyll a (Chlac,

Chla, corrected and uncorrected for phaeophytin, respectively). Details regarding filter size and

type, sample storage, and sample analysis are given in Appendix 1.

To better understand nutrient trends in these aquatic systems, a number of nutrient ratios

were calculated. For example, ratios of POC:Chla (mg) were assessed to help elucidate whether

carbon sources were predominantly allochthonous (>100) or autochthonous (<100, EPPLEY et al.

1977), and were also used to assess general nutrient deficiency trends (μmol:μg) (HECKY et al.

1993). Total N:Total P ratios were calculated to assess whether sites were more likely to be N or

P limited (DOWNING & MCCAULEY 1992, SAKAMOTO 1966, GUILDFORD & HECKY 2000). We

also calculated molar ratios of POC:PON to examine the degree of N-deficiency in lakes (HECKY

et al. 1993). Finally, P deficiencies were further assessed by examining particulate matter molar

ratios of PON:POP and POC:POP (HECKY et al. 1993).

Statistical analyses

Variables that were below the detection limit in >50% of the sites were removed from

further analyses. Some variables were below the analytical detection limit in <50% of sites and

were replaced with half the detection limit for further statistical analyses to avoid zero values and

to provide an estimate of the true value in those sites. Some values of some variables for a few

sites were unavailable due to broken bottles during shipping; these missing values were estimated

by performing a linear regression between two highly correlated variables and using the equation

of the resulting line to derive the missing value. To assess the normality of distributions,

limnological variables were first examined with exploratory graphing techniques using

CALIBRATE version 1.0 (JUGGINS & TER BRAAK 1992). All variables that were not normally-

distributed were normalized by transforming them using either square-root, log x or log x+1

transformations. The transformed data were subsequently used in the following statistical

analyses. A Pearson correlation with Bonferroni-adjusted probabilities was run using the

program Systat version 9.0 to assess pairs of significantly correlated variables. Principal

components analysis (PCA) was executed using CANOCO 4.0 (TER BRAAK & ŠMILAUER 1998),

to determine the main directions of variation in limnological variables in the dataset. Any

variable whose distribution could not be normalized using the aforementioned transformations, as

well as the spatial measurements of elevation, latitude and longitude, were run passively (e.g.

they were plotted onto the PCA after patterns of variation had been derived from the active

variables and thus did not influence the ordination) in the PCA. Finally, to further investigate

differences amongst lakes and ponds located in different bioclimatic zones, t-tests (assuming

unequal variances) were performed using Systat version 9.0.

Results & Discussion

Descriptive analyses

Physical variables

The Melville Island dataset represents the largest elevational gradient yet sampled in high

Arctic limnological surveys, ranging from ~3 m above sea level (asl) (MVAH) to 655 m asl

(MVAE). Other physical gradients covered on Melville Island included a range of latitude from

74°25.14’N (MVQ) to 76°38.66’N (MVAM), and a range of longitude from 106°03.33’W

(MVAQ) to 114°41.53’W (MVAE). Increasing latitude and elevation typically result in harsher

climatic conditions, and thus are expected to play important roles in water chemistry. Despite the

large latitudinal gradient on Melville Island, the complex topography promotes elevation as a

main driver in many of the water quality parameters in this system.

Water temperature during the short sampling window we used (i.e. 15-24 July 2002)

ranged from 2.5°C to 15.5°C, with a mean temperature of 8.8°C (Table 1). The warmest sites,

not surprisingly, were amongst the smallest ponds we sampled (MVP, MVY, MVX, MVS,

MVAD). Water temperature in small Arctic ponds often closely tracks daily air temperature

(DOUGLAS & SMOL 1994), and all these sites were measured over the course of one relatively

warm day. These shallow sites also had relatively lush catchments with mosses and grasses (MV-

P, MV-S, MV-X), and/or had noticeably dense populations of zooplankton, including copepods,

Daphnia, and Anostraca (fairy shrimp). Conversely, the three coolest sites, MVAP, MVAL and

MVAF, were relatively large lakes that were partially ice-covered at the time of sampling, had

little vegetation in their catchments, and were depauperate in zooplankton.

While we did not measure water clarity directly, the turbid nature of many sites during

windy days was noted. Turbidity affects the re-suspension of sediment particles, and thus likely

influences some of our chemical measurements.

Metals

Similar to other high Arctic surveys (e.g. LIM et al. 2001, MICHELUTTI et al. 2002a,

MICHELUTTI et al. 2002b, LIM & DOUGLAS 2003, ANTONIADES et al. 2003a, ANTONIADES et al.

2003b, LIM et al. 2005), many metals were below detection limit in > 50% of the sites (Ag, Be,

Cd, Co, Cr, Mo, Ni, Pb, and V) and are not discussed further. This resulted in a subset of metals

consisting of Al, Ba, Cu, Fe, Li, Mn, Sr, and Zn.

All metal concentrations, except Al and Fe, examined in the Melville Island dataset were

within ranges for natural Canadian waters (MCNEELY et al. 1979). However, mean

concentrations of Al (0.34 mg/L) and Fe (0.49 mg/L) were higher than most other Arctic sites

(Table 1) with similar means and ranges of Al to Ellef Ringnes Island (ANTONIADES et al.

2003b), and Fe with both Ellef Ringnes and Banks islands (ANTONIADES et al. 2003b, LIM et al.

2005). Al was significantly correlated with both Fe and Mn concentrations on Melville Island

(Table 3) and, not surprisingly, the sites with the highest Al typically had high Fe and Mn values.

These sites included two large lakes, MVAO and MVAL, and two small ponds, MVAH and

MVH, all of which were turbid on the day of sampling. In addition to high Fe and Al values,

these sites also had high TP concentrations (see below). Together, this suggests that the re-

suspension of sediment particles is likely the source of elevated metal concentrations in these

sites.

SiO2

Silica concentrations were also within the ranges reported for natural Canadian surface

waters (MCNEELY et al. 1979) with the largest concentrations occurring in the same sites with

very high levels of Al, Fe, Mn, and TP. This likely represents, once again, the presence of re-

suspended material in these sites.

pH and Specific Conductivity

Sites on Melville Island were circumneutral to alkaline in pH, ranging from 6.9 (MVAF)

to 8.8 (MVB), (mean pH = 7.8). This is a relatively small range of pH, given the large geological

gradients covered in this survey, but it is similar to those found on most high Arctic islands (Fig.

3). The pH changes between sites are likely related mainly to local geology, but also possibly to

longer ice cover (PSENNER & SCHMIDT 1992). The lowest pH (6.9) was encountered in a large,

mostly frozen (~80%) lake located at a high elevation which was underlain by sandstone,

siltstone and shale (MVAF). When ice cover persists, photosynthesis is limited, and the

associated CO2 build up can contribute to declining pH (PSENNER & SCHMIDT 1992). The highest

pH (8.8) was from a small pond (MVB) located on calcareous till in the most diverse bioclimatic

zone on Melville Island, and was inhabited by 4 eider ducks on the day of sampling. The

combination of alkaline surficial material, duck guano and relatively high production may have

elevated the pH of this site relative to others in this region.

Specific conductivity in Arctic lakes and ponds is often inversely related to distance from

the coast. On Melville Island, specific conductivity ranged from 5 μS/cm to 1250 μS/cm, with a

mean of 157 μS/cm. This is similar to the average specific conductivities measured from

Bathurst Island, Banks Island, Cape Herschel, and Prince Patrick Island, but is much lower than

those recorded at Axel Heiberg and Ellef Ringnes islands, as well as that recorded at Alert on

Ellesmere Island (Fig. 3., DOUGLAS & SMOL 1994, LIM et al. 2001, MICHELUTTI et al. 2002b,

ANTONIADES et al. 2003a, ANTONIADES et al. 2003b, LIM et al. 2005). Not surprisingly, the

highest conductivities on Melville Island came from coastal ponds that were influenced by sea

spray; they also had the highest concentrations of Na+ and Cl- (MVAI, MVAB). Interestingly,

some sites on Melville Island recorded the lowest conductivities (5 μS/cm) yet reported from the

High Arctic (MVAP, MVAQ, MVAE, MVAF). The lack of catchment vegetation (which would

accelerate the flow of run-off into the lakes and reduce time for solute concentration in run-off),

the influence of ice cover (which would reduce atmospheric input into the water), the relatively

large volumes of water in these sites, and/or the location at high elevations near permanent ice

caps (which would reduce both the size of the catchment area and the influence of marine-derived

aerosols) all contributed to the very low specific conductivities in these sites.

Major and minor ions

While specific conductivity provides summary information about total ionic composition,

the relative contributions of the major cations and anions allow us to more fully understand the

water chemistry of a given system. Mean concentrations (mg/L) of major cations followed the

pattern Na+>Ca2+>Mg2+>K+, while major anions follow Cl->SO42->DIC

(carbonates/bicarbonates). This is likely due to the large influence of coastal sites with very high

Na+ and Cl- concentrations, as well as the influence of sites with very little Ca2+ (see below). To

remove the influence of these few sites, median ion concentrations were also calculated. Median

concentrations give the following cation pattern of Ca2+>Mg2+>Na+>K+, which is similar to cation

concentration patterns found on Victoria (MICHELUTTI et al. 2002a), Devon (LIM & DOUGLAS

2003), and Bathurst (LIM et al. 2001) islands. The median concentrations of anions follow the

pattern of DIC>Cl->SO42-, which is similar to that described from Alert (ANTONIADES et al.

2003a), but a departure from the more typical anion concentration pattern of DIC>SO42->Cl-

found in many other Arctic sites (e.g. LIM et al. 2001, MICHELUTTI et al. 2002a, MICHELUTTI et

al. 2002b, LIM & DOUGLAS 2003). The elevated concentrations of Cl- relative to SO42-

may be

attributed to the highly irregular shoreline of Melville Island; as such, no site is greater than ~40

km from the coast.

The direct influence of the above coastal proximity, as well as the indirect influence of

elevation, are evident in both Cl- and Na+ concentrations, which were significantly (positively)

correlated with each other, as well as significantly (negatively) correlated with elevation (Table

3). Concentrations of Cl- were relatively high compared to other Arctic islands, with a mean

(37.39 mg/L) and a range similar to those found on Ellef Ringnes Island (ANTONIADES et al.

2003b). Sodium concentrations were close to the average for other Arctic islands (mean 19.1

mg/L, range 0.1 mg/L to 297.0 mg/L). Mean Na+ concentrations on Melville Island were lower

than those reported for Axel Heiberg and Ellef Ringnes islands (MICHELUTTI et al. 2002b,

ANTONIADES et al. 2003b), but higher than Bathhurst and Victoria islands (LIM et al. 2001,

MICHELUTTI et al. 2002a). Two small coastal sites had the highest Cl- and Na+ concentrations,

(MVAB and MVAI), while the two ice-covered, high elevation sites (MVAE and MVAF) had

Na+ concentrations below the detection limit, as well as relatively low Cl- concentrations.

The varied surficial and bedrock geology found on Melville Island is likely responsible

for a range of DIC concentrations that was amongst the largest reported for high Arctic islands,

(Fig. 3, mean 12.3 mg/L, median 8.4 mg/L, maximum 47.8 mg/L, minimum 0.2 mg/L), except for

values documented on Axel Heiberg Island (MICHELUTTI et al. 2002b). For example, calcareous

materials are largely restricted to the Dundas Peninsula and are associated with glacial tills

originating from Victoria Island, rather than from local bedrock (HODGSON 1989). Not

surprisingly, the sites located on this glacial till (MVA through MVM) had the highest DIC

concentrations in our suite of lakes and ponds. The lowest DIC sites were located on the central

and eastern part of the Melville Island (MVAJ, MVAP, and MVAQ), an area underlain mainly by

sandstones, shales, and evaporites (BARNETT et al. 1977).

Surficial geology also plays a role in both Ca2+ and SO42- concentrations. For example,

while the mean concentrations of both SO42- and Ca2+ were similar to previously recorded Arctic

values (SO42-: mean 13.8 mg/L, median 1.8 mg/L; Ca2+: mean 15.8 mg/L and median 7.5 mg/L),

the location of pond MVAM near the base of a sulphate and calcium-rich gypsum evaporite dome

on the Sabine Peninsula resulted in anomalously high concentration of SO42- (357 mg/L) and Ca2+

(179 mg/L). A similar trend was noted on Ellef Ringnes Island in ponds close to gypsum

outcrops (ANTONIADES et al. 2003b). Other relatively high Ca2+ values were found in sites with

high DIC, located on calcareous glacial till (see above).

Interestingly, Melville Island also recorded the lowest Ca2+ and Mg2+ concentrations yet

reported from the High Arctic (Fig. 3; Ca2+ only). Low Ca2+ concentrations were also reported by

Stewart & Bernier (1982) in their survey of 7 Melville Island lakes. Once again, high elevation

and ice cover characterised the sites with the lowest Ca2+ and Mg2+ concentrations (MVAE and

MVAF). These sites also had very low specific conductivities, reflecting the dominant role of

Ca2+ in conductivity measurements for sites located away from coastal areas. Similar low Ca2+

values to sites from Melville Island are reported from SubArctic areas in Alaska (GREGORY-

EAVES et al. 2002), and across treeline sites from the Northwest Territories (RÜHLAND & SMOL

1998).

Potassium concentrations (mean 1.5 mg/L, range 0.1 mg/L to 14.1 mg/L) were similar to

the range recorded on Banks Island (LIM et al. 2005). Lim et al. (2005) found that high

concentrations of K+ on Banks Island were likely attributable to higher terrestrial production

evidenced by low Na:K ratios (MCNEELY et al. 1979), and it is possible that the low Na:K (Table

2) ratios of some of the Melville Island sites are related to relatively high plant production.

However, the sites located in the most lushly vegetated zone did not have significantly lower

Na:K ratios than other sites (see below). The highest K+ concentration was recorded in a small

coastal pond (MVAI); as K+ is significantly correlated with Cl-, the influence of sea spray was

likely a driving force in this high potassium concentration.

Nutrients and related variables

In general, Arctic lakes and ponds are quite oligotrophic relative to temperate sites.

However, lakes and ponds surrounded by relatively lush vegetation would be expected to have

higher nutrients (phosphorus and nitrogen) and related variables (organic carbon) due to increased

input from the catchment. The increased availability of nutrients is expected to translate into

higher primary productivity (as measured by chlorophyll a). Many of our sites followed this

expected pattern, with some notable exceptions.

Many nitrogen species were measured from Melville Island. Ammonia (NH3) and nitrite

(NO2) levels were within ranges reported for other Arctic sites, although many lakes and ponds

situated in the High Arctic are typically below the detection limit for these nitrogen species; this

was the case for nitrate-nitrite (NO3-NO2) values from most sites in our dataset. Total Kjeldahl

nitrogen (TKN) is a measure of ammonia and organic nitrogen; TKN values were similar to those

reported from Bathurst, Victoria and Ellef Ringnes islands (LIM et al. 2001, MICHELUTTI et al.

2002a, ANTONIADES et al. 2003b). In summary, total nitrogen values (TN =

PON+TKN+NO3NO2) were calculated and were within the ranges reported from other high

Arctic surveys (Fig. 3). Total dissolved nitrogen (TdN) values were not directly measured for

other high Arctic limnological surveys. Not surprisingly, the sites with the highest TdN, TN, and

TKN values in our dataset (MVL, MVM) came from small ponds located in a very lush area

around Winter Harbour that was characterized by mosses and grasses. These sites also contained

large populations of Daphnia and Anostraca, suggesting that they are relatively productive. As

expected, the lowest TdN values were found in generally high elevation sites MVAE, MVAF,

and MVAJ that also had low DOC. TdN values were significantly correlated with DOC, DIC,

TKN, Ca, Mg, pH, specific conductivity, and (negatively) to elevation (Table 3).

The only nitrogen fraction that did not fit this pattern was PON, which appears to be

more closely related to particulate matter (POC, TP) in the water column than to other nitrogen

parameters. PON concentrations (mean = 0.051 mg/L, median = 0.041 mg/L) were higher than

all other Arctic sites, although Axel Heiberg Island (MICHELUTTI et al. 2002b) was the only other

high Arctic island with >50% of sites above detection limit. The sites with the highest

concentrations of PON (MVAK: 0.296 mg/L, MVAH: 0.122 mg/L, and MVP: 0.210 mg/L) also

had high POC and, except for MVP, high TP values.

Unlike TN or TdN values, the highest TP values were not found in small ponds located in

lushly vegetated sites, but instead in ponds and lakes that likely experience sediment re-

suspension. On Melville Island, TP values were elevated relative to other high Arctic ponds and

lakes (Fig. 3, mean 21.8 μg/L, median 13.4 μg/L). The range of values, 6.2 μg/L to 135.0 μg/L

(Table 1), was most similar to those reported from Banks Island, the most lushly vegetated Arctic

island yet studied (LIM et al. 2005), but lower than those values reported from Ellef Ringnes

Island (ANTONIADES et al. 2003b). The two sites with the highest TP values (MVAR: 135 μg/L

and MVAO: 123 μg/L) had values of TPF, SRP, and TdN that were below their respective means.

Since MVAO was a large, mostly ice-covered lake that was quite turbid, also had high Fe and Al

values (see above), and was located on phosphorus-rich shale, the high TP value is likely a result

of sediment re-suspension; a similar phenomenon was noted and used to explain very high TP

values from sites located on Ellef Ringnes Island (ANTONIADES et al. 2003b). Sites MVAL (73.0

μg/L) and MVAQ (34.5 μg/L) were also large lakes that had little apparent vegetation in their

catchments, yet had fairly high TPF and SRP values. The lowest TP values were from sites MVV

(6.2 μg/L), a large lake with >70% ice cover containing very little vegetation in its catchment;

MVAM (8.8 μg/L), a pond located right by the gypsum outcropping on the Sabine Peninsula; and

MVI (8 μg/L), a small pond characterized by many mosses and grasses in its catchment.

To avoid problems associated with particulate P, total phosphorus filtered (TPF) may be

used. Unfortunately, TPF values suggest that some of our samples may have been contaminated,

as TPF was much higher than TP in some sites (Table 1, MVAC, MVAH, MVAM, MVAN).

When these suspicious sites are removed, mean and median TPF is 10.0 μg/L and 8.6 μg/L,

respectively. These are the highest average values for TPF yet reported in the High Arctic, but

are similar to values reported for two of the most lushly vegetated high Arctic regions yet studied,

Banks Island (LIM et al. 2005) and Prince Patrick Island (ANTONIADES et al. 2003a). As

expected, sites with high TPF concentrations were typically small ponds in lush, grass-moss

meadows (MVE: 32.2 μg/L, MVF: 24.9 μg/L, MVAA: 22.7 μg/L, MVAK: 25.6 μg/L). Site

MVAL (22.4 μg/L) was unusual in this group as it was a very large, mostly ice-covered lake with

little terrestrial vegetation in its catchment. Interestingly, TPF was uncorrelated with any other

variable in our dataset (Table 3).

Like TPF, soluble reactive phosphorus (SRP) values were uncorrelated with any other

variable used in this dataset and were also relatively high; the mean concentration was the highest

yet reported from high Arctic surveys (mean = 2.8 μg/L). The two sites with the highest SRP

values from Melville Island were MVAJ (13.5 μg/L), a site with very little vegetation and

sediment-covered rocks at the bottom, and MVAM (11.3 μg/L), a pond located near a gypsum

dome.

Dissolved organic carbon (DOC) in lakes and ponds is highly related to the presence and

type of vegetation in the catchment. Due to relatively sparse vegetation cover and very harsh

climate (where water and soils are frozen ~10 months of the year), DOC values from high Arctic

lakes and ponds are often much lower than those present in Subarctic sites, where more extensive

vegetation growth, increased leaf litter, and a longer season for run-off act in concert to increase

DOC. On average, DOC values from Melville Island were in the higher range of values reported

for Canadian Arctic islands (Fig. 3, mean 5.45 mg/L and median 4.65 mg/L). Not surprisingly,

these values were much lower than those reported from Subarctic sites located in Alaska

(GREGORY-EAVES et al. 2000), likely reflecting the more restricted summer growing season in

the High Arctic. Some small ponds on Melville Island had quite high DOC concentrations (MVL,

MVM, MVF, and MVAK); these are likely due to their locations in the most lushly vegetated

regions of Melville Island, which would increase DOC in run-off from the catchment. The lowest

DOC concentrations were found in MVAE and MVAF, two high elevation sites characterized by

persistent ice cover, very low conductivities and low TdN values.

Particulate organic carbon (POC) values were within the range reported for other high

Arctic sites (mean 0.503 mg/L, range 0.162 mg/L to 2.710 mg/L). These values were higher than

those recorded on Victoria Island (MICHELUTTI et al. 2002a), approximately the same as ponds on

Axel Heiberg Island (MICHELUTTI et al. 2002b), but lower than Ellef Ringnes and Banks islands

(ANTONIADES et al. 2003b, LIM et al. 2005). The sites with the highest POC values (MVAK,

MVP, MVAH) also had high DOC and TdN values. POC was correlated with both PON and

DOC (Table 3).

Chlorophyll a

Both chlorophyll a uncorrected (Chla) and corrected (Chlac) for phaeophytin were

measured; however, like most high Arctic studies, Chlac was below the detection limit levels in

nearly all sites. This suggests that much of the algal material in these systems was degraded.

Chla had a mean of 1.41 μg/L, a median value of 0.9 μg/L, and a range between <0.1 μg/L to

10.1 μg/L. The mean and median values are similar to most other high Arctic sites, but the

maximum Chla concentration reported was higher than almost all other measured Chla values

except for Banks Island (Fig. 3, LIM et al. 2005). The site of maximum Chla, MVAK, was a very

small, shallow pond (depth = approximately 15 cm) located in the middle of a grassy moss

meadow, and had moss banks extending into the water. An orange algal crust was present on the

sediment, and the cyanobacterium Nostoc was also prevalent. The lowest Chla values came from

relatively large, ice-covered sites that also had low conductivities and low total dissolved nitrogen

(MVAE, MVAF, MVAG, MVAJ, MVY). Chla was significantly correlated with both POC and

PON, which likely suggests their co-variance as particulate matter in the water column.

Interestingly, Chla was uncorrelated with any other variable in the dataset. Typically

Chla is correlated with TP (DILLON & RIGLER 1974) but a lack of correlation between Chla and

either TN or TP has also been found in most other high Arctic lakes and ponds (MICHELUTTI et

al. 2002a, MICHELUTTI et al. 2002b, ANTONIADES et al. 2003a, ANTONIADES et al. 2003b, LIM et

al. 2005), and suggests that neither N or P are controlling planktonic algal production in these

systems. This may be explained in a number of ways. First, most of the aquatic primary

production in the Arctic is from periphyton, especially in shallow systems. As Antoniades et al.

(2003a) highlighted, our Chla values measure concentrations in the water column, and are not

indicative of periphytic production. Secondly, limitation by factors such as temperature may

affect the responses of primary producers such that Arctic systems may not respond to nutrient

enrichment in the same fashion as temperate lakes (FLANAGAN et al. 2003). Finally, recent

studies have shown that N and P concentrations in interstitial waters associated with

cyanobacterial mats in the High Arctic may be many times higher that those measured in the

pelagic zone (VILLENEUVE et al. 2001), and that these mats are not nutrient limited (BONILLA et

al. 2005). Thus, pelagic nutrient concentrations may not reflect what is functionally available to

periphyton.

Nutrient ratios

There are a number of ways to assess the nutrient status of a lake system, with TP

concentration being the most commonly used. Classification of trophic status based solely on TP

(WETZEL 1983) suggests that 22% of our sites were oligotrophic (<10 μg/L), 67% were

mesotrophic (10-30 μg/L), 6% were eutrophic (30-100 μg/L) and 4% were hypereutrophic (>100

μg/L). However, this TP-based classification may be misleading. In some high Arctic shallow

ponds, high TP values recorded in water samples taken near the shoreline (where we do our

sampling) may reflect the presence of re-suspended sediment particles from high wind and wave

action, and do not necessarily indicate eutrophic systems (e.g. ANTONIADES et al. 2003b). Thus,

other methods of assessing nutrient availability (or deficiency) might be preferable. An

examination of particulate organic carbon (POC) versus Chla, for example, could indicate general

nutrient deprivation (HECKY et al. 1993). On Melville Island, POC:Chla ratios suggest that 96%

are severely nutrient deprived, while the other 4% are moderately nutrient deprived (Table 2). Of

course, some caution must be used when interpreting these ratios, as the input of terrestrial

sources of C, especially in shallow ponds, would inflate our POC values. Indeed, when we

examine the POC:Chla (by weight) ratio, we observed that 96% of the lakes are likely receiving

most of their carbon from allochthonous sources (EPPLEY et al. 1977).

In addition to understanding general nutrient deficiency, it is useful to know whether

systems are generally N or P limited, especially as the input of anthropogenic N becomes of

increasing concern (e.g. VITOUSEK et al. 1997, WOLFE et al. 2001). Sakamoto (1966) suggested

that lakes with TN:TP (μg/L) >17:1 would be P limited, where sites with ratios of <14:1 would

more likely exhibit N-limitation (DOWNING & MCCAULEY 1992). When examined this way, 56%

of sites can be considered P deficient, 35% could be considered N deficient and the remaining 9%

might be limited by either nutrient. Examining TN:TP ratios using molar concentrations indicates

that 47% of our sites are P-deficient, 33% are N-deficient, and 20% could be limited by either

nutrient (GUILDFORD & HECKY 2000, Table 2). Regardless of how we calculate these TN:TP

ratios, they suggest that, while the majority of our sites are P limited, many are N limited.

Further examination of nutrient ratios from particulate matter can provide a better

understanding of nutrient limitation in these systems. For example, POC:PON (molar ratios)

indicate the degree of N-deficiency, where values < 8.3 indicate no N-deficiency, values between

8.3 and 14.6 indicate moderate N deficiency and values > 14.6 suggest severe N deficiency

(HECKY et al. 1993). According to these criteria, all of our sites are moderately N-deficient

except for MVM, which is classified as severely N-deficient (Table 2).

Likewise, P limitation can be examined by calculating molar ratios of PON:POP and

POC:POP (HECKY et al. 1993). We have restricted our calculation of these ratios to 37 sites, due

to suspect TPF values from 9 sites (see above). PON:POP ratios < 23 indicate no P limitation,

while ratios > 23 indicate severe P limitation (HECKY et al. 1993). According to these criteria, 12

sites are P deficient while 25 sites are not P deficient. POC:POP ratios < 133 indicate no P

limitation, while those >133 suggest moderate to severe P limitation (HECKY et al. 1993).

According to POC:POP ratios, 21 sites are P deficient while 16 are not (Table 2).

Taken together, these data suggest that, while all our sites may be considered N deficient,

not all are considered P deficient. These data support previous findings from Alaskan lakes

which suggest that N deficiency might be more common in Arctic lake systems than in temperate

systems (LEVINE & WHALEN 2001). In the Canadian high Arctic, Bathurst Island and Ellef

Ringnes Island were the only two regions studied thus far where the majority of sites are thought

to have been limited by N rather than P, based on TN:TP (by weight) (LIM et al. 2001,

ANTONIADES et al. 2003b). Furthermore, POC:Chla ratios indicate that all our sites are at least

moderately nutrient deprived. These findings corroborate classification of trophic status based on

Chla, but not TP concentrations alone, suggesting that, despite high TP values from some sites on

Melville Island, it is unlikely that any of these systems could be genuinely considered eutrophic.

Statistical analyses

PCA summarizes the main gradients of environmental variables into linear components.

Many of our parameters were highly correlated (Table 3), and thus we selected representative

variables to include in the PCA in an attempt to make the results of the PCA less complex. For

physical and chemical variables from Melville Island, the first two PCA axes explained 55.2% of

the variance in the environmental data. The first axis explained 31.5% of the variance and

represented gradients of pH, specific conductivity and related variables (Ca, Na, K, Cl, Mg, DIC),

TdN, DOC, and elevation (Fig. 4). Axis 2 explained 23.7% of the variance in the environmental

data and was mainly influenced by variables related to nutrients, including Chl a, TP, TPF, and

the metals Mn and SiO2, (Fig. 4).

The location of lakes and ponds on the PCA biplot indicates their relationship to multiple

gradients. Sites did not sort according to any pattern when split into physiographic regions (data

not shown). This is not surprising given that geology, rather than physiography, is known to be

an important factor influencing water chemistry elsewhere in the Arctic. Interestingly, when we

sorted sites based on bedrock geology (Fig. 4a), patterns were also difficult to distinguish,

possibly due to the predominance of sites located on carbonate and sandstone bedrock. When

sites are distinguished based on bioclimatic zones (Fig. 4b), only the sites located in the most

botanically diverse and abundant zone (zone 4) tended to group together in the lower left

quadrant of the ordination biplot. The sites located in bioclimatic zones 3, 2, and 1 are widely

spread throughout the PCA biplot, suggesting that, except for bioclimatic zone 4, variations

amongst these zones are not reflected in differences in water chemistry.

Bioclimatic zone 4 is characterized by shrubs such as Salix arctica and Dryas

integrifolia, several species of sedges and cotton grasses, Rosaceae, Asteraceae, Taraxacum spp.,

legumes (e.g. Oxytropis spp. and Astragalus alpinus), as well as several emergent and aquatic

species (EDLUND 1994). Together, the sites from bioclimatic zone 4 have significantly higher

pH, DOC, DIC, TdN, and TN:TP (Fig. 5, t-test, p < 0.05) than the rest of the sites combined.

When the sites from zone 4 are tested relative to those in individual zones, TdN remains

significantly higher in zone 4 than in all other zones except for zone 3. Interestingly, these sites

are not characterized by higher concentrations of other nutrients and associated variables (TP,

Chla, SRP, TPF, PON). In fact, the other zones combined had significantly higher TP values

than zone 4 (Fig. 5, t-test, p < 0.05).

The higher DIC and pH values may be explained by the location of most of these sites

near Winter Harbour on calcareous glacial till. The abundance of vegetation in this area, which

would contribute terrestrial organic C and would slow the percolation of run-off through the

catchment, is likely to have caused increased DOC in these sites. Both Antoniades et al. (2003a)

and Lim et al. (2005) suggested that the presence of vegetation near some of their sites was likely

responsible for higher DOC, TP, and TN values. On Melville Island, however, both TdN and

DOC are separated from both TP and Chla in the PCA biplot. This is puzzling, and may be

related to both the greater abundance of vegetation as well as the presence of leguminous plants

only in bioclimatic zone 4. Indeed, differences in vegetation type have been known to affect

fluxes of DOC and dissolved organic N in Arctic soils more so than climatic differences (NEFF &

HOOPER 2002). The presence of nitrogen-fixing Alnus near lakes in Alaska has been known to

increase nitrogen concentrations in lakewater due to both the terrestrial decomposition of N-

enriched plant matter as well as through input of N-rich leaves to the lake (GOLDMAN 1960,

DUGDALE & DUGDALE 1961). A similar phenomenon may be occurring in these zone 4 sites (the

only sites in which legumes are present) on Melville Island, albeit on a much smaller scale. Thus,

the higher DOC and TdN concentrations in these sites likely reflects the abundance, and possibly

the type, of vegetation in relatively lush areas.

Conclusions

The 46 ponds and lakes on Melville Island represent wide-ranging limnological

conditions. When examined as mean or median values, most of the measured environmental

variables are within the ranges previously reported for other high Arctic lakes and ponds.

Individual sites, however, recorded the lowest specific conductivities and Ca2+ concentrations yet

reported from the Canadian high Arctic. These sites occur at high elevations and are in proximity

to the small permanent ice caps on Melville Island. The influence of surficial geology was

apparent in both the locally high Ca2+ and SO42- concentrations occurring in sites located on

calcareous glacial till, and near a gypsum outcrop, respectively.

Nutrient measurements indicate some very high TP values on Melville Island. Most of

these can be explained by the influence of re-suspended sediments, and are not likely

representative of nutrient-rich conditions. Likewise, Chla concentrations suggest that most sites

are oligotrophic. Various particulate nutrient ratios from sites on Melville Island suggest that all

the ponds and lakes experience both general nutrient deficiency, as well as nitrogen deficiency.

TN:TP molar ratios suggest that the majority of sites are P deficient (47%), while 33% could be

considered N deficient.

PCA shows that 55.2 % of the variation in the environmental data can be explained by

the first two axes. Axis one explains 31.5 % of the variance and represents primarily pH,

elevation, DOC, TdN, and conductivity and related variables, while axis two explains 23.7 % of

the variance and represents a gradient of other nutrients, and trace metals. Ponds and lakes

occurring in different bioclimatic zones did not fall along any recognizable patterns of the PCA

axes, except for those occurring in the most lushly vegetated zone. This likely reflects the limited

influence of vegetation type on limnological characteristics in most high Arctic bioclimatic zones.

The sites located in bioclimatic zone 4 had significantly higher DOC and TdN than all other lakes

and ponds. This may reflect the influence of both higher vegetation cover, as well as the presence

of legumes in proximity to the ponds.

The physical and chemical characteristics of lakes and ponds from Melville Island may

have implications for both their resident biota and in the way these lakes will respond to climate

change. For example, DOC is known to act as a UV filter that may have important effects for

heterotrophic bacteria, cyanobacteria, algae and zooplankton in these systems (e.g. RAE &

VINCENT 1998; RAUTIO & KORHOLA 2002). Furthermore, as the Arctic continues to warm as a

result of climate change, the catchments of lakes and ponds will likely become increasingly

vegetated with concomitant changes in water chemistry (e.g. increased DOC and nutrient

cycling). These effects may be most pronounced in regions within the high Arctic which, at

present, are especially depauperate in terrestrial vegetation. Likewise, the reduction of seasonal

ice cover within lakes will have major implications for pH (e.g. PSENNER & SCHMIDT 1992),

conductivity (e.g. DOUGLAS & SMOL 1999), and light regimes within lakes, which will likely

have associated effects on aquatic biota.

Acknowledgements

This project was funded through Natural Sciences and Engineering Research Council (NSERC)

grants to BEK, JPS and MSVD, and a Northern Scientific Training Program grant to BEK. We

are very grateful to X. Wang and D. Muir for water chemistry analyses at the Canadian Centre for

Inland Waters, Environment Canada. The Polar Continental Shelf Project (PCSP) provided

logistical support. Many thanks to J.R. Glew, D. Antoniades, and N. Michelutti for help in the

field, as well as to D. Antoniades, N. Michelutti, and D.S.S. Lim for graciously providing their

data. Finally, thank you to D. Antoniades, A. Poulain, K. Rühland, D. Selbie and anonymous

reviewers for helpful comments on the manuscript. This PCSP/ÉPCP contribution # 032-06.

References

ACIA. (2004): Impacts of a warming Arctic. Arctic Climate Impact Assessment. –

Cambridge University Press, Cambridge, UK. 146 pp. ANTONIADES, D., DOUGLAS, M.S.V., & SMOL, J.P. (2003a): Comparative physical and

chemical limnology of two Canadian High Arctic regions: Alert (Ellesmere Island, NU) and Mould Bay (Prince Patrick Island, NWT). – Arch. Hydrobiol. 158: 485 – 516.

ANTONIADES, D., DOUGLAS, M.S.V., & SMOL, J.P. (2003b): The physical and chemical limnology of 24 ponds and one lake from Isachsen, Ellef Ringnes Island, Canadian High Arctic. – Internat. Rev. Hydrobiol. 88: 519 – 538.

ANTONIADES, D., DOUGLAS, M.S.V., & SMOL, J.P. (2005): Quantitative estimates of recent environmental changes in the Canadian High Arctic inferred from diatoms in lake and pond sediments. – J. Paleolimnol. 33: 349 – 360.

BARNETT, D.M., EDLUND, S.A., & DREDGE, L.A. (1977): Terrain characterization and evaluation: an example from eastern Melville Island. – Geol. Surv. Can. Paper 76-23. 18 pp.

BONILLA, S., VILLENEUVE, V., & VINCENT, W.F. (2005): Benthic and planktonic algal communities in a High Arctic lake: pigment structure and contrasting responses to nutrient enrichment. – J. Phycol. 41: 1120 – 1130.

CANADIAN ICE SERVICE (2002). Sea ice climatic atlas: northern Canadian waters 1971-2000. Minister of Public Works and Government Services Canada, Ottawa. 266 pp.

CHRISTIE, R.L. (1994): Introduction. – In: CHRISTIE, R.L., & MCMILLAN, N.J. (eds.): The Geology of Melville Island, Arctic Canada. – Geol. Surv. Can. Bull. 450: pp. 7 – 11.

DILLON, P.J., & RIGLER, F.H. (1974): Phosphorus-chlorophyll relationships in lakes. – Limnol. Oceanogr. 19: 767 – 773.

DOUGLAS, M.S.V., & SMOL, J.P. (1994): Limnology of high Arctic ponds (Cape Herschel, Ellesmere Island, N.W.T.). – Arch. Hydrobiol. 131: 410 – 434.

DOUGLAS, M.S.V., SMOL, J.P., & BLAKE JR., W. (1994): Marked post-18th century environmental change in high-Arctic ecosystems. – Science. 266: 416 – 419.

DOWNING, J.A., & MCCAULEY, E. (1992): The nitrogen-phosphorus relationship in lakes. – Limnol. Oceanogr. 37: 936 – 945.

DUGDALE, R.C., & DUGDALE, V.A. (1961): Sources of phosphorus and nitrogen for lakes on Afognak Island. – Limnol. Oceanogr. 6: 13 – 23.

EDLUND, S.A. (1994): The distribution of plant communities on Melville Island, Arctic Canada. – In: CHRISTIE, R.L., & MCMILLAN, N.J. (eds.): The Geology of Melville Island, Arctic Canada. – Geol. Surv. Can. Bull. 450: pp. 247 – 255.

ENVIRONMENT CANADA (1994): Manual of Analytical Methods – National Laboratory for Environmental Testing, Canadian Centre for Inland Waters. Burlington, Ontario, 2 vols., 1097 pp.

EPPLEY, R.W., HARRISON, W.G., CHISHOLM, S.W., & STEWART, E. (1977): Particulate organic matter in surface waters off Southern California and its relationship to phytoplankton. – J. Mar. Res. 35: 671 – 696.

FLANAGAN, K.M., MCCAULEY, E., WRONA, F., & PROWSE, T. (2003): Climate change: the potential for latitudinal effects on algal biomass in aquatic ecosystems. – Can. J. Fish. Aquat. Sci. 60: 635 – 639.

GOLDMAN, C.R. (1960): Primary productivity and limiting factors in 3 lakes of the Alaska Peninsula. – Ecol. Monogr. 30: 193 – 223.

GREGORY-EAVES, I., SMOL, J.P., FINNEY, B.P., LEAN, D.R.S., & EDWARDS, M.E. (2000): Characteristics and variation in lakes along a north-south transect in Alaska. –

Archiv. Hydrobiol. 147: 193 – 223. GUILDFORD, S.J., & HECKY, R.E. (2000): Total nitrogen, total phosphorus, and nutrient

limitation in lakes and oceans: is there a common relationship? – Limnol. Oceanogr. 45: 1213 – 1223.

HAMILTON, P.B., GAJEWSKI, K., MCNEELY, R.N., & LEAN, D.R.S. (2000): Physical, chemical, and biological characteristics of lakes from the Slidre River basin on Fosheim Peninsula, Ellesmere Island, Nunavut. – In: Garneau, M., & Alt, B.T. (eds.): Environmental response to climate change in the Canadian High Arctic. – Geol. Surv. Can. Bull. 539: 235 – 248.

HAMILTON, P.B., GAJEWSKI, K., ATKINSON, D.E., & LEAN, D.R.S. (2001): Physical and chemical limnology of 204 lakes from the Canadian Arctic Archipelago. – Hydrobiologia. 457: 133 – 148.

HARRISON, J.C. (1994): Melville Island and adjacent smaller islands, Canadian Arctic Archipelago, District of Franklin, Northwest Territories. Map 1844A, Scale 1:250 000. Geological Survey of Canada: Ottawa.

HECKY, R.E., CAMPBELL, P., & HENDZEL, L.L. (1993): The stoichiometry of carbon, nitrogen, and phosphorus in particulate matter of lakes and oceans. – Limnol. Oceanogr. 38: 709 – 724.

HINZMAN, L.D., BETTEZ, N.D., BOLTON, W.R., CHAPIN, F.S., DYURGEROV, M.B., FASTIE, C.L., GRIFFITH, B., HOLLISTER, R.D., HOPE, A., HUNTINGTON, H.P., JENSEN, A.M., JIA, G.J., JORGENSON, T., KANE, D.L., KLEIN, D.R., KOFINAS, G., LYNCH, A.H., LLOYD, A.H., MCGUIRE, A.D., NELSON, F.E., OECHEL, W.C., OSTERKAMP, T.E., RACINE, C.H., ROMANOVSKY, V.E., STONE, R.S., STOW, D.A., STURM, M., TWEEDIE, C.E., VOURLITIS, G.L., WALKER, M.D., WALKER, D.A., WEBBER, P.J., WELKER, J.M., WINKER, K., & YOSHIKAWA, K. (2005): Evidence and implications of recent climate change in northern Alaska and other Arctic regions. – Clim. Change. 72: 251 – 298.

HODGSON, D.A. (1989): Surficial materials (Queen Elizabeth Islands). – In: FULTON, R.J. (ed.): Quaternary Geology of Canada and Greenland. – Geol. Surv. Can. pp. 448-452.

JUGGINS, S., & TER BRAAK, C.J.F. (1992): CALIBRATE – a program for species- environment calibration by [weighted averaging] partial least squares regression. – University College, London.

KORHOLA, A., TIKKANEN, M., & WECKSTROM, J. (2005): Quantification of Holocene lake-level changes in Finnish Lapland using a cladocera-lake depth transfer model. – J. Palelimnol. 34: 175 – 190.

LEVINE, M.A., & WHALEN, S.C. (2001): Nutrient limitation of phytoplankton production in Alaskan Arctic foothill lakes. – Hydrobiologia. 455: 189 – 201.

LIM, D.S.S., & DOUGLAS, M.S.V. (2003): Limnological characteristics of 22 lakes and ponds in the Haughton Crater region of Devon Island, Nunavut, Canadian High Arctic. – Arct. Antarct. Alp. Res. 35: 509 – 519.

LIM, D.S.S., DOUGLAS, M.S.V., & SMOL, J.P. (2005): Limnology of 46 lakes and ponds on Banks Island, NWT, Canadian Arctic Archipelago. – Hydrobiologia. 545: 11 – 32.

LIM, D.S.S., DOUGLAS, M.S.V., SMOL, J.P., & LEAN, D.R.S. (2001): Physical and chemical limnological characteristics of 38 lakes and ponds on Bathurst Island, Nunavut, Canadian High Arctic. – Internat. Rev. Hydrobiol. 86: 1 – 22.

MAGNUSON, J.J., ROBERTSON, D.M., BENSON, B.J., WYNNE, R.H., LIVINGSTON, D.M., ARAI, T., ASSEL, R.A., BARRY, R.G., CARD, V., KUUSISTO, E., GRANIN, N.G., PROWSE, T.D., STEWART, K.M., & VUGLINSKI, V.S. (2000): Historical trends in lake and river ice cover in the Northern Hemisphere. – Science. 289: 1743 – 1746.

MAJOROWICZ, J.A., SKINNER, W.R., & SAFANDA, J. (2005): Ground surface warming

history in northern Canada inferred from inversions of temperature logs and comparison with other proxy climate reconstructions. – Pure and Applied Geophysics. 162: 109 – 128.

MCNEELY, R.N., NEIMANIS, V.P., & DWYER, L. (1979): Water Quality Sourcebook: a guide to water quality parameters. Minister of Supply and Services Canada. Ottawa, Ontario, 88 pp.

METEOROLOGICAL SERVICE OF CANADA. (2002): Canadian Climate Normals 1971 – 2000. Environment Canada, Ottawa.

MICHELUTTI, N., DOUGLAS, M. S. V., LEAN, D. R. S. & SMOL, J. P. (2002a): Physical and chemical limnology of 34 ultra-oligotrophic lakes and ponds near Wynniatt Bay, Victoria Island, Arctic Canada. – Hydrobiologia 482: 1–13.

MICHELUTTI, N., DOUGLAS, M. S. V., MUIR, D.C.G., WANG, X., & SMOL, J. P. (2002b). Limnological characteristics of 38 lakes and ponds on Axel Heiberg Island, High Arctic Canada. – Internat. Rev. Hydrobiol. 87: 385 –399.

MICHELUTTI, N., WOLFE, A.P., VINEBROOKE, R.D., RIVARD, B., & BRINER, J.P. (2005): Recent primary production increases in Arctic lakes. – Geophysical Research Letters. 32: L19715 – doi: 10.1029/2005GL023693.

NEFF, J.C., & HOOPER, D.U. (2002): Vegetation and climate controls on potential CO2, DOC and DON production in northern latitude soils. – Global Change Biology. 8: 872 – 884.

OVERPECK, J. J., HUGHEN, K., HARDY, D., BRADLEY, R., CASE, R., DOUGLAS, M., FINNEY, B., GAJEWSKI, K., JACOBY, G., JENNINGS, A., LAMOUREUX, S., LASCA, A., MACDONALD, G., MOORE, J., RETELLE, M., SMITH, S., WOLFE, A., & ZIELINSKI, G. (1997): Arctic environmental change of the last four centuries. – Science. 278: 1251 – 1256.

PSENNER, R., & R. SCHMIDT. (1992): Climate-driven pH control of remote alpine lakes and effects of acid deposition. – Nature. 356: 781-783.

RAE, R., & W. F. VINCENT. (1998): Effects of temperature and ultraviolet radiation on microbial foodweb structure: potential responses to global change. – Fresh. Biol. 40: 747-758.

RAUTIO, M. & KORHOLA, A. (2002): Effects of ultraviolet radiation and dissolved organic carbon on the survival of subarctic zooplankton. – Polar Biology. 25: 460-468.

ROUSE, W. R., DOUGLAS, M. S. V., HECKY, R. E., HERSHEY, A. E., KLING, G. W., LESACK. L., MARSH, P., MCDONALD, M., NICHOLSON, B. J., ROULET, N. T. & SMOL, J. P. (1997): Effects of climate change on the waters of Arctic and SubArctic North America. – Hydrological Processes 11: 973 –902.

RÜHLAND, K., & SMOL, J.P. (1998): Limnological characteristics of 70 lakes spanning Arctic treeline from Coronation Gulf to Great Slave Lake in the Central Northwest Territories, Canada. – Internat. Rev. Hydrobiol. 83: 183 – 203.

SAKAMOTO, M. (1966): Primary productivity by phytoplankton communities in some Japanese lakes and its dependence on lake depth. – Arch. Hydrobiol. 62: 1 - 28.

SMOL, J.P., WOLFE, A.P., BIRKS, H.J.B., DOUGLAS, M.S.V., JONES, V.J., KORHOLA, A., PIENITZ, R., RUHLAND, K., SORVARI, S., ANTONIADES, D., BROOKS, S.J., FALLU, M.A., HUGHES, M., KEATLEY, B.E., LAING, T.E., MICHELUTTI, N., NAZAROVA, L., NYMAN, M., PATERSON, A.M., PERREN, B., QUINLAN, R., RAUTIO, M., SAULNIER-TALBOT, E., SIITONENI, S., SOLOVIEVA, N., & WECKSTROM, J. (2005): Climate-driven regime shifts in the biological communities of Arctic lakes. - Proc. Natl. Acad. Sci. U.S.A. 102: 4397 - 4402.

SOLOVIEVA, N., JONES, V.J., NAZAROVA, L., BROOKS, S.J., BIRKS, H.J.B., GRYTNES, J.A., APPLEBY, P.G., KAUPPILA, T., KONDRATENOK, B., RENBERG, I., & PONOMAREV, V.

(2005): Palaeolimnological evidence for recent climatic change in lakes from the northern Urals, Arctic Russia. – J. Paleolimnol. 33: 463 - 482.

STEWART, D.B. & BERNIER, L.M.J. (1982): An Aquatic Resource Survey of Islands Bordering Viscount Melville Sound. Land Use Information Series - Environment Canada Lands Directorate, Background Report. No. 2: 110 pp.

TER BRAAK, C. J. F. & ŠMILAUER, P. (1998): CANOCO and CANOPost, Version 4 for Windows. – Microcomputer Power: Ithaca, N.Y., USA.

TOZER, E.T. & THORSTEINSSON, R. (1964): Western Queen Elizabeth Islands, Arctic Archipelago – Geol. Surv. Can. Memoir 332: 242 pp.

VILLENEUVE, V., VINCENT, W.F., & KOMÁREK, J. (2001): Community structure and microhabitat characteristics of cyanobacterial mats in an extreme high Arctic environment: Ward Hunt Lake. - Nova Hedwigia 123: 199 - 224.

VITOUSEK, P.M., MOONEY, H.A., LUBCHENCO, J., & MELILLO, J.M. (1997): Human domination of Earth's ecosystems. - Science 277: 494 - 499.

WETZEL, R.G. (1983): Limnology. 2nd ed. – Saunders Publishing, Philadelphia, PA, 767 pp.

WOLFE, A.P., BARON, J.S., & CORNETT, R.J. (2001): Anthropogenic nitrogen deposition induces rapid ecological changes in alpine lakes of the Colorado Front Range (USA). – J. Paleolimnol. 25: 1 - 7.

Figure Captions Figure 1. Regional map of the Canadian Arctic Archipelago with the locations of both Melville Island and previous modern limnological studies to which references are made in the text. The inset map A) indicates the location of the Canadian High Arctic within Canada, and the circle on the main map B) indicates Melville Island, the focus of our study. Figure 2. Map of Melville Island, with sites differentiated according to previously defined a) bioclimatic zones (after EDLUND 1994), and b) bedrock geology (after HARRISON 1994). On both maps, numbers indicate the following geographical features: 1) Liddon Gulf, 2) Murray Inlet, 3) Purchase Bay, 4) Leopold Glacier, 5) unnamed ice caps, 6) Bridport Inlet, 7) Sabine Bay, 8) Hecla and Griper Bay. 2a) Bioclimatic zone 1 is the most sparsely vegetated region, while zone 4 has the greatest number and abundance of terrestrial vegetation, including woody shrubs (EDLUND 1994). Figure 3. Box plots showing variability of selected environmental variables from islands of the Canadian Arctic Archipelago. Solid lines indicate median values, dashed lines indicate mean values, whiskers represent 10th and 90th percentiles, and dots are 5th and 95th percentiles. Data sources are as follows: Prince Patrick Island (ANTONIADES et al. 2003a), Banks Island (LIM et al. 2005), Victoria Island (MICHELUTTI et al. 2002a), Bathurst Island (LIM et al. 2001), Devon Island (LIM & DOUGLAS 2003), Ellef Ringnes Island (ANTONIADES et al. 2003b), Axel Heiberg Island (MICHELUTTI et al. 2002b), Alert (ANTONIADES et al. 2003a). Figure 4. Principle components analysis (PCA) biplot of measured environmental variables (arrows) and sampling sites (symbols). The light lines represent variables that were run passively in the PCA. Sampling sites are differentiated into: 4a) dominant type of bedrock geology (after HARRISON 1994), or 4b) bioclimatic zones (after EDLUND 1994). See caption of Figure 2a and 2b for symbol legends. Figure 5. Box plots of selected variables showing sites from bioclimatic zone 4, the mostly lushly vegetated zone on Melville Island (n=16) versus all other bioclimatic zones combined (n=30). Solid lines indicate median values, dashed lines indicate mean values, whiskers represent 10th and 90th percentiles, and dots are 5th and 95th percentiles.

Figure 1.

Figure 2.

Melville

Prince

Patr

ick

Banks

Victori

a

Bathurs

t

Devon

Ellef R

ingne

s

Axel H

eibergAler

t

TP u

nfilt

ered

(μg/

L)

0102030405060708090

120140

Tota

l nitr

ogen

(mg/

L)

0.00

0.25

0.50

0.75

1.00

1.25

2.00

Melville

Prince

Patrick

Banks

Victori

a

Bathurs

t

Devon

Ellef R

ingne

s

Axel H

eibergAler

t

Ca

(mg/

L)

0

25

50

75

100

125200

DIC

(mg/

L)0

10

20

30

40

50

Chl

a(μ

g/L)

0

1

2

3

4

5

6

DO

C (m

g/L)

02468

10121416

pH

4.55.05.56.06.57.07.58.08.59.0

Spe

c. c

ond.

(μS

/cm

)

0150300450600750900

105015002000

Figure 3.

-1.0 1.0

Figure 4.

-1.0

1.0

SO4

SiO2

CHLa

POC

DOC

Mn

TdN

TP

TPF

pH

COND

TEMPLONG

LAT

ELEVAB

C

DE

F

G

H

I

JKL

M NO

P

QRS

T U

V

W

XY Z

AA

AB

AC

AD

AE

AF

AG

AH

AIAJ

AKAL

AM

AN

AO

AP

AQ

AR

AS

AT

Axis 1λ = 0.315

Axi

s 2 λ

= 0.

237

-1.0 1.0

-1.0

1.0

SO4

SiO2

CHLa

POC

DOC

Mn

TdN

TP

TPF

pH

COND

TEMPLONG

LAT

ELEVAB

C

DE

F

G

H

I

JKL

M NO

P

QRS

T U

V

W

XY Z

AA

AB

AC

AD

AE

AF

AG

AH

AIAJ

AKAL

AM

AN

AO

AP

AQ

AR

AS

AT

Axis 1λ = 0.315

Axis

2λ

= 0.

237

a)

b)

-1.0 1.0

-1.0

1.0

SO4

SiO2

CHLa

POC

DOC

Mn

TdN

TP

TPF

pH

COND

TEMPLONG

LAT

ELEVAB

C

DE

F

G

H

I

JKL

M NO

P

QRS

T U

V

W

XY Z

AA

AB

AC

AD

AE

AF

AG

AH

AIAJ

AKAL

AM

AN

AO

AP

AQ

AR

AS

AT

Axis 1λ = 0.315

Axi

s 2 λ

= 0.

237

-1.0 1.0

-1.0

1.0

SO4

SiO2

CHLa

POC

DOC

Mn

TdN

TP

TPF

pH

COND

TEMPLONG

LAT

ELEVAB

C

DE

F

G

H

I

JKL

M NO

P

QRS

T U

V

W

XY Z

AA

AB

AC

AD

AE

AF

AG

AH

AIAJ

AKAL

AM

AN

AO

AP

AQ

AR

AS

AT

Axis 1λ = 0.315

Axis

2λ

= 0.

237

a)

b)

Chl

a(μ

g/L)

0

1

2

3

48

TN:T

P ra

tio

20406080100120140160180

zone 4 others

TdN

(mg/

L)

0.00

0.25

0.50

0.75

1.00

zone 4 others

TP u

nfilt

ered

(μg/

L)

0

10

20

30

40

50

60120140

pH

7.007.257.507.758.008.258.508.759.00

Spe

c. c

ond.

(μS

/cm

)

010020030040050060070010001200

DIC

(mg/

L)

0

10

20

30

40

50

DO

C (m

g/L)

024681012141618

Figure 5.

Table 1. The chemical and physical parameters for 46 freshwater sites on Melville Island. The values in italics represent variables that were below the method detection limit and were replaced with half the detection limit. Bold values with asterisks indicate missing values that were replaced using a linear regression equation with a highly correlated variable. No variables were highly correlated to SRP, thus missing/suspicious values were replaced with the mean SRP value across all sites.

Sample Pond or Latitude Longitude ELEV SA pH CONDID Lake m asl ha μS/cm

MVA P 74° 45.434'N 110° 37.057'W 5.0 0.478 8.7 340MVB P 74° 45.270'N 110° 35.772'W 5.0 0.059 8.8 800MVC P 74° 45.447'N 110° 38.237'W 5.0 0.785 8.7 231MVD P 74° 45.860'N 110° 38.528'W 10.0 0.024 8.6 187MVE P 74° 48.270'N 110° 36.784'W 15.0 0.039 8.5 149MVF P 74° 48.207'N 110° 36.773'W 15.0 0.020 8.5 161MVG P 74° 47.478'N 110° 44.990'W 70.0 2.356 8.5 109MVH P 74° 45.688'N 110° 45.846'W 65.0 7.854 8.4 101MVI P 74° 45.983'N 110° 44.217'W 65.0 1.178 8.4 120MVJ P 74° 48.323'N 110° 48.732'W 65.0 1.257 7.8 18MVK P 74° 48.728'N 110° 47.416'W 70.0 2.945 8.1 88MVL P 74° 45.936'N 110° 39.913'W 5.0 0.031 8.5 318MVM P 74° 45.865'N 110° 39.793'W 15.0 0.126 8.5 292MVN P? 74° 39.367'N 111° 04.364'W 40.0 4.909 8.3 150MVO P? 74° 27.887'N 112° 15.15'W 121.9 5.498 8.0 23MVP P 74° 28.04'N 112° 18.69'W 152.4 0.079 8.0 18MVQ P? 74° 25.138'N 112° 51.359'W 274.3 7.069 8.1 70MVR L? 74° 33.689'N 113° 04.570'W 304.8 19.635 8.0 39MVS P? 74° 47.301'N 113° 12.205'W 243.8 3.142 8.1 48MVT P 74° 53.809'N 110° 46.241'W 76.2 0.332 8.2 79MVU P 74° 44.619'N 110° 44.995'W 70.0 31.416 8.1 112MVV P? 74° 37.999'N 111° 19.198'W 80.0 78.540 8.0 90MVW P? 75° 09.427'N 111° 55.432'W 213.4 7.069 7.7 21MVX P 75° 24.080'N 111° 46.290'W 228.6 0.079 7.9 56MVY P 75° 38.291'N 111° 31.048'W 109.7 0.008 7.7 13MVZ P 76° 01.188'N 112° 22.334'W 173.7 0.785 7.3 10MVAA P? 76° 12.280'N 114° 00.330'W 85.3 4.909 8.0 32MVAB P 76° 19.755'N 114° 09.392'W 70.1 19.635 7.9 1040MVAC P 76° 23.812'N 114° 04.243'W 82.3 0.283 8.2 51MVAD P? 76° 12.910'N 113° 25.307'W 91.4 0.385 8.1 43MVAE L 75° 50.410'N 114° 41.532'W 655.3 19.635 8.1 5MVAF L? 75° 29.174'N 114° 24.009'W 442.0 9.425 6.9 5MVAG P? 75° 29.749'N 113° 33.54'W 579.1 12.566 7.5 16MVAH P 74° 57.907'N 109° 10.562'W 3.0 1.021 8.1 245MVAI P 74° 57.491'N 108° 44.407'W 15.2 3.534 8.0 1230MVAJ P? 75° 27.120'N 110° 19.200'W 121.9 11.781 7.8 7MVAK P 75° 56.722'N 109° 19.519'W 61.0 0.159 7.9 28MVAL L 76° 12.605'N 109° 33.637'W 15.2 80.994 7.5 83MVAM P 76° 38.657'N 108° 55.793'W 106.7 0.008 8.4 520MVAN P? 76° 32.363'N 108° 43.071'W 61.0 12.566 7.8 39MVAO L 76° 23.840'N 108° 33.277'W 30.5 58.905 7.7 77MVAP L 75° 40.114'N 107° 00.473'W 274.3 0.785 7.1 5MVAQ P? 75° 53.673'N 106° 03.333'W 274.3 2.688 7.7 5MVAR P 75° 16.517'N 106° 19.807'W 182.9 1.767 8.0 25MVAS P 75° 08.256'N 107° 37.298'W 76.2 0.238 8.3 77MVAT P 75° 19.020'N 111° 25.184'W 120.0 0.785 8.1 39MEAN 127.1 9.082 7.8 157MEDIAN 76.2 1.512 8.1 74HIGH 655.3 80.994 8.8 1230LOW 5.0 0.008 6.8 5

Table 1. Continued. Sample TEMP TP TPF TdN-N TKN-N NH3-N NO2-N

ID ° Celsius μg/L μg/L mg/L mg/L mg/L mg/LMVA 8.0 10.1 7.7 0.598 0.66 0.015 0.001MVB 8.0 14.8 6.9 0.703 0.734 0.013 0.001MVC 7.0 9.9 3.8 0.587 0.604 0.01 0.001MVD 9.0 9.8 3.9 0.672 0.615 0.007 0.001MVE 7.5 12.1 32.2 0.769 0.702 0.013 0.001MVF 7.0 14.4 24.9 0.892 0.938 0.008 0.001MVG 8.5 13.4 13.1 0.309 0.311 0.007 0.001MVH 9.5 16.8 16.9 0.184 0.177 0.007 0.001MVI 9.5 8 6.1 0.21 0.202 0.009 0.001MVJ 5.0 14.7 4.6 0.201 0.179 0.007 0.001MVK 8.0 22.1 6.1 0.29 0.277 0.0025 0.001MVL 11.0 20.6 5.2 0.977 1.18 0.013 0.002MVM 11.5 12.4 5.7 0.966 1.14 0.009 0.001MVN 11.5 14.4 5.1 0.179 0.175 0.006 0.001MVO 6.5 9.1 4.4 0.131 0.113 0.011 0.001MVP 15.5 20.1 14.9 0.309 0.316 0.008 0.003MVQ 11.0 6.67* 4 0.134 0.108 0.008 0.001MVR 6.0 12.3 4.5 0.087 0.063 0.005 0.001MVS 15.0 8.6 7.1 0.245 0.221 0.009 0.001MVT 14.0 12.2 13.9 0.333 0.321 0.008 0.001MVU 11.0 11.8 11.6 0.239 0.218 0.013 0.002MVV 6.0 6.2 4.9 0.067 0.049 0.006 0.001MVW 10.0 24.4 9.4 0.146 0.096 0.011 0.002MVX 14.0 15.4 2 0.81 0.838 0.024 0.002MVY 15.5 11.6 9 0.351 0.347 0.006 0.006MVZ 13.0 8.4 6.1 0.062 0.032 0.01 0.002MVAA 13.0 45.4 22.7 0.274 0.26 0.01 0.001MVAB 9.0 23.5 9.7 0.213 0.186 0.011 0.001MVAC 13.5 13 10 0.611 0.644 0.03 0.001MVAD 14.5 13 10.7 0.403 0.394 0.015 0.001MVAE 4.0 12.4 8.9 0.046 0.007 0.008 0.001MVAF 3.0 16.8 15.3 0.03 0.007 0.006 0.0005MVAG 11.0 9.5 13.7 0.056 0.032 0.006 0.001MVAH 9.0 27.3 10 0.393 0.399 0.013 0.001MVAI 9.0 12.9 9.4 0.327 0.331 0.016 0.001MVAJ 7.5 14.2 8.3 0.04 0.02 0.009 0.002MVAK 6.5 27.4 25.6 0.601 0.59 0.012 0.002MVAL 2.5 73.1 22.4 0.125 0.106 0.019 0.002MVAM 5.5 8.8 10 0.406 0.412 0.015 0.001MVAN 7.5 20.9 10 0.119 0.093 0.016 0.001MVAO 5.5 123 7.3 0.181 0.087 0.013 0.003MVAP 2.5 10 5.5 0.032 0.007 0.009 0.001MVAQ 3.5 34.5 6.9 0.061 0.04 0.013 0.001MVAR 5.0 135 8.8 0.286 0.275 0.012 0.001MVAS 7.5 13.4 3.9 0.236 0.227 0.024 0.001MVAT 6.0 12.7 7.8 0.641 0.648 0.011 0.001MEAN 8.8 21.5 10.0 0.338 0.334 0.011 0.001MEDIAN 8.3 13.2 8.6 0.260 0.244 0.010 0.001HIGH 15.5 135.0 32.2 0.977 1.180 0.030 0.006LOW 2.5 6.2 2.0 0.030 0.007 <0.005 <0.001

Table 1. Continued. Sample CHLa POC PON DOC DIC Cl SO4

ID ug/L mg/L mg/L mg/L mg/L mg/L mg/LMVA 0.5 0.414 0.041 7.6 29.9 84.5 1.9MVB 0.5 0.443 0.045 9 41.5 271 21.8MVC 0.5 0.38 0.034 8 34.4 26.5 7.3MVD 0.5 0.346 0.03 8.8 23.7 29.7 1.8MVE 1 0.436 0.036 10 26.3 11.6 1MVF 0.05 0.404 0.033 14.1 27.9 14.6 1.3MVG 0.6 0.674 0.066 5.6 23 2.8 0.9MVH 0.9 0.437 0.054 4.7 16.1 6.78 1.5MVI 1 0.218 0.02 3.1 20.6 6.53 1.3MVJ 1.3 0.692 0.082 3.8 1.4 2.6 1MVK 0.8 0.473 0.055 4 14.3 4.59 1.3MVL 1.4 0.44 0.037 16.3 47.8 11.7 2.5MVM 0.9 0.524 0.04 16.3 45.1 11.5 1.7MVN 0.9 0.351 0.038 3.8 24.7 12.1 1.8MVO 1.3 0.315 0.034 2.7 2.4 2.24 2.2MVP 5.6 2.3 0.21 6.6 2.5 1.37 0.9MVQ 0.5 0.278 0.028 2.5 10 1.18 5.9MVR 0.4 0.254 0.021 1.7 6.2 0.82 2.1MVS 0.5 0.338 0.035 3 7.2 2.12 0.7MVT 0.5 0.299 0.032 4.6 15.4 2.2 0.2MVU 1.3 0.393 0.043 3.2 17.4 6.8 1.3MVV 1.9 0.162 0.018 1.7 13.4 3.46 2.2MVW 0.9 0.412 0.043 2.8 2.3 1.14 3.1MVX 0.5 0.372 0.043 9.4 9.5 2.2 0.6MVY 0.3 0.448 0.05 5.9 2.7 2.25 0.8MVZ 0.9 0.2 0.02 2 0.7 0.62 1.8MVAA 1.4 0.414 0.051 4.9 3.8 3.29 3MVAB 3.7 0.677 0.083 6.7 10.9 412 45.7MVAC 1.5 0.484 0.051 7.2 5.3 7.51 0.7MVAD 1 0.653 0.054 5.2 2 8.62 1.5MVAE 0.3 0.178 0.011 0.8 0.7 0.4* 28.3MVAF 0.05 0.336 0.033 1.1 0.6 0.47 0.4MVAG 0.05 0.238 0.022 1.5 2.3 0.46 2.3MVAH 1.3 1.1 0.122 5.6 10.9 77 17.3MVAI 0.5 0.448 0.06 3.3 13.5 588 47.6MVAJ 0.3 0.205 0.021 1.2 0.2 1.49 0.5MVAK 10.1 2.71 0.296 12.3 2.4 4.45 0.8MVAL 0.4 0.379 0.039 5 2.1 20.2 18.7MVAM 0.9 0.434 0.043 7.2 19.4 3.8 357MVAN 0.9 0.268 0.025 3 1.6 5.28 12.6MVAO 7.9 0.503 0.041 7.2 1.8 19 18.6MVAP 0.05* 0.198 0.017 1.3 0.3 0.64 0.7MVAQ 3.6 0.507 0.062 1.1 0.4 0.77 0.9MVAR 1.6 0.526 0.06 3.9 2.5 1.28 4.1MVAS 1.1 0.32 0.03 3 13.1 3.52 1.7MVAT 1.6 0.559 0.053 8.2 4.8 1.93 4.8MEAN 1 0.503 0.051 5.5 12.3 36.59 13.8MEDIAN 1 0.413 0.041 4.7 8.4 3.66 1.8HIGH 10.1 2.710 0.296 16.3 47.8 588.00 357.0LOW <0.1 0.162 0.011 0.8 0.2 0.40 0.2

Table 1. Continued. Sample Ca Mg Na K SiO2 SRP Al

ID mg/L mg/L mg/L mg/L mg/L mg/L mg/LMVA 28.8 20.3 43.7 2.1 0.26 0.0006 0.02MVB 45.9 42.2 140 7.2 0.2 0.0011 0.03MVC 29.1 21.2 14.4 2.8 0.07 0.001 0.01MVD 27 12.2 13.2 1.1 0.27 0.0007 0.01MVE 25.2 11.6 4.8 0.8 0.14 0.0085 0.01MVF 26.6 14.9 6.1 1.3 0.05 0.0015 0.01MVG 21 10.3 1.3 1.1 0.1 0.001 0.25MVH 13.8 8.9 2.5 1.8 0.11 0.0091 2.01MVI 19.2 9.9 2.4 1.2 0.11 0.0009 0.04MVJ 0.8 0.7 1.1 0.4 0.1 0.0001 0.37MVK 12.3 7.7 1.6 0.9 0.24 0.0006 0.23MVL 40.5 28.5 5.7 2.4 0.32 0.0008 0.03MVM 38.7 27 5.3 2 0.23 0.0006 0.02MVN 20.8 15 5.6 2.1 0.06 0.0004 0.92MVO 1.8 1.6 1.6 0.5 0.12 0.0001 0.04MVP 1.5 1.1 0.7 0.1 0.2 0.0007 0.03MVQ 10.7 4.8 0.6 0.3 0.12 0.0016 0.05MVR 5.8 3 0.4 0.3 0.23 0.0001 0.06MVS 5.6 3.6 0.9 0.7 0.19 0.0002 0.03MVT 12.5 7.3 1.2 0.9 0.33 0.0003 0.1MVU 16.1 8 2.6 1.1 0.12 0.0001 0.37MVV 12.4 5.7 2 0.6 0.07 0.0007 0.01MVW 2.6 1.2 0.6 0.8 0.55 0.0008 1.75MVX 8.6 4 1 0.9 0.21 0.0006 0.01MVY 3.3 0.8 1 0.4 0.37 0.0006 0.1MVZ 0.8 0.3 0.2 0.2 0.4 0.0008 0.03MVAA 4.4 1.3 1.6 0.9 0.11 0.0063 0.08MVAB 22.1 30.3 219 8.1 0.22 0.0013 0.27MVAC 6.3 1.7 4.7 0.3 0.13 0.0028 0.005MVAD 2.2 1.2 3.6 0.8 0.25 0.0017 0.01MVAE 0.1 0.1 0.1 0.1 0.06 0.0011 0.08MVAF 0.1 0.2 0.1 0.1 0.16 0.0003 0.21MVAG 0.9 1.5 0.3 0.7 0.08 0.0013 0.42MVAH 13.3 13.5 44.4 3.2 0.09 0.0028 2.43MVAI 31.4 48.1 297 14.1 0.02 0.0049 0.07MVAJ 0.5 0.5 2.1 0.2 0.04 0.0135 0.19MVAK 1.8 0.7 1.8 0.1 1.48 0.0079 0.005MVAL 2.2 2.1 14.7 2 3.31 0.009 1.96MVAM 179 6.8 4.2 1.2 2.92 0.0113 0.005MVAN 2.9 1.8 4 0.9 0.18 0.0028 0.09MVAO 3.8 2.7 13.2 1.7 3.23 0.0075 2.29MVAP 0.2 0.2 0.3 0.2 0.21 0.0028 0.42MVAQ 0.3 0.3 0.3 0.1 0.2 0.006 0.47MVAR 1.8 1.7 1.6 0.5 0.59 0.004 0.13MVAS 15.1 3.9 1.7 1.1 0.06 0.0028 0.03MVAT 5.2 2.7 1.5 0.4 0.27 0.0057 0.02MEAN 15.8 8.5 19.1 1.5 0.41 0.0028 0.34MEDIAN 7.5 3.8 1.9 0.9 0.20 0.001 0.07HIGH 179.0 48.1 297.0 14.1 3.31 0.014 2.43LOW 0.1 0.1 <0.2 <0.2 0.02 <0.0002 <0.01

Table 1. Continued. Sample Ba Cu Fe Li Mn Sr Zn

ID mg/L mg/L mg/L mg/L mg/L mg/L mg/LMVA 0.0109 0.001 0.121 0.002 0.0135 0.0719 0.0005MVB 0.0283 0.002 0.14 0.008 0.0062 0.173 0.0005MVC 0.0121 0.002 0.041 0.002 0.004 0.0457 0.0005MVD 0.0055 0.002 0.075 0.002 0.0057 0.0368 0.0005MVE 0.007 0.002 0.111 0.001 0.0034 0.0221 0.0005MVF 0.0043 0.002 0.041 0.001 0.001 0.0203 0.0005MVG 0.0056 0.002 0.331 0.001 0.0125 0.0162 0.001MVH 0.0117 0.003 1.65 0.004 0.0192 0.0202 0.004MVI 0.0049 0.001 0.053 0.001 0.0018 0.0208 0.0005MVJ 0.0039 0.002 0.852 0.001 0.0129 0.0028 0.002MVK 0.0047 0.002 0.438 0.001 0.018 0.0126 0.002MVL 0.0108 0.002 0.124 0.001 0.0033 0.0246 0.0005MVM 0.0089 0.002 0.104 0.001 0.0027 0.0293 0.0005MVN 0.009 0.001 0.779 0.002 0.0084 0.0317 0.002MVO 0.0008 0.002 0.065 0.001 0.006 0.0063 0.0005MVP 0.00025 0.001 0.826 0.0005 0.0061 0.0031 0.001MVQ 0.0063 0.001 0.149 0.0005 0.0055 0.0208 0.0005MVR 0.0028 0.001 0.086 0.001 0.0029 0.0094 0.0005MVS 0.0015 0.001 0.083 0.0005 0.0098 0.0055 0.0005MVT 0.0031 0.002 0.152 0.001 0.0041 0.0132 0.001MVU 0.0057 0.002 0.372 0.001 0.0054 0.0209 0.003MVV 0.0028 0.001 0.021 0.001 0.0024 0.015 0.0005MVW 0.0053 0.001 0.99 0.001 0.0124 0.007 0.003MVX 0.0028 0.001 0.185 0.002 0.0109 0.0124 0.001MVY 0.0013 0.001 0.213 0.0005 0.0052 0.0098 0.001MVZ 0.0033 0.0005 0.05 0.001 0.0015 0.0043 0.001MVAA 0.0146 0.0005 0.247 0.003 0.0017 0.024 0.001MVAB 0.0373 0.001 0.327 0.011 0.0045 0.2 0.001MVAC 0.0086 0.0005 0.145 0.001 0.0027 0.0199 0.0005MVAD 0.0072 0.0005 0.085 0.005 0.0006 0.0158 0.0005MVAE 0.0027 0.0005 0.128 0.0005 0.0053 0.00025 0.001MVAF 0.0017 0.002 0.364 0.001 0.0089 0.00025 0.001MVAG 0.0023 0.001 0.434 0.0005 0.0046 0.0087 0.002MVAH 0.0129 0.003 2.02 0.004 0.0274 0.083 0.005MVAI 0.0069 0.002 0.083 0.007 0.003 0.185 0.001MVAJ 0.0035 0.001 0.324 0.0005 0.0065 0.0028 0.002MVAK 0.003 0.001 0.2 0.001 0.0062 0.0128 0.002MVAL 0.0086 0.002 2.67 0.008 0.0167 0.0317 0.006MVAM 0.0254 0.0005 0.026 0.004 0.0055 1 0.001MVAN 0.0024 0.0005 0.099 0.006 0.0018 0.0218 0.0005MVAO 0.0133 0.006 5.39 0.01 0.0347 0.0481 0.013MVAP 0.0072 0.001 0.168 0.0005 0.0035 0.0017 0.004MVAQ 0.003 0.0005 0.918 0.0005 0.0102 0.0013 0.002MVAR 0.0021 0.002 0.273 0.001 0.0163 0.0083 0.001MVAS 0.0053 0.0005 0.146 0.0005 0.0094 0.0399 0.0005MVAT 0.0023 0.001 0.353 0.001 0.0052 0.0112 0.0005MEAN 0.0073 0.001 0.488 0.002 0.0078 0.0516 0.002MEDIAN 0.0053 0.001 0.160 0.001 0.0055 0.0181 0.001HIGH 0.0373 0.006 5.390 0.011 0.0347 1.0000 0.013LOW <0.0005 <0.001 0.021 <0.001 0.0006 <0.0005 <0.001

Table 1. Continued.

Sample Turbid? In/Out flow? Nostoc? Inverts?ID

MVA n n y fairy shrimp, Daphnia, red & brown chironomidsMVB n n n DaphniaMVC n n n DaphniaMVD n n y nMVE n n n nMVF n n n nMVG y n n nMVH y n y nMVI n n n nMVJ y n n copepodsMVK n n n fairy shrimp MVL n n y Daphnia, chironomidsMVM n n n fairy shrimpMVN n n n fairy shrimpMVO n n n nMVP n n n copepods, chironomidsMVQ n n y fairy shrimp (small)MVR n n n fairy shrimpMVS n n n nMVT n n y fairy shrimpMVU n n n fairy shrimpMVV n n n nMVW y n n nMVX n n y fairy shrimpMVY n n n fairy shrimp, Daphnia MVZ n n n nMVAA n n y fairy shrimpMVAB y n n nMVAC n n n fairy shrimpMVAD n n n copepods, fairy shrimp, DaphniaMVAE n n n nMVAF y inflow (small) n nMVAG n n n red chironomidMVAH n n n nMVAI n n n fairy shrimp, 2 sizesMVAJ n n n nMVAK n n y nMVAL y inflow n nMVAM n n y nMVAN n outflow n nMVAO y n n nMVAP n n n nMVAQ n n n nMVAR n n n nMVAS n n n nMVAT n n n fairy shrimp, Daphnia, chironmids

Table 2. Selected nutrient ratio parameters for 46 freshwater sites on Melville Island.

Sample TN:TP POC:PON PON:POP POC:POP POC:CHLa POC:Chla Na:KID molar molar molar molar mg/L:mg/L μmol/L:μg/L mg/L:mg/L

MVA 139.9 11.78 37.76 445 828.0 68.9 20.81MVB 111.7 11.48 12.59 145 886.0 73.8 19.44MVC 138.7 13.04 12.32 161 760.0 63.3 5.14MVD 158.3 13.45 11.24 151 692.0 57.6 12.00MVE 147.1 14.13 NA NA 436.0 36.3 6.00MVF 142.0 14.28 NA NA 8080.0 672.8 4.69MVG 61.9 11.91 486.32 5793 1123.3 93.5 1.18MVH 31.3 9.44 NA NA 485.6 40.4 1.39MVI 63.6 12.72 23.27 296 218.0 18.2 2.00MVJ 42.6 9.84 17.95 177 532.3 44.3 2.75MVK 34.5 10.03 7.60 76 591.3 49.2 1.78MVL 108.8 13.87 5.31 74 314.3 26.2 2.38MVM 179.3 15.28 13.20 202 582.2 48.5 2.65MVN 33.3 10.78 9.03 97 390.0 32.5 2.67MVO 40.1 10.81 15.99 173 242.3 20.2 3.20MVP 57.1 12.78 89.27 1141 410.7 34.2 7.00MVQ 45.0 11.58 23.18 269 556.0 46.3 2.00MVR 19.4 14.11 5.95 84 635.0 52.9 1.33MVS 72.0 11.27 51.58 581 676.0 56.3 1.29MVT 66.1 10.90 NA NA 598.0 49.8 1.33MVU 52.8 10.66 475.27 5067 302.3 25.2 2.36MVV 30.3 10.50 30.61 321 85.3 7.1 3.33MVW 17.1 11.18 6.34 71 457.8 38.1 0.75MVX 122.4 10.09 7.09 72 744.0 61.9 1.11MVY 76.4 10.45 42.51 444 1493.3 124.3 2.50MVZ 21.6 11.67 19.22 224 222.2 18.5 1.00MVAA 15.8 9.47 4.97 47 295.7 24.6 1.78MVAB 27.8 9.51 13.30 127 183.0 15.2 27.04MVAC 112.6 11.07 NA NA 322.7 26.9 15.67MVAD 77.7 14.11 51.90 732 653.0 54.4 4.50MVAE 10.2 18.88 6.95 131 593.3 49.4 1.00MVAF 8.3 11.88 48.63 578 6720.0 559.5 1.00MVAG 18.1 12.62 NA NA 4760.0 396.3 0.43MVAH 41.7 10.52 NA NA 846.2 70.5 13.88MVAI 66.3 8.71 37.90 330 896.0 74.6 21.06

Table 2. continued.

Sample TN:TP POC:PON PON:POP POC:POP POC:CHLa POC:Chla Na:KID molar molar molar molar mg/L:mg/L μmol/L:μg/L mg/L:mg/L

MVAJ 9.5 11.39 7.87 90 683.3 56.9 10.50MVAK 72.4 10.68 363.51 3882 268.3 22.3 18.00MVAL 5.0 11.34 1.70 19 947.5 78.9 7.35MVAM 112.8 11.77 NA NA 482.2 40.2 3.50MVAN 15.2 12.51 NA NA 297.8 24.8 4.44MVAO 4.0 14.31 0.78 11 63.7 5.3 7.76MVAP 10.8 13.59 8.35 113 3960.0 330.0 1.50MVAQ 7.9 9.54 4.97 47 140.8 11.7 3.00MVAR 5.7 10.23 1.05 11 328.8 27.4 3.20MVAS 43.9 12.44 6.98 87 290.9 24.2 1.55MVAT 120.8 12.30 23.91 294 349.4 29.1 3.75MEAN 60.2 11.8 53.7 609.8 987.5 82.2 5.7MEDIAN 44.4 11.5 13.2 160.6 544.2 45.3 2.9MAXIMUM 179.3 18.88 475.27 5793 8080 672.8 27.04MINIMUM 4 8.71 <0 <0 63.7 5.3 0.43

Tabl

e 3.

Pea

rson

cor

rela

tion

mat

rix w

ith B

onfe

rron

i-adj

uste

d pr

obab

ilitie

s. S

igni

fican

tly c

orre

late

d va

riabl

es a

re in

dica

ted

in b

old

(p<0

.01)

or i

talic

s (p<

0.05

). N

H3

NO

2C

lSO

4Si

O2

CH

La

POC

PON

DO

CD

ICT

KN

SRP

Ba

FeM

nSr

Ca

NO

2-0

.02

1C

l0.

234

-0.1

181

SO4

0.17

4-0

.109

0.55

81

SiO

20.

148

0.33

8-0

.119

0.22

41

CH

La

0.13

50.

220.

039

0.06

0.41

31

POC

0.11

20.

209

0.18

60.

009

0.27

60.

621

PON

0.12

60.

203

0.17

3-0

.003

0.23

70.

601

0.96

91

DO

C0.

255

0.16

0.44

60.

093

0.29

60.

226

0.55

30.

483

1D

IC-0

.03

-0.1

930.

489

0.09

3-0

.139

-0.2

110.

009

-0.0

470.

618

1T

KN

0.27

60.

008

0.37

1-0

.043

0.04

9-0

.022

0.34

20.

260.

882

0.73

1SR

P0.

339

0.00

30.

146

0.35

20.

218

0.27

0.15

70.

156

0.13

8-0

.132

0.02

81

Ba

0.23

9-0

.278

0.60

30.

501

0.11

8-0

.013

-0.0

34-0

.016

0.34

30.

482

0.25

50.

344

1Fe

-0.0

160.

247

-0.0

3-0

.014

0.24

20.

319

0.37

90.

41-0

.062

-0.3

25-0

.282

0.20

40.

045

1M

n-0

.004

0.12

30.

019

0.07

70.

274

0.18

30.

255

0.30

9-0

.046

-0.1

2-0

.194

0.05

90.

029

0.70

51

Sr0.

32-0

.058

0.61

80.

569

0.17

80.

074

0.19

0.22

70.

575

0.51

70.

397

0.25

60.

672

-0.1

45-0

.019

1C

a0.

151

-0.1

150.

526

0.29

8-0

.004

-0.0

810.

119

0.13

30.

653

0.76

80.

583

0.00

30.

534

-0.2

97-0

.068

0.86

91

Mg

0.03

7-0

.241

0.70

40.

307

-0.1

9-0

.146

0.08

40.

093

0.57

30.

844

0.56

7-0

.068

0.55

7-0

.204

-0.0

140.

758

0.89

6N

a0.

306

-0.0

780.

886

0.50

5-0

.007

0.10

20.

263

0.29

50.

550.

498

0.41

30.

224

0.62

50.

001

0.04

60.

813

0.66

5K

0.17

6-0

.149

0.83

50.

542

-0.1

16-0

.095

0.09

10.

138

0.41

50.

573

0.33

0.09

30.

660.

016

0.06

10.

763

0.67

3A

l0.

017

0.14

50.

134

0.23

0.28

10.

178

0.14

30.

168

-0.0

46-0

.181

-0.2

790.

265

0.28

0.78

30.

603

0.10

4-0

.069

Cu

-0.1

350.

111

0.29

10.

121

0.21

20.

201

0.18

30.

158

0.29

20.

206

0.10

50.

046

0.22

0.46

70.

494

0.19

0.18

1L

i0.

261

0.00

80.

661

0.66

50.

30.

228

0.15

50.

182

0.22

90.

034

-0.0

40.

331

0.56

70.

274

0.11

20.

571

0.24

9Z

n0.

040.

285

0.07

80.

204

0.44

10.

402

0.11

80.

109

-0.0

3-0

.262

-0.2

930.

302

0.23

0.72

20.

546

0.05

4-0

.173

TdN

0.30

60.

039

0.43

0.00

10.

097

0.04

60.

443

0.37

50.

931

0.67

70.

964

0.05

40.

275

-0.2

37-0

.132

0.50

20.

662

TP

0.16

0.16

60.

049

0.22

20.

467

0.43

30.

364

0.37

80.

117

-0.2

64-0

.104

0.42

20.

130.

655

0.45

70.

026

-0.1

86T

PF0.

235

-0.0

40.

143

0.15

20.

092

0.09

70.

345

0.33

10.

148

-0.1

640.

024

0.42

40.

108

0.20

5-0

.063

0.14

3-0

.01

pH0.

077

-0.2

320.

506

0.12

3-0

.218

-0.0

970.

147

0.09

50.

591

0.78

10.

634

-0.0

40.

349

-0.3

43-0

.087

0.54

70.

782

CO

ND

0.19

2-0

.215

0.77

10.

484

-0.0

17-0

.034

0.16

90.

183

0.61

50.

739

0.53

10.

052

0.65

3-0

.199

-0.0

220.

889

0.91

5E

LE

V-0

.177

0.03

8-0

.729

-0.2

5-0

.002

-0.0

16-0

.261

-0.2

32-0

.674

-0.7

14-0

.618

-0.1

24-0

.535

0.02

6-0

.046

-0.6

36-0

.648

TE

MP

0.07

20.

342

-0.0

34-0

.305

-0.1

37-0

.067

0.15

40.

176

0.26

90.

119

0.25

8-0

.26

-0.1

28-0

.109

-0.2

790.

155

0.27

5L

ON

G-0

.194

-0.0

25-0

.048

-0.1

18-0

.213

-0.2

73-0

.119

-0.1

46-0

.081

-0.1

01-0

.046

-0.4

17-0

.126

-0.1

52-0

.308

-0.2

24-0

.116

LA

T0.

418

0.15

2-0

.016

0.33

50.

465

0.25

80.

029

0.02

8-0

.067

-0.4

81-0

.208

0.53

20.

226

0.12

7-0

.128

0.03

4-0

.324

Tabl

e 3.

Con

tinue

d.

M

gN

aK

Al

Cu

Li

Zn

TdN

TP

TPF

pHC

ON

DE

LE

VT

EM

PL

ON

GN

a0.

771

K0.

840.

884

1A

l-0

.013

0.16

70.

221

Cu

0.31

10.

341

0.35

20.

584

1L

i0.

343

0.70

80.

691

0.40

80.

389

1Z

n-0

.139

0.11

70.

101

0.82

40.

711

0.45

61

TdN

0.59

80.

492

0.37

5-0

.236

0.15

0.03

6-0

.25

1T

P-0

.157

0.12

30.

058

0.50

70.

426

0.40

90.

582

-0.0

551

TPF

-0.0

660.

168

0.02

10.

283

0.03

40.

202

0.13

50.

075

0.22

11

pH0.

737

0.48

90.

434

-0.2

350.

08-0

.017

-0.3

50.

691

-0.2

4-0

.076

1C

ON

D0.

950.

855

0.87

20.

026

0.27

90.

514

-0.0

680.

596

-0.0

780.

040.

683

1E

LE

V-0

.738

-0.8

08-0

.694

-0.1

81-0

.436

-0.3

75-0

.097

-0.6

45-0

.054

-0.1

63-0

.596

-0.7

381

TE

MP

0.15

80.

021

0.06

7-0

.18

-0.2

05-0

.128

-0.2

650.

311

-0.2

360.

018

0.20

90.

127

0.00

91

LO

NG

-0.1

03-0

.216

-0.1

43-0

.216

-0.2

61-0

.091

-0.3

11-0

.063

-0.3

320.

002

-0.0

5-0

.11

0.29

80.

436

1L

AT

-0.4

48-0

.026

-0.1

170.

119

-0.1

90.

427

0.28

6-0

.195

0.34

20.

4-0

.466

-0.2

310.

24-0

.126

-0.0

29

CHAPTER 3

EVALUATING THE ROLE OF ENVIRONMENTAL AND SPATIAL VARIABLES ON DIATOM SPECIES

DISTRIBUTIONS ON MELVILLE ISLAND (CANADIAN HIGH ARCTIC)

BRONWYN E. KEATLEY, MARIANNE S.V. DOUGLAS, AND JOHN P. SMOL

This chapter has been submitted to the Canadian Journal of Botany: Keatley, B.E., M.S.V. Douglas, and J.P. Smol. In review. Evaluating the role of environmental and spatial variables on diatom species distributions on Melville Island (Canadian High Arctic). Submitted: 7 June 2007.

Abstract

Diatom species assemblages were identified and enumerated from the surface sediments of 45

lakes and ponds across a wide spectrum of spatial and environmental gradients on Melville

Island, Nunavut/N.W.T, High Arctic Canada. Whereas the most common taxa were similar to

those recorded elsewhere in the Canadian High Arctic, significant differences in assemblages

existed between sites located in the different bioclimatic zones of Melville Island. For example,

taxa recorded in the most lushly vegetated bioclimatic zone were similar to those found in richly

vegetated regions elsewhere in the Canadian Arctic Archipelago, and generally different from

diatoms in the poorly vegetated regions on Melville Island. Of the measured environmental

variables, pH, specific conductivity, surface area, elevation, and chlorophyll a explained

significant portions of the variance in diatom assemblage composition when diatoms were

considered at the scale of the entire island. However, only total dissolved nitrogen was an

important explanatory variable within the most lushly vegetated bioclimatic zone. Ecological

relationships between diatom species distributions and pH were moderately strong, as indicated

by weighted averaging transfer functions (r2boot = 0.432 to 0.746, RMSEP = 0.341 to 0.242).

Interestingly, spatial factors were of little importance, confirming that diatoms are not likely to be

dispersal limited, at least at the landscape scale explored in this study.

Introduction

Polar regions are particularly sensitive to climate change (ACIA 2004), and these

changes are already resulting in dramatic consequences for biological communities across the

circumpolar north (e.g. ACIA 2004; Smol et al. 2005). As long-term monitoring data are often

unavailable for remote Arctic regions, paleolimnological data are being increasingly used to

assess the rate, magnitude and direction of environmental changes (Douglas et al. 2004).

Diatoms, unicellular algae with siliceous cell walls, are one of the most commonly used proxy

indicators used in Arctic paleolimnological studies because they preserve well in lake sediments,

they are prevalent in most aquatic systems, and certain species often have well defined optima to

given environmental variables (Stoermer and Smol 1999). However, in order to use proxy

indicators such as diatoms effectively, it is important to better understand the ecological and

biogeographical characteristics of taxa (Smol 2002).

Over the past approximately two decades, a concerted effort has been made to better

understand the factors influencing diatom species distributions in the Canadian High Arctic. The

resulting body of work includes diatom calibration sets from nine regions of the Canadian High

Arctic (Fig. 1): Cape Herschel (Douglas and Smol 1993, 1995), Bathurst Island (Lim et al. 2001a,

b), Isachsen (Antoniades et al. 2004), Alert and, Mould Bay (Antoniades et al. 2005), Axel

Heiberg Island (Michelutti et al. 2006), Banks Island (Lim et al. 2007), Devon Island (Lim 2004)

and Cornwallis Island (Michelutti et al. in press). Although considerable progress has been made,

critical gaps remain in our understanding of diatom species distributions across the vast and

ecologically diverse High Arctic landscape, and this is especially true for the western High Arctic

(Fig. 1). For example, in the western portion of the Canadian High Arctic, a region for which

climate change research is particularly scarce, our understanding of modern diatom ecology is

based largely on two surveys (Antoniades et al. 2005; Lim et al. 2007), one of which is restricted

to a region less than 5 km2 (Antoniades et al. 2005). Data from these studies revealed large

differences in environmental characteristics and diatom species distributions, underscoring the

need for additional data from the western Arctic.

Here, we explore diatom species-environment relationships from Melville Island, which,

due to its large size (i.e. 42 149 km2), relatively high relief (from sea level to ~655 m above sea

level), multiple bioclimatic zones (Edlund 1994), and marked gradients of water chemistry

(Keatley et al. [2]), provides us with a unique study site in which to determine the environmental

factors influencing diatom species distributions in the western Canadian High Arctic.

The large geographical scale of our sampling regime on Melville Island (spanning nearly

the whole island) also provides a database which allows us to explore the relative roles of spatial

versus environmental factors in structuring diatom communities. Although dispersal is often

thought to be of minimal importance in influencing microscopic species distributions (e.g.

Kristiansen 1996; Finlay 2002; Finlay et al. 2002), as their small size allows them to reach new

habitats via wind, waterfowl, or other vectors, some researchers have argued that dispersal in

diatoms may be more important than initially thought (e.g. Kociolek and Spaulding 2000; Telford

et al. 2006). The lack of any direct human activities makes Melville Island an ideal location to

examine the relative importance of dispersal versus environmental factors in diatom species

distributions.

Our main study objectives are to: 1) describe the ecological characteristics of diatoms

from Melville Island; 2) elucidate which environmental and/or spatial factors are able to explain

diatom species distributions across this island; and 3) assess whether diatom species distributions

can be classified according to previously defined bioclimatic zones. We show that diatom

assemblages are significantly influenced by environmental, but not spatial, variables at the whole

island scale, with pH, specific conductivity, surface area, elevation and chlorophyll a

concentrations explaining significant portions of the diatom variance. Interestingly, diatoms

within the most lushly vegetated bioclimatic zone on Melville Island are significantly different

from those found in all other zones, and a significant proportion of the diatom species variance

within this zone can be explained by total dissolved nitrogen. Finally, we construct quantitative

transfer functions for pH, the most ecologically-relevant variable explaining diatom species

abundances on Melville Island.

Methods

Site description

A detailed site description of Melville Island and the limnological characteristics of the

45 study sites can be found in Keatley et al. ([2]). In brief, Melville Island is located in the

western Canadian High Arctic (Fig. 1), with the border between Nunavut and the Northwest

Territories running longitudinally through the middle of the island. It is the fourth largest island

of the Queen Elizabeth Islands (consisting of the Arctic islands north of the Parry Channel), and

represents five bioclimatic zones (Edlund 1994) and four physiographic regions (Tozer and

Thorsteinsson 1964). The nearest long-term weather station is located at Mould Bay, Prince

Patrick Island (~ 250 km from the centre of Melville Island), where average temperatures for July

and February are 4.0°C and -34.0°C, respectively, and the mean annual precipitation is 111.0 mm

(Environment Canada 2004). Melville Island is the only island in the western Arctic that

currently retains small permanent ice caps.

The surficial geology of the island is largely composed of weathered bedrock, which is

highly variable within regions. In general, the bedrock of eastern Melville Island is comprised of

sandstones, mudstones, limestones, shales, and evaporites (Barnett et al. 1977). Western Melville

Island is dominated by sandstones and carbonates, whereas the Dundas Peninsula in the

southwest portion of the island contains glacial deposits, including calcareous tills derived from

southern locations (Tozer and Thorsteinsson 1964; Hodgson and Vincent 1984).

Of the five previously described bioclimatic zones (Zones 0 through 4) found on Melville

Island (Edlund 1994), four (Zones 1 through 4) were encompassed by our sampling regime (Fig.

1). These zones are characterized both by their climate and vegetation, with Zone 4 representing

the most diverse and lushly vegetated regions, and Zone 0 representing areas completely devoid

of vegetation (Edlund 1994).

Sampling

Surface sediment diatoms were sampled from 46 lakes and ponds (Fig. 1c), chosen to

encompass a large range of physical and chemical characteristics (Keatley et al. [2]). Using

identical sampling protocols used in all our previous Arctic diatom surveys, we sampled the

uppermost ca. 0.5 cm of sediments by hand from the near-shore environment from each site, and

stored the sediment in Whirlpak® sample bags. It was impossible to obtain sediment from pond

MVAF, so we sampled the surfaces of rocks as the dominant substrate for diatom samples. The

samples were kept cool and dark until they were processed in the lab at PEARL, Queen’s

University, Kingston, Canada.

At the same time that the sediment samples were obtained, the lake or pond was sampled

for water chemical variables (e.g. pH, specific conductivity, nutrients, major ions, and trace

metals) and physical characteristics (e.g. temperature, elevation, morphometric parameters). The

physical and chemical limnological methodologies and characteristics of these sites has been fully

described elsewhere (Keatley et al. [2]). Table 1 provides a summary of some key limnological

variables, separated by bioclimatic zones. Briefly, the limnological characteristics from Melville

Island sites represent one of the largest environmental gradients sampled to date in the Canadian

High Arctic. Of particular interest is the distinctive water chemistry of the lakes and ponds

located in bioclimatic Zone 4 (the most richly vegetated zone on Melville Island), which are

characterized by high pH, dissolved organic carbon, and total dissolved nitrogen (Keatley et al.

[2]). In addition, our sampling regime also included several sites close to high elevation ice caps,

which exhibit the lowest specific conductivity (~5 µS/cm) and calcium (0.1 mg/L) yet recorded in

the Canadian Arctic (Keatley et al. [2]).

Diatom analyses

The sediment samples were treated using standard diatom preparation techniques

(Battarbee et al. 2001). Briefly, organic material was digested from the sediment samples using a

50/50 mixture of concentrated HNO3 and H2SO4 at 80 °C for ~8 hours. The resultant slurries

were rinsed with deionized water until they reached a neutral pH. Diatoms were distributed onto

pre-cleaned coverslips, and then mounted onto microscope slides with Naphrax®, a mounting

medium with a high refractive index. Diatoms from each site were enumerated and identified

using Krammer and Lange-Bertalot (1991), Krammer (2002) and Antoniades et al. (in press). A

minimum of 300 valves were identified from each site.

Statistical analyses

Diatom counts were converted to relative abundances and screened for frequency of

occurrence. Any species that were present in >1% relative abundance in at least one site were

retained for further analyses. A detrended correspondence analysis (DCA) was used to assess the

length of the gradient encompassed by the diatom taxa, and to determine whether linear or

unimodal statistical techniques should be used to assess the relationships between environmental

variables and diatom species.

The similarity of diatom assemblages within and between bioclimatic zones was assessed

using the program PRIMER v5.2.9 (Clarke and Gorley 2001). An analysis of similarities

(ANOSIM, based on a Bray-Curtis similarity matrix) was used to determine whether the diatom

species assemblages were significantly different between pairs of bioclimatic zones. The

PRIMER program SIMPER (similarity of percentages) was used to determine which diatom

species were most typical of each zone.

Environmental variables

In order to assess which measured environmental variables were most closely related to

diatom species distributions, we used a series of constrained multivariate ordinations. Any

environmental variable that was below the method detection limit in >50% of the sites was

removed from further analysis. A Pearson correlation matrix was then used to assess pairs of

significantly correlated variables; this information was also used to truncate the dataset by

removing highly collinear variables. This subset of variables was subsequently entered into a

canonical correspondence analysis (CCA) to determine which of the measured environmental

variables best explained the diatom species distributions. All ordinations were performed in

CANOCO 4.5 (ter Braak and Šmilauer 2002).

Spatial variables

The latitude and longitude of each site were converted to x,y coordinates (in metres),

based on the Lambert Canadian Conic projection to reduce the distortion found at high latitudes

in the more commonly used Universal Transverse Mercator coordinate system (Lo and Yeung

2007). These coordinates were used to create a dataset of spatial variables derived from principal

coordinates of neighbour matrices (PCNM) using the program SpacemakeR in R Statistical

Environment, following the methods of Dray et al. (2006). Distance-based eigenvector maps

were used to create matrices of spatial variables according to predefined criteria for “neighbours”.

Because it was impossible to determine a priori which neighbourhood model would be most

realistic, we tested five different models (based on Delaunay triangulation, Gabriel graph, Sphere

of Influence, Relative Neighbourhood, and a distance-based criterion which listed sites as

neighbours only if they were within 50 km of each other). Based on previous concepts regarding

the dispersal of diatoms, we felt that the most connected model (i.e. based on Delaunay

triangulation) was likely the most realistic, but we nevertheless examined all of the models.

SpacemakeR generated a series of spatial variables that were subsequently used in a CCA

with forward selection to determine what proportion of the variance in species distributions could

be explained by spatial structure. Finally, both the environmental data and spatial data were used

together, using the program VARCAN (Peres-Neto et al. 2006), to assess the relative importance

of the measured environmental variables and the spatial variables. VARCAN is capable of

performing an adjustment for bias correction, as well as assessing the significance of the

proportion of variation explained by both environmental and spatial (Peres-Neto et al. 2006).

Exploratory statistical analysis and transfer functions

Variables identified as explaining significant portions of the diatom species variance

were each run in individually constrained CCAs in order to determine the explanatory power of

each particular variable and examine which variables were best suited to the construction of a

diatom transfer function. This resulted in a subset of variables that explained significant portions

of the diatom variance individually; each of these was then tested with a detrended CCAs

(DCCA) to assess whether linear or unimodal techniques were most appropriate.

As the DCCAs indicated gradient lengths of > 2 standard deviations, we used unimodal

techniques (Lepš and Šmilauer 2003) to assess the possibility of building transfer functions to

reconstruct pH and specific conductivity (the two most ecologically significant variables shaping

diatom distributions). We examined both weighted averaging (WA) and weighted averaging

partial least squares (WA-PLS) techniques, in the program C2 v1.43 (Juggins 2003) to assess

which model was most appropriate for the given environmental variable. We included taxa that

reached either at least 1% relative abundance in a minimum of three sites or greater than 10%

relative abundance in at least one site. We also screened taxa to assess whether or not they had:

a) any significant response, or b) a significant unimodal response, using the program Huisman-

Olff-Fresco (HOF v2.3; www.helsinki.fi/~jhoksane/, Huisman et al. 1993).

Results and Discussion

Ecological characteristics of diatoms from Melville Island

Although our initial dataset consisted of 46 ponds and lakes, no diatoms were found in

site MVV (discussed below). This resulted in a subset of 45 lakes and ponds, from which a total

of 293 species were identified. Just under half of all taxa identified occurred in less than 1%

relative abundance and were removed from the dataset. All ordinations and statistical analyses

are based on the 164 taxa that occurred in at least 1% relative abundance in at least one site. Table

2 lists all taxa present in at least 1% relative abundance in at least three sites, or at least 10%

relative abundance in at least one site.

No diatoms were found in site MVV, a fairly large lake located in a poorly vegetated

catchment that was mostly ice-covered at the time of sampling. The persistent ice cover on this

lake was not unique to site MVV, as many other lakes on Melville Island were similarly covered

with ice during our field season of July 2002. What is particularly interesting to note, however, is

the extremely low SiO2 concentration (0.07 mg/L) found in MVV. In temperate systems, diatoms

are known to become less abundant when silica concentrations in the water column fall below 0.5

mg/L (Wetzel 2001). Thus, it is possible that the extremely low silica in MVV may have been

partially responsible for restricting diatom growth in this pond. However, other factors must also

have been acting in concert to exclude diatoms from this site as the rest of the water chemistry

was relatively unremarkable compared to some other sites on Melville Island, some of which

recorded even lower SiO2 concentrations. Perhaps a combination of persistent ice cover, low

nutrients, and low SiO2 precluded diatoms from this site.

Our survey revealed several diatom taxa common among our study sites (Fig. 2), which

have also been recorded elsewhere throughout the Canadian Arctic. For example, Nitzschia

perminuta, found in almost every lake sampled on Melville Island, is one of the most ubiquitous

diatoms of the Canadian Arctic (Antoniades et al. 2005; Lim et al. 2007). Other common taxa

found on Melville Island include Fragilaria capucina, Achnanthidium minutissimum,

Psammothidium marginulatum, and Chaemaepinnularia soehrensis (Fig. 2). These species are

all relatively small, benthic diatoms that are common in circumneutral pH habitats in High Arctic

freshwater ecosystems including Alert, Ellesmere Island, Axel Heiberg Island, Cornwallis Island,

and Devon Island (Lim 2004; Antoniades et al. 2005; Michelutti et al. 2006; Michelutti et al. in

press).

Although the diatom species were overwhelmingly dominated by benthic taxa, one site

(MVO) was characterized by high relative abundances of small, planktonic Cyclotella spp., which

have been linked to reduced ice cover and stronger stratification regimes in Subarctic lakes (e.g.

Sorvari et al. 2002; Rühland et al. 2003). Interestingly, MVO is a fairly large lake located in a

lush grassy area that would be predicted to record increases in Cyclotella spp. with warmer

temperatures.

A few lakes within this calibration set were also notable due to their location near high

elevation ice caps and their very low specific conductivity (~5 µS/cm) and nutrients (MVAE,

MVAF; Keatley et al. [2]). These high elevation sites were characterized by diatom species such

as Caloneis aerophila, Eunotia spp., Staurosira construens var. venter, F. capucina and

Achnanthes spp. (especially Psammothidium helveticum and A. rupestris). Elsewhere in the High

Arctic, many of these species have been recorded in sites with relatively low pH (<7, especially

C. aerophila and Eunotia spp., Antoniades et al. 2004) and low specific conductivity (e.g. C.

aerophila, P. helveticum, F. capucina, Antoniades et al. 2005). A. rupestris has previously been

reported as an uncommon taxon, yet when present can reach high abundances (e.g. >25% relative

abundance in a given site, Antoniades et al. 2004). However, previous surveys have reported

widely divergent specific conductivity (603 µS/cm, Antoniades et al. 2004; 66 µS/cm, Michelutti

et al. 2006) and pH optima for this species (pH: 6.4, Antoniades et al. 2004; 7.9, Michelutti et al.

2006). On Melville Island, A. rupestris reached very high relative abundances (>50%) in these

high elevation and extremely low specific conductivity sites (5 µS/cm), suggesting that this taxon

has a low specific conductivity optimum.

Differences between bioclimatic zones

Significant differences in diatom species assemblages between bioclimatic Zone 4 (the

most lush zone) and each of the other zones were identified using ANOSIM (Table 3). The

largest differences were found, not surprisingly, between Zone 4 (the most lush) and Zone 1 (the

least lush zone, Table 3). There were no significant differences between diatom assemblages in

zones 3, 2, and 1 (Table 3). The diatom species data track the general trend of differences in

water chemistry between these bioclimatic zones (Keatley et al. [2]). As climatic factors

undoubtedly influence water chemistry, it is difficult to completely untangle these signals.

The program SIMPER was used to determine which species within each zone contributed

to both: a) the similarity within a zone, and b) the dissimilarity between zones. Interestingly, taxa

with the highest relative abundances in each zone were remarkably similar, with Nitzshia

perminuta, for example, reaching an average of between 6-10% relative abundance in each zone,

although it was least abundant in zone 1. Likewise, Psammothidium marginulatum was quite

common in all zones. The differences in species assemblages between zones are therefore being

influenced by the relatively rare species found within each zone. The results from SIMPER also

indicate that sites located in zone 4 have more species in common (average similarity = 35%)

than those located in zone 1 (average similarity = 14%).

Although each zone was typified by similar taxa, the species contributing most to the

dissimilarity between zones provides insights in to the characteristics of each zone. Because the

greatest dissimilarities were found between Zone 4 and Zone 1, we restrict our discussion to the

taxa from these two zones. The species found in Zone 4 that contributed the most to the

dissimilarity from Zone 1 are listed in Table 4, and include several Cymbella spp., Achnanthes

spp., Nitzschia spp., Navicula spp., and Denticula kuetzingii. Zone 4, by definition a lushly

vegetated region, was characterized by sites with higher pH, DOC, and total dissolved nitrogen

relative to the sites in other zones (Keatley et al. [2]). Based on water chemistry, Zone 4 sites

most resemble those of Mould Bay, Prince Patrick Island (Antoniades et al. 2003) and Banks

Island (Lim et al. 2005). Not surprisingly, the diatom assemblages between these three regions

also share some similarities. The diatoms of the Mould Bay sites, for example, were notable for

their higher relative abundances of Cymbella and Achnanthes spp. (Antoniades et al. 2005), as

compared to other High Arctic diatom calibration sets. Likewise, on the very lush Banks Island,

sites characterized by higher total nitrogen typically contained Cymbopleura angustata var.

spitzbergensis and Denticula kuetzingii (Lim et al. 2007), both species that contributed to the

dissimilarity between Zone 4 and Zone 1 on Melville Island.

Species from Zone 1 sites that contributed to the dissimilarity from Zone 4 sites included

Encyonema silesiacum, F. capucina, and Navicula vulpina. While these species have not

previously been noted for characterizing particularly poorly-vegetated regions, they are common

components of High Arctic diatom studies (e.g. Lim et al. 2001a; Antoniades et al. 2004;

Antoniades et al. 2005; Michelutti et al. 2006). However, due to the relatively few taxa shared

between the sites located in Zone 1, the use of the species mentioned above as discriminators is

somewhat limited.

Previous studies have suggested that small, benthic Fragilaria taxa (e.g., Staurosira

venter and Staurosirella pinnata, synonyms F. construens var. venter and F. pinnata,

respectively) are common pioneering species, found in cold, dilute, nutrient-poor systems (e.g.

Douglas and Smol 1999; Lim et al. 2007). On Melville Island, these taxa, although present in

zone 1, were also found in very high relative abundances in Zone 3 sites with low silica (< 0.2

mg/L; MVO, MVQ, MVAS). The presence of these small, benthic Fragilaria sensu lato taxa in

these sites provides evidence of their opportunistic nature; perhaps these taxa are able to exploit

habitats that are not commonly favourable for most other diatom species, whether the harsh

conditions are due to cold temperatures, low light, limited availability of nutrients or, as in this

case, low silica. Recent physiological studies of nutrient limitation of S. pinnata from alpine

lakes supports this hypothesis, in that S. pinnata were found to have low nutrient requirements for

N, P, and Si (Michel et al. 2006).

Multivariate ordinations

In order to determine the relative influence of both measured environmental variables and

spatial variables on diatom species distributions, we used a series of exploratory CCAs.

Interestingly, spatial variables explained very little of the variance of diatom species on Melville

Island (Appendix 2). Based on the CCA, individual spatial variables were sometimes selected as

explaining significant proportions of the diatom variation, depending on the spatial

neighbourhood model chosen. However, all spatial variables were less important than the

environmental variables of pH, specific conductivity and surface area. Indeed, when the

significance of the environmental versus spatial relationship was assessed (Peres-Neto et al.

2006), spatial variables could only explain a significant amount of the variation in the diatom

species data in one out of the five models tested (i.e. when sites located < 50 km apart were

designated as neighbours, Appendix 2). Although Beisner et al. (2006) examined the role of

environmental versus spatial variables on the structure of phytoplankton communities, no

previous diatom surveys, to our knowledge, have assessed the relative significance of PCNM-

derived spatial versus environmental factors in determining diatom species distributions, making

comparisons impossible. Our results suggest, however, that assessing measured environmental

variables is likely capturing most of the variance in diatom species distributions and that dispersal

related factors are not important for diatoms, at least at the landscape scale identified in this

study. Our data support the hypothesis that diatoms in Arctic regions do not experience dispersal

limitations, even in the absence of human-mediated transport.

The canonical correspondence analysis (CCA) indicated that five environmental variables

could significantly explain ~25% of the variance in diatom species (0.944 out of 3.789). All four

axes were significant as tested by a series of partial CCAs each constrained with the sample

scores as co-variables. Axis 1 (λ = 0.319) of the CCA explained 8.4% of the variation in the

diatom data and represents gradients of pH, surface area, and chlorophyll a (Fig. 3). Axis 2 (λ =

0.255) explained 6.7% of the variation in the diatom data and represents gradients of specific

conductivity and elevation (Fig.3).

Diatom species characteristic of higher pH and specific conductivity plot in the upper left

quadrant of the CCA, and include Amphora spp. (A. spitzbergensis, A. copulata, A. inariensis),

Caloneis silicula, Denticula tenue, D. kuetzingii, many Achnanthes spp. (Achnanthidium kriegeri,

Psammothidium bioretti, Eucocconeis flexella, Eucocconeis laevis,), Cymbella spp. (Cymbella

cleve-eulerae, Cymbopleura angustata var. spitzbergensis, Encyonopsis descripta, Cymbella

designata, Encyonema fogedii, Cymbella subaequalis), Navicula spp. (N. cryptocephala, N. sp. cf

chiarae, N. vulpina), Pinnularia balfouriana (synonym Hygropetra balfouriana), and Nitzschia

spp. (N. inconspicua, N. frustulum). Species plotting in the bottom right quadrant of the CCA are

more common in lower pH, lower specific conductivity and higher elevation sites on Melville

Island and include the only Cyclotella spp. (C. sp. aff. comensis, C. stelligera), small Achnanthes

spp. sensu lato (Psammothidium broenlundensum, P. helveticum, A. ingratiformis, P.

marginulatum, Rossithidium petersenii, P. ventralis), Eunotia praerupta, Fragilaria capucina,

Staurosirella pinnata, Cavinula pseudoscutiformis, Nitzschia perminuta, Pinnularia spp. (P.

grunowii, P. krammeri), and Stauroneis anceps.

Of the measured environmental variables, pH, specific conductivity and surface area

were the most important for explaining diatom variance (p < 0.01). This is a finding common to

other diatom surveys in the Canadian High Arctic (e.g. Antoniades et al. 2004; Antoniades et al.

2005; Michelutti et al. 2006), and is not unexpected as pH and related variables have been

described as the single most important controlling variable on diatom species composition in

freshwater systems (Battarbee et al. 2001). When based on water chemistry data alone, the sites

located within bioclimatic Zone 4 (the most lushly vegetated zone) also had higher total dissolved

nitrogen and related variables (e.g. DOC) than the other sites (Keatley et al. [2]). However, this

distinctiveness was not captured in the diatom data, which did not show a significant response to

any nutrients or related variables when all sites were included together in the CCA. This is likely

due to the overriding influence of pH and specific conductivity on many diatom species.

As the sites from Zone 4 contained significantly different diatom species assemblages

compared to each of the other zones based on our ANOSIM results (see above), we performed an

additional ordination restricted to only Zone 4 sites. In this case, the short gradient length of the

DCA (1.9 standard deviations) indicated that linear methods, such as redundancy analysis (RDA),

would be most appropriate. The RDA with forward selection identified total dissolved nitrogen

as the only significant explanatory variable in explaining diatom abundances in Zone 4 sites (p <

0.01, data not shown). The importance of total dissolved nitrogen as an explanatory variable for

diatom species abundances is similar to the relationships found in the more southerly and lush

Banks Island (Lim et al. 2007), and the well-vegetated, High Arctic Bathurst Island (Lim et al.

2001a). Interestingly, Antoniades et al. (2005) did not find a significant relationship between

diatom assemblages and nutrients (i.e. N, P) in their relatively lush Mould Bay ponds; instead,

DOC explained a significant portion of the variance. The results from Zone 4 suggest that, in

Arctic environments that are warmer and more favourable for vegetation growth, different factors

may control diatom species assemblages as compared to those from other regions.

Quantitative relationships between diatom species and environmental variables

The environmental variables that explained significant proportions of diatom variation

(pH, specific conductivity, surface area, elevation and chlorophyll a) were tested individually in a

detrended CCA (DCCA) to assess whether they could explain significant portions of the variation

in the diatom species data. The DCCA results showed that pH, specific conductivity, surface

area, and elevation each explained significant amounts of the variation in the diatom data (p <

0.01), whereas chlorophyll a did not. However, since pH had the strongest relationship and is

ecologically-relevant for paleoenvironmental reconstructions, we focus here on pH models.

Exploratory models relating specific conductivity to diatom species distributions were also

generated, however, due to the highly structured nature of the residuals and the large error

associated with our conductivity models, we provide them as supplementary material only

(Appendices 3, 4).

Because the gradient length of species turnover with respect to pH and specific

conductivity were relatively long (3.18 standard deviation units), we compared models using the

unimodal modelling techniques of weighted averaging (WA, with and without tolerance

downweighting, as well as with either classical or inverse deshrinking), and weighted averaging

partial least squares (WA-PLS, with 1 to 5 components). All models were validated using the

bootstrapping technique with Monte Carlo permutations (n = 1000). We also examined the

effects of more stringent cut-off criteria for our diatom data by assessing model performance

with: a) all taxa present in at least three sites or in one site if greater than 10% (n = 90); b) taxa

with significant responses to pH (n = 70); and c) only taxa with significant unimodal responses to

pH (n = 46).

None of the WA-PLS models performed significantly better than the WA models, and are

not discussed further. The results of the various WA models are given in Table 5.

The most robust pH model was constructed using WAtol inv for our less stringent

inclusion criteria, with cross-validated r2boot = 0.432 and 0.495 respectively, but when only

species with a significant unimodal response were used (n = 46), WAinv was the best model with

r2boot = 0.738 (Table 5, Fig. 4). Although the model with the more stringent cut-off criteria

produced a stronger relationship, this model also limited the number of species for which pH

optima could be estimated. In any case, our pH models performed as well as, or better, than those

previously generated from Axel Heiberg Island (Michelutti et al. 2006), Isachsen, Ellef Ringnes

Island (Antoniades et al. 2004), Alert, Ellesmere Island and Mould Bay, Prince Patrick Island

(Antoniades et al. 2005). Taken together, these data show that pH is an important explanatory

variable influencing diatom species in the Canadian High Arctic and provide important insights

into the pH preferences of High Arctic diatom taxa. In order to better visualize the relationships

between the most common taxa from Melville Island and pH, species response curves are

provided in Figure 5.

Summary

We characterized the diatom species distributions across Melville Island, NU/N.W.T, a

large geographical region covering five bioclimatic zones and broad environmental gradients in

the western Canadian High Arctic. While the most common species were similar to previously

reported High Arctic species, significant differences in species assemblages occurred between the

most lushly vegetated bioclimatic zone and all other bioclimatic zones. The diatom taxa

contributing to the dissimilarity of the most lush zone from others included species previously

reported to be common in other richly vegetated areas of the Canadian Arctic.

The most important factors explaining diatom species distributions included pH, specific

conductivity, surface area, elevation, and chlorophyll a. Within the most lush bioclimatic zone,

total dissolved nitrogen also explained a significant portion of diatom species distribution,

suggesting that nutrient dynamics may be more important in warmer, more richly vegetated

regions of the Arctic. Spatial variables did not explain a significant portion of the species data,

suggesting that dispersal was not limiting diatom species distributions on Melville Island.

Acknowledgements Funding for this research came from NSERC grants to B.E.K., M.S.V.D., and J.P.S., and an NSTP grant to B.E.K. Logistical support was provided by P.C.S.P. We thank J.R. Glew, N. Michelutti, and D. Antoniades for help in the field, G. Barber and D. Atkinson for help with geographical projections and N. Michelutti and K. Rühland for many useful comments on the manuscript. This is PCSP contribution # (TBA).

References ACIA. 2004. Impacts of a warming Arctic. Cambridge University Press, Cambridge, UK. Available online: http://amap.no/acia/

Antoniades, D., Douglas, M.S.V., and Smol, J.P. 2003. Comparative physical and chemical limnology of two Canadian High Arctic regions: Alert (Ellesmere Island, NU) and Mould Bay (Prince Patrick Island, NWT). Archiv. Hydrobiol. 158: 485-516. doi: 10.1127/0003-9136/2003/0158-0485.

Antoniades, D., Douglas, M.S.V., and Smol, J.P. 2004. Diatom species-environment relationships and inference models from Isachsen, Ellef Ringnes Island, Canadian High Arctic. Hydrobiologia, 529: 1-18. doi: 10.1007/s10750-004-4942-8.

Antoniades, D., Douglas, M.S.V., and Smol, J.P. 2005. Benthic diatom autecology and inference model development from the Canadian High Arctic Archipelago. J. Phycol. 41: 30-45. doi: 10.1111/j.1529-8817.2005.04049.x

Antoniades, D., Hamilton, P.B., Douglas, M.S.V., and Smol, J.P. in press. Freshwater diatoms from the Canadian High Arctic. Iconographica Diatomologica.

Barnett, D.M., Edlund, S.A., and Dredge, L.A. 1977. Terrain characterization and evaluation: an example from eastern Melville Island. Geological Survey of Canada Report 76-023. Ottawa.

Battarbee, R.W., Jones, V.J., Flower, R.J., Cameron, N.G., Bennion, H., Carvalho, L., and Juggins, S. 2001. Diatoms. In Tracking Environmental Change Using Lake Sediments: Terrestrial, Algal, and Siliceous Indicators. Edited by J. P. Smol, H. J. B. Birks and W. M. Last. Kluwer Academic Publishers, Dordrecht, Netherlands. pp.155-202.

Beisner, B.E., Peres-Neto, P.R., Lindstrom, E.S., Barnett, A., and Longhi, M.L. 2006. The role of environmental and spatial processes in structuring lake communities from bacteria to fish. Ecology, 87: 2985-2991. doi: 10.1890/0012-9658(2006)87[2985:TROEAS]2.0.CO;2.

Clarke, K.R. and Gorley, R.N. 2001. Primer v 5.0. Primer-E Ltd., Plymouth Marine Laboratory, Plymouth, UK.

Douglas, M.S.V., and Smol, J.P. 1993. Fresh-water diatoms from High Arctic ponds (Cape Herschel, Ellesmere-Island, NWT). Nova Hedwigia, 57: 511-552.

Douglas, M.S.V., and Smol, J.P. 1995. Periphytic diatom assemblages from High arctic ponds. J. Phycol. 31: 60-69. doi: 10.1111/j.0022-3646.1995.00060.x

Douglas, M.S.V., and Smol, J.P. 1999. Freshwater diatoms as indicators of environmental change in the High Arctic. In The Diatoms: Applications for the Environmental and Earth Sciences. Edited by E.F. Stoermer and J.P. Smol. Cambridge University Press, Cambridge. pp. 227-244.

Douglas, M.S.V., Hamilton, P.B., Pienitz, R., and Smol, J.P. 2004. Algal indicators of environmental change in Arctic and Antarctic lakes and ponds. In Long-term environmental change in Arctic and Antarctic lakes. Edited by R. Pienitz, M.S.V. Douglas and J.P. Smol. Springer, Dordrecht. pp. 117-158.

Dray, S., Legendre, P., and Peres-Neto, P.R. 2006. Spatial modelling: a comprehensive framework for principal coordinate analysis of neighbour matrices (PCNM). Ecol. Model. 196: 483-493. doi: 10.1016/j.ecolmodel.2006.02.015

Edlund, S.A. 1994. The distribution of plant communities on Melville Island, Arctic Canada. In The Geology of Melville Island, Arctic Canada. Edited by R.L. Christie and N.J. McMillan. Geological Survey of Canada, Ottawa. pp. 247-255.

Environment Canada. 2004. Canadian Climate Normals 1971-2000. Environment Canada. http://www.climate.weatheroffice.ec.gc.ca/climate_normals/index_e.html,

Finlay, B.J. 2002. Global dispersal of free-living microbial eukaryote species. Science, 296: 1061-1063. doi: 10.1126/science.1070710

Finlay, B. J., Monaghan, E.B., and Maberly, S.C. 2002. Hypothesis: The rate and scale of dispersal of freshwater diatom species is a function of their global abundance. Protist, 153: 261-273. doi: 10.1078/1434-4610-00103.

Hill, M.O. 1973. Diversity and evenness: a unifying notation and its consequences. Ecology, 54: 427-432.

Hodgson, D.A., and Vincent, J.-S. 1984. Surficial geology, Central Melville Island, Northwest Territories; Geological Survey of Canada, Map 1583A, scale 1:250 000.

Huisman, J., H. Olff, & L. F. M. Fresco, 1993. A hierarchical set of models for species response analysis. Journal of Vegetation Science, 4: 37-46.

Juggins, S. 2003. C2: software for ecological and palaeocological data analysis and visualisation, version 1.3. Department of Geography, University of Newcastle.

Keatley, B. E., Douglas, M.S.V., and Smol, J.P. [2]. Physical and chemical limnological characteristics of 46 lakes and ponds across environmental gradients on Melville Island, High Arctic Canada. Fundamental and Applied Limnology.

Kociolek, J.P., and Spaulding, S.A. 2000. Freshwater diatom biogeography. Nova Hedwigia, 71: 223-241.

Krammer, K. 2002. Diatoms of Europe: Cymbella. A.R.G. Gantner Verlag, Ruggell.

Krammer, K., and Lange-Bertalot, H. 1986-1991. Bacillariophyceae. In Sußwasswerflora von Mittleeuropa. Edited by H. Ettl, J. Gerloff, H. Heynig and D. Mollenhauer. Gustav Fischer Verlag, Stuttgart/Jena.

Kristiansen, J. 1996. Dispersal of freshwater algae - A review. Hydrobiologia, 336: 151-157. doi: 10.1007/BF00010829.

Lepš, J., and Šmilauer, P. 2003. Multivariate analysis of ecological data using CANOCO. Cambridge University Press, Cambridge.

Lim, D.S.S. 2004. Limnology and diatom paleoecology of diatoms on Banks Island, N.W.T., and Devon Island, Nunavut, Canadian Arctic. PhD thesis, Department of Geology, University of Toronto, Toronto, ON.

Lim, D.S.S., Douglas, M.S.V., and Smol, J.P. 2001a. Diatoms and their relationship to environmental variables from lakes and ponds on Bathurst Island, Nunavut, Canadian High Arctic. Hydrobiologia, 450: 215-230. doi: 10.1023/A1017553112643.

Lim, D.S.S., Douglas, M.S.V., and Smol, J.P. 2005. Limnology of 46 lakes and ponds on Banks Island, NWT, Canadian Arctic Archipelago. Hydrobiologia, 545: 11-32. doi: 10.1007/s10750-005-1824-7.

Lim, D.S.S., Kwan, C., and Douglas, M.S.V. 2001b. Periphytic diatom assemblages from Bathurst Island, Nunavut, Canadian High Arctic: An examination of community relationships and habitat preferences. J. Phycol. 37: 379-392. doi: 10.1046/j.1529-8817.2001.03703379.x.

Lim, D.S.S., Smol, J.P., and Douglas, M.S.V. 2007. Diatom assemblages and their relationships to lakewater nitrogen levels and other limnological variables from 36 lakes and ponds on Banks Island, N.W.T., Canadian Arctic. Hydrobiologia, doi: 10.1007/s10750-007-0623-8.

Lo, C.P., and Yeung, A.K.W. 2007. Concepts and Techniques in Geographic Information Systems. Pearson Prentice Hall, Upper Saddle River, NJ.

Michel, T.J., Saros, J.E., Interlandi, S.J., and Wolfe, A.P. 2006. Resource requirements of four freshwater diatom taxa determined by in situ growth bioassays using natural populations from alpine lakes. Hydrobiologia, 568: 235-243. doi: 10.1007/s10750-006-0109-0.

Michelutti, N., Douglas, M.S.V., and Smol, J.P. 2006. Ecological characteristics of modern diatom assemblages from Axel Heiberg Island (High Arctic Canada) and their application to paleolimnological inference models. Can. J. Bot. 84: 1695-1713. doi: 10.1139/BO6-122.

Michelutti, N., Douglas, M.S.V., and Smol, J.P. in press. Evaluating diatom community composition in the absence of marked limnological gradients in the High Arctic: A surface sediment calibration set from Cornwallis Island (Nunavut, Canada). Polar Biology.

Peres-Neto, P.R., Legendre, P., Dray, S., and Borcard, D. 2006. Variation partitioning of species data matrices: Estimation and comparison of fractions. Ecology, 87: 2614-2625. doi: 10.1890/0012-9658(2006)87[2614:VPOSDM]2.0.CO;2.

Rühland, K.M., Smol, J.P., and Pienitz, R. 2003. Ecology and spatial distributions of surface-sediment diatoms from 77 lakes in the subarctic Canadian treeline region. Can. J. Bot. 81: 57-73. doi: 10.1139/BO3-005.

Smol, J.P. 2002. Pollution of lakes and rivers: a paleoenvironmental perspective. Arnold, London.

Smol, J.P., Wolfe, A.P., Birks, H.J.B., Douglas, M.S.V., Jones, V.J., Korhola, A., Pienitz, R., Rühland, K., Sorvari, S., Antoniades, D., Brooks, S.J., Fallu, M.A., Hughes, M., Keatley, B.E., Laing, T.E., Michelutti, N., Nazarova, L., Nyman, M., Paterson, A.M., Perren, B., Quinlan, R., Rautio, M., Saulnier-Talbot, E., Siitoneni, S., Solovieva, N., and Weckstrom, J. 2005. Climate-driven regime shifts in the biological communities of arctic lakes. Proc. Natl. Acad. Sci. U.S.A. 102: 4397-4402. doi: 10.1073/pnas.0500245102.

Sorvari, S., Korhola, A., and Thompson, R. 2002. Lake diatom response to recent Arctic warming in Finnish Lapland. Global Change Biology, 8: 171-181. doi: 10.1046/j.1365-2486.2002.00463.x.

Stoermer, E.F., and Smol, J.P. (eds). 1999. The diatoms: applications for the environmental and earth sciences. Cambridge University Press, Cambridge.

Telford, R.J., Vandvik, V., and Birks, H.J.B. 2006. Dispersal limitations matter for microbial morphospecies. Science, 312: 1015. doi: 10.1126/science.1125669.

ter Braak, C.J.F., and Šmilauer, P. 2002. CANOCO reference manual and CANOdraw for Windows user guide: Software for Canonical Community Ordination (Version 4.5).

Tozer, E.T., and Thorsteinsson, R. 1964. Western Queen Elizabeth Islands, Arctic Archipelago. Geological Survey of Canada Memoir 332. Ottawa.

Wetzel, R. G., 2001. Limnology. Academic Press, San Diego, CA, USA.

Figure captions

Figure 1. Map showing the location of (a) Melville Island in relation to Canada, (b) existing diatom calibration sets in the Canadian Arctic, and (c) the 45 lakes and ponds in this study. The numbers given in (b) correspond to the following studies: 1 = Mould Bay, Prince Patrick Island (Antoniades et al. 2005); 2 = Isachsen, Ellef Ringnes Island (Antoniades et al. 2003b); 3 = Axel Heiberg Island (Michelutti et al. 2006); 4 = Alert, Ellesmere Island (Antoniades et al. 2005); 5 = Cape Herschel, Ellesmere Island (Douglas and Smol 1993, 1995), 6 = Devon Island (Lim 2004); 7 = Cornwallis Island (Michelutti et al. in press); 8 = Bathurst Island (Lim et al. 2001a, b); 9 = Banks Island (Lim et al. 2007). Figure 2. Histograms of the common species found in the surface sediments of 45 Melville Island lakes and ponds. Only species present in at least 1% relative abundance in at least 10 sites are shown, arranged in order of DCA axis 1 species scores. The 45 study sites are also arranged in order of DCA axis 1 sample scores. Measured pH is plotted to the far right of the figure. Figure 3. Canonical Correspondence Analysis (CCA) of a) sites and environmental variables, and b) diatom species and environmental variables from Melville Island. See Table 2 for species code numbers. Figure 4. Relationship between observed and estimated pH values based on the WAtol inv model (n = 90) for a) bootstrapped pH values (r2

boot = 0.432), and b) bootstrapped pH residuals. Figure 5. Species response curves of common species (found in at least 1% relative abundance in at least 10 sites) in relation to pH. The number beside the species name corresponds to the HOF model that best fits the species distribution data (Huisman et al. 1993) and corresponds to the following models: I – no response, II – monotonic with a plateau at a theoretical maximum value, III – monotonic reaching a plateau less than the maximum value, IV – symmetric unimodal, V – skewed unimodal.

Figure 1.

FL

AM

MA

EC

DA

CB

PQ

IA

SG

AT

KS

AG

AK

AH

NA

DR

XT

HO

AA

UA

IJ

AB

YW

AR

AN

AQ

ZA

OA

FA

PA

LA

EA

J0

20

Psammoth

idium

chlid

anos

0Achna

nthes

scoti

ca

020

40

Navicu

lage

rloffi 20

4060

Psammoth

idium

marginu

latum

2040

Diades

misco

ntenta 20Psa

mmothidi

umve

ntrali

s

Nitzsc

hiapu

silla

020Frag

ilaria

capu

cina

0Amphora

inari

ensis

0Navicu

lacry

ptoce

phala

020

40Nitz

schia

perm

inuta

Encyo

nema

minutum

020Nitz

schia

incon

spicu

a

20Rossit

hidium

peter

senn

i

2040

Nitzsc

hiafru

stulum

0Eucoc

cone

islae

vis

020Ach

nanth

idium

minutis

simum

0Cymbe

llacle

ve-eu

lerae

0Navicu

lach

iarae

0Eucoc

cone

isfle

xella

020Nav

icula

soeh

rensis

67

89

pH

Rel

ativ

e ab

unda

nce

(%)

Figu

re 2

.

Fi

gure

3.

A

observed pH6.5 7.0 7.5 8.0 8.5 9.0

WA

toli

nv p

H e

stim

ates

6.5

7.0

7.5

8.0

8.5

9.0

r2boot = 0.432

B

observed pH6.5 7.0 7.5 8.0 8.5 9.0

WA

toli

nv p

H re

sidu

als

-1.0

-0.5

0.0

0.5

1.0

1.5

RMSEP = 0.341

A B

Figure 4.

Psammothidium chlidanos (IV)

0

5

10

15

20

25

30Eucocconeis flexella (IV)

0

2

4

6

8Eucocconeis laevis (V)

0

2

4

6

8

Psammothidium marginulatum (II)

Rel

ativ

e ab

unda

nce

(%)

0

10

20

30

40

50

60

70Achnanthidium minutissimum (II)

0

5

10

15

20

25Rossithidium petersenii (II)

0

2

4

6

8

10

12

14

Cymbella cleve-eulerae (II)

6.5 7.0 7.5 8.0 8.5 9.0

0

1

2

3

4

5Encyonema minutum (IV)

pH

6.5 7.0 7.5 8.0 8.5 9.0

0

2

4

6

8

10Chamaepinnularia soehrensis (IV)

6.5 7.0 7.5 8.0 8.5 9.0

0

4

8

12

16

20

Psammothidium scotica (IV)

0

2

4

6

8

10Psammothidium ventralis (I)

0

4

8

12

16

20Amphora inariensis (IV)

0

2

4

6

8

Figure 5.

Diadesmis contenta (IV)

0

5

10

15

20

25

30

35Fragilaria capucina (I)

0

5

10

15

20

25

30Navicula chiarae (IV)

0

2

4

6

8

10

12

Navicula cryptocephala (I)

Rel

ativ

e ab

unda

nce

(%)

0

2

4

6

8

10Navicula gerloffi (IV)

0

5

10

15

20

25

30

35

Nitzschia frustulum (IV)

6.5 7.0 7.5 8.0 8.5 9.0

0

10

20

30

40

50

Nitzschia inconspicua (I)

0

2

4

6

8

10

12

14

Nitzschia perminuta (II)

pH

6.5 7.0 7.5 8.0 8.5 9.0

0

5

10

15

20

25

30

35Nitzschia pusilla (IV)

6.5 7.0 7.5 8.0 8.5 9.0

0

1

2

3

4

Figure 5. continued.

Tabl

e 1.

Sum

mar

y of

sele

cted

lim

nolo

gica

l cha

ract

eris

tics f

or si

tes f

rom

Mel

ville

Isla

nd.

The

valu

es p

rese

nted

are

mea

n va

lues

for

each

bio

clim

atic

zon

e, w

ith m

ean

valu

es fo

r all

45 si

tes g

iven

at t

he b

otto

m.

Lim

nolo

gica

l det

ails

for e

ach

sam

plin

g si

te a

re d

escr

ibed

in

Kea

tley

et a

l. (2

007

[2])

.

EL

EV

SApH

CO

ND

TE

MP

TP

TdN

CH

La

DO

CD

ICSi

O2

m

asl

ha

μS/c

m° C

elsi

usμg

/L

mg/

Lµg

/Lm

g/L

mg/

Lm

g/L

Zone

4 m

ean

(n=1

5)

34.7

3.6

8.4

211.

78.

813

.7

0.5

0.8

7.9

26.3

0.2

Zone

3 m

ean

(n=1

0)

113.

22.

08.

118

1.9

9.9

28.3

0.

31.

44.

67.

80.

2Zo

ne 2

mea

n (n

=10)

15

1.8

14.8

7.8

136.

49.

725

.3

0.3

1.9

4.9

4.5

0.7

Zone

1 m

ean

(n=1

0)

259.

711

.77.

776

.67.

126

.2

0.2

1.8

3.6

3.4

0.7

All

site

s mea

n (n

=45)

12

6.4

7.49

37.

815

78.

821

.5

0.33

81

5.5

12.3

0.41

All

site

s med

ian

76.2

1.90

08.

174

8.3

13.2

0.

260

14.

78.

40.

20A

ll si

tes m

axim

um

655.

380

.994

8.8

1230

15.5

135.

0 0.

977

10.1

16.3

47.8

3.31

All

site

s min

imum

5.

00.

008

6.8

52.

56.

2 0.

030

<0.1

0.8

0.2

0.02

Tabl

e 2.

Lis

t of d

iato

m sp

ecie

s fou

nd in

Mel

ville

Isla

nd su

rfac

e se

dim

ent s

ampl

es in

>1%

rela

tive

abun

danc

e fr

om a

t lea

st th

ree

site

s, or

>10

% re

lativ

e ab

unda

nce

in a

t lea

st o

ne si

te.

Col

umn

head

ings

are

as f

ollo

ws:

N (t

he n

umbe

r of o

ccur

renc

es),

N2

(the

effe

ctiv

e nu

mbe

r of o

ccur

renc

es b

ased

on

Hill

’s N

2; H

ill 1

973)

, max

% (t

he m

axim

um re

lativ

e ab

unda

nce)

, pH

(est

imat

ed p

H o

ptim

um a

s ba

sed

on W

Ato

l inv

).

Cod

e N

ame

N

Max

%

N2

pH

1 Ps

amm

othi

dium

bio

retti

(Ger

mai

n) B

ukht

iyar

ova

& R

ound

14

4.

87

6 8.

16

2 Ps

amm

othi

dium

bro

enlu

nden

sum

(Fog

ed) H

amilt

on, A

nton

iade

s & S

iver

5

11.9

1 2

8.19

3

Psam

mot

hidi

um c

hlid

anos

(Hoh

n &

Hel

lerm

ann)

Lan

ge-B

erta

lot

32

28.0

6 15

7.

75

4 Eu

cocc

onei

s fle

xella

(Küt

zing

) Cle

ve

23

7.16

11

8.

43

5 Ps

amm

othi

dium

hel

vetic

um (H

uste

dt) B

ukht

iyar

ova

& R

ound

11

19

.06

3 8.

03

6 Ac

hnan

thes

ingr

atifo

rmis

Lan

ge-B

erta

lot

8 8.

00

5 8.

03

7 Ac

hnan

thid

ium

kri

eger

i (K

rass

ke) H

amilt

on, A

nton

iade

s & S

iver

15

5.

47

7 8.

06

8 Ps

amm

othi

dium

kry

ophi

lum

(Pet

erse

n) R

eich

ardt

11

4.

33

7 7.

93

9 Eu

cocc

onei

s lae

vis (

Øst

rup)

Lan

ge-B

erta

lot

23

7.49

12

8.

33

10

Psam

mot

hidi

um m

argi

nula

tum

(Gru

now

) Buk

htiy

arov

a &

Rou

nd

41

59.8

2 10

7.

82

11

Achn

anth

idiu

m m

inut

issi

mum

(Küt

zing

) Cza

rnec

ki

29

22.6

5 17

8.

30

12

Ross

ithid

ium

pet

erse

nii (

Hus

tedt

) Buk

htiy

arov

a &

Rou

nd

29

12.2

1 13

8.

22

13

Achn

anth

es ru

pest

ris K

rass

ke

4 63

.21

2 7.

61

14

Psam

mot

hidi

um ro

ssii

(Hus

tedt

) Buk

htiy

arov

a &

Rou

nd

12

9.14

8

7.83

15

Ps

amm

othi

dium

sp. [

cf. P

. sub

atom

oide

s (H

uste

dt) B

ukht

iyar

ova

& R

ound

] 8

7.79

3

7.99

16

Ps

amm

othi

dium

ven

tral

is (K

rass

ke) B

ukht

iyar

ova

& R

ound

29

13

.88

13

8.01

17

Ad

lafia

sp. [

cf A

. bry

ophi

la (P

eter

sen)

Mos

er &

Lan

ge-B

erta

lot

14

5.21

6

8.35

18

Am

phor

a co

pula

ta (K

ützi

ng) S

choe

man

& A

rchi

bald

12

1.

39

8 8.

29

19

Amph

ora

inar

iens

is K

ram

mer

16

6.

78

8 8.

22

20

Amph

ora

spitz

berg

ensi

s Van

Land

ingh

am

7 20

.67

3 8.

16

21

Cal

onei

s aer

ophi

la B

ock

10

28.4

0 3

7.63

22

C

alon

eis s

ilicu

la (E

hren

berg

) Cle

ve

19

2.80

12

8.

29

23

Cal

onei

s fas

ciat

a (L

ager

sted

t) C

leve

(gro

up 2

) 13

5.

10

8 8.

25

24

Cav

inul

a ja

erne

felti

i (H

uste

dt) D

. Man

n &

Stic

kle

7 9.

83

3 8.

11

25

Cha

mae

pinn

ular

ia g

andr

upii

(Pet

erse

n) L

ange

-Ber

talo

t & K

ram

mer

4

20.9

4 2

7.77

26

En

cyon

opsi

s sp.

[cf.

Nav

icul

a so

ehre

nsis

Kra

sske

] 16

17

.35

9 8.

53

27

Cyc

lote

lla c

omen

sis G

runo

w

1 14

.08

1 n/

a 28

C

yclo

tella

stel

liger

a (C

leve

& G

runo

w) V

an H

eurc

k 2

21.2

4 1

7.94

29

C

ymbo

pleu

ra a

mph

icep

hala

(Nae

geli)

Kra

mm

er

13

2.55

9

8.42

30

C

ymbo

pleu

ra a

ngus

tata

var

. spi

tzbe

rgen

sis K

ram

mer

23

5.

40

11

8.24

31

C

ymbe

lla b

otel

lus (

Lage

rste

dt) S

chm

idt

14

11.3

4 5

8.32

32

En

cyon

opsi

s ces

atii

(Rab

enho

rst)

Kra

mm

er

12

10.2

0 4

8.21

33

C

ymbe

lla c

leve

-eul

erae

Kra

mm

er

25

4.46

19

8.

36

34

Ency

onop

sis d

escr

ipta

(Hus

tedt

) Kra

mm

er

11

5.08

5

8.48

35

C

ymbe

lla d

esig

nata

Kra

mm

er

14

4.31

7

8.44

36

En

cyon

ema

foge

dii K

ram

mer

13

4.

99

7 8.

37

37

Ency

onem

a la

nge-

bert

alot

ii K

ram

mer

11

2.

18

8 8.

10

38

Cym

bopl

eura

lapp

onic

a (G

runo

w) K

ram

mer

5

12.3

7 2

8.05

39

En

cyon

ema

late

ns (K

rass

ke) D

. Man

n 15

6.

71

7 8.

14

40

Ency

onem

a m

inut

um (H

ilse)

D. M

ann

27

8.24

13

8.

14

41

Ency

onem

a si

lesi

acum

(Ble

isch

) D. M

ann

27

3.26

16

8.

21

42

Cym

bella

suba

equa

lis G

runo

w in

Van

Heu

rck

13

2.52

8

8.44

43

D

entic

ula

kuet

zing

ii G

runo

w

13

4.07

8

8.49

44

D

entic

ula

tenu

e K

ützi

ng

3 13

.03

1 8.

42

45

Dia

desm

is sp

. [cf

. D. c

onte

nta

(Gru

now

) D. M

ann]

26

32

.40

10

8.01

46

D

iade

smis

gal

lica

W. S

mith

1

23.7

4 1

n/a

47

Dia

desm

is p

arac

onte

nta

Lang

e-B

erta

lot &

Wer

um

4 6.

46

2 7.

58

48

Dia

tom

a te

nue

Aga

rdh

17

86.2

7 1

7.98

49

D

iplo

neis

sp. [

cf. D

. mar

gine

stri

ata

Hus

tedt

] 19

2.

80

12

8.36

50

D

iplo

neis

ocu

lata

(Bré

biss

on) C

leve

6

2.39

4

8.24

51

Eu

notia

bilu

nari

s (Eh

renb

erg)

Mill

s 8

3.03

4

7.49

52

Eu

notia

bor

eote

nuis

Nör

pel-S

chem

pp &

Lan

ge-B

erta

lot

6 19

.72

3 7.

59

53

Euno

tia sp

. [cf

. E. p

raer

upta

Ehr

enbe

rg]

22

3.20

14

7.

91

54

Frag

ilari

a ca

puci

na D

esm

aziè

res

31

26.0

7 15

7.

97

55

Frag

ilari

a ca

puci

na [c

f. F.

cap

ucin

a va

r. gr

acili

s (Ø

stru

p) H

uste

dt]

14

3.70

8

8.16

56

St

auro

sire

lla p

inna

ta (E

hren

berg

) Will

iam

s & R

ound

9

48.1

9 3

8.00

57

G

omph

onem

a sp

MV

AO

1

10.3

4 1

n/a

58

Gom

phon

ema

sp. a

ff e

xigu

um K

ützi

ng

1 17

.70

1 n/

a 59

N

avic

ula

chia

rae/

cryp

toce

phal

a 2

11.2

2 2

8.64

60

Nav

icul

a sp

. [cf

. N. c

hiar

ae L

ange

-Ber

talo

t & G

enka

l] 21

9.

94

12

8.41

61

N

avic

ula

cryp

toce

phal

a K

ützi

ng

27

8.80

14

8.

14

62

Nav

icul

a sp

. [cf

. Mic

roco

stat

us e

greg

rius

(Hus

tedt

) Lan

ge-B

erta

lot

5 4.

18

4 7.

84

63

Nav

icul

a ge

rlof

fi Sc

him

ansk

i 21

31

.42

3 7.

92

64

Nav

icul

a gr

egar

ia D

onki

n 9

2.33

7

7.91

65

N

avic

ula

ingr

ata

Kra

sske

1

11.1

1 1

n/a

66

Nav

icul

a ph

ylle

pta

Küt

zing

2

36.6

7 2

8.05

67

C

avin

ula

pseu

dosc

utifo

rmis

(Hus

tedt

) D. M

ann

& S

tickl

e 12

5.

36

7 8.

02

68

Sella

phor

a pu

pula

(Küt

zing

) Mer

esch

kow

sky

16

2.18

9

8.26

69

N

avic

ula

sp. [

cf. N

. ven

eta

Küt

zing

] 11

2.

93

7 8.

25

70

Nav

icul

a vu

lpin

a K

ützi

ng

16

3.68

10

8.

41

71

Nei

dium

affi

ne (E

hren

berg

) Pfit

zer

8 8.

57

4 7.

71

72

Nitz

schi

a cl

ausi

i Han

tzsc

h 7

5.81

4

7.78

73

N

itzsc

hia

diss

ipat

a va

r. m

edia

(Han

tzsc

h) G

runo

w

10

1.58

8

8.15

74

N

itzsc

hia

frus

tulu

m (K

ützi

ng) G

runo

w

29

43.2

0 9

8.22

75

N

itzsc

hia

hom

burg

iens

is L

ange

-Ber

talo

t 14

8.

46

4 7.

69

76

Nitz

schi

a in

cons

picu

a G

runo

w

21

12.0

9 7

8.22

77

N

itzsc

hia

pale

acea

Gru

now

8

3.70

5

7.91

78

N

itzsc

hia

perm

inut

a (G

runo

w) M

. Per

agal

lo

37

30.4

1 26

8.

16

79

Nitz

schi

a pu

ra H

uste

dt

18

3.10

12

8.

42

80

Nitz

schi

a pu

silla

Gru

now

18

3.

69

14

8.10

81

N

itzsc

hia

sp. [

cf. N

. suc

hlan

dtii

Hus

tedt

] 7

3.11

6

8.20

82

H

ygro

petr

a ba

lfour

iana

(Gru

now

ex

Cle

ve) K

ram

mer

& L

ange

-Ber

talo

t 5

59.8

7 2

8.43

83

Pi

nnul

aria

inte

rmed

ia (L

ager

sted

t) C

leve

7

5.71

4

7.59

84

Pi

nnul

aria

kra

mm

eri M

etze

ltin

29

2.68

19

8.

05

85

Pinn

ular

ia sp

MV

T 4

10.7

0 1

7.98

86

Pl

anot

hidi

um p

erag

allii

(Bru

n &

Hér

ibau

d) R

ound

& B

ukht

iyar

ova

5 21

.00

1 8.

14

87

Stau

rone

is a

ncep

s Ehr

enbe

rg

31

14.2

9 6

7.87

88

St

auro

neis

sp. [

cf. S

. pro

min

ula

(Gru

now

) Hus

tedt

] 6

2.70

4

7.77

89

Su

rire

lla m

inut

a B

rébi

sson

4

25.7

1 1

7.86

90

Ta

bella

ria

flocc

ulos

a (R

oth)

Kut

zing

14

7.

64

6 7.

75

Table 3. Analysis of similarity (ANOSIM) results. An * following a p value indicates pairs of zones containing significantly different diatom assemblages.

zones R statistic p value 4 vs 3 0.299 0.001* 4 vs 2 0.337 0.002* 4 vs 1 0.447 0.001* 3 vs 2 0.023 0.281 3 vs 1 0.097 0.045 2 vs 1 -0.02 0.59

Table 4. List of species contributing the most to the dissimilarity (Diss/SD; average dissimilarity/standard deviation) between Zones 4 and 1, based on an analysis of similarity percentages (SIMPER). Zone 4 Zone 1 Species Mean % Mean % Diss/SD Cymbella cleve-eulerae 1.85 0.34 1.54 Nitzschia perminuta 10.66 5.92 1.33 Psammothidium chlidanos 1.41 7.15 1.26 Nitzschia pura/palea 1.1 0.21 1.24 Pinnularia krammeri 0.32 0.3 1.22 Achnanthidium minutissimum 6.95 3.06 1.18 Chamaepinnularia soehrensis 6.27 0.02 1.17 Diploneis marginestriata 0.99 0 1.17 Navicula chiarae 3.1 0.61 1.09 Eucocconeis laevis 2 0 1.07 Rossithidium petersenni 1.83 0.45 1.07 Eunotia praerupta 0.51 0.5 1.06 Eucocconeis flexella 2.15 0.35 1.02 Encyonema silesiacum 0.57 0.64 1.01 Cymbopleura angustata var. sptizbergensis 0.87 0.54 0.97 Cymbella amphicephala 0.62 0.25 0.96 Caloneis silicula 0.65 0.24 0.95 Sellaphora bacillum 0.26 0.11 0.93 Fragilaria capucina 2.57 5.02 0.92 Navicula vulpina 0.61 0.67 0.91 Note: average dissimilarity between Zones 1 and 4 = 83.6 %

Table 5. Summary statistics of various weighted averaging models for pH with a) all species >1% relative abundance in three sites or >10% relative abundance in one site (n=90), b) all species with a significant response to pH (n=70), and c) all species with a significant unimodal response to pH (n=46). All values shown are cross-validated using bootstrapping techniques. “RMSEP” is the Root Mean Squared Error of Prediction. a) WAInv WACla WATol Inv WATol Cla

r2boot 0.344 0.368 0.432 0.446

Average bias boot -0.029 -0.039 -0.046 -0.055 RMSEP 0.350 0.372 0.341 0.351

b) WAInv WACla WATol Inv WATol Cla

r2boot 0.402 0.424 0.495 0.508

Average bias boot -0.030 -0.043 -0.046 -0.054 RMSEP 0.340 0.356 0.326 0.334

c) WAInv WACla WATol Inv WATol Cla

r2boot 0.738 0.742 0.746 0.748

Average bias boot -0.031 -0.037 -0.048 -0.057 RMSEP 0.230 0.237 0.242 0.246

CHAPTER 4

LIMNOLOGICAL CHARACTERISTICS OF A HIGH ARCTIC OASIS AND COMPARISONS ACROSS

NORTHERN ELLESMERE ISLAND

BRONWYN E. KEATLEY, MARIANNE S.V. DOUGLAS AND JOHN P. SMOL

This chapter is forthcoming in September 2007. Keatley, B.E., M.S.V. Douglas, and J.P. Smol. 2007. Limnological characteristics of a high Arctic oasis and comparisons across northern Ellesmere Island. Arctic, in press.

ABSTRACT

Rapidly warming temperatures in the Arctic are predicted to markedly alter the limnology of

tundra lakes and ponds. These changes include increases in aquatic production, pH, specific

conductivity, and nutrient levels. However, baseline limnological data from High Arctic regions

are typically restricted to single sampling events or to repeated samplings of a few select sites,

thereby limiting our ability to assess the influence of climatic change. We employ two techniques

to examine the influence of a warmer climate on High Arctic aquatic ecosystems. First we

compare limnological characteristics during July 2003 from 23 ponds and lakes from an

atypically warm High Arctic oasis on Ellesmere Island, to 32 ponds and lakes located across

northern Ellesmere Island, where climatic conditions are much cooler and more typical of High

Arctic environments. Second, we resample 13 sites originally analysed in 1963 to assess the

influence that 40 years of rising temperatures (as documented by meteorological records) have

had on the limnological characteristics of these freshwater ecosystems. The specific conductivity

values, as well as the concentrations of nutrients and related variables (especially dissolved

organic carbon, DOC), from the Arctic oasis sites are amongst the highest yet reported from the

Canadian High Arctic, and are significantly higher than those from the polar desert around

northern Ellesmere Island. Comparison of the modern and historical data indicated that most

oasis sites currently have higher pH than they did in 1963, consistent with documented warming

temperatures.

INTRODUCTION

The Canadian High Arctic is broadly classified as a polar desert due to the limited

precipitation and harsh annual climate of the region (Muc and Bliss, 1977). Given the vastness of

the High Arctic landscape, however, it is not surprising that climate is heterogeneous. Arctic

oases, regions of great biological production and diversity, are associated with greater availability

of local water sources relative to the surrounding polar desert and are generally found at small

scales (often less than 5 km2; Edlund and Alt, 1989). In the Canadian High Arctic, oases have

been identified on Devon Island (including Truelove Lowland, Bliss, 1977a), and on Ellesmere

Island including Eureka (Edlund and Alt, 1989), Alexandra Fiord (Freedman et al., 1994),

Tanquary Fiord and Lake Hazen (Edlund and Alt, 1989). Similar areas occur at Polar Bear Pass

on Bathurst Island, at Sherard Bay on Melville Island, and at Mould Bay on Prince Patrick Island

(Aiken et al., 1999 onwards). However, even among Arctic oases, the oasis of our study area at

Lake Hazen is strikingly warm and lush, particularly given its extreme location north of latitude

80°N.

Arctic oases are of particular interest to ecologists examining the effects of recent

climatic changes as they represent a glimpse of what the more typical polar desert ecosystems

might become under a warmer climate. By assessing the biological, physical and chemical

processes occurring in Arctic oases, we may better recognize the effects of climate change in

other Arctic regions. Because of their ecological importance and their uniqueness in the High

Arctic, Arctic oases have been relatively well-studied compared to their polar desert counterparts.

For example, terrestrial faunal (Bliss, 1977b; France, 1993) and botanical surveys (Muc and

Bliss, 1977; Soper and Powell, 1985; Henry et al., 1990) have been reported from Lake Hazen,

Truelove Lowland, and Alexandra Fiord (botanical only). However, aquatic biological research

from Arctic oases has largely been limited to a few lakes in the Lake Hazen area (zooplankton,

McLaren, 1964; non-diatom algae, Croasdale, 1973; cyanobacteria, Quesada et al., 1999) and to

three lakes at Truelove Lowland (Minns, 1977).

While Arctic oases are largely defined as regions of greater biological production and

diversity, little is known about the baseline limnological conditions that characterize lakes and

ponds from these regions. For example, limited limnological investigations were undertaken on

Truelove Lowland (Minns, 1977), and across northern Ellesmere Island (Hamilton et al. 1994;

Hamilton et al., 2001), including some sites in the oasis at Lake Hazen. More recent aquatic

work on dissolved organic carbon (DOC) and ultra violet (UV) penetration has been conducted

on Skeleton Lake in the Hazen oasis (Laurion et al., 1997). Also near Lake Hazen, a physical and

chemical limnological survey of ponds and lakes was carried out by the Department of Defence

Research Board (DRB, Canada) in 1963, with some additional observations in 1964 (Oliver and

Corbet, 1966). This valuable dataset includes seasonal measurements of important limnological

variables such as pH, specific conductivity, and major ions, but does not provide comparison data

to aquatic systems at similar latitudes outside of the Arctic oasis zone. Nonetheless, this early

1960’s dataset provides important reference data to assess whether these sites have changed over

the past ~40 years, a time of documented climate change in northern Ellesmere Island

(Environment Canada, 2004).

Excluding the oasis region of Lake Hazen on northern Ellesmere Island, previous

limnological survey data have been provided for aquatic systems near Alert, Ellesmere Island

(Antoniades et al. 2003a). Basic limnological data have also been provided for some lakes to the

south of Lake Hazen (Smith, 2002). In addition, detailed limnological analyses have been

undertaken in complex lakes along the northern coast of Ellesmere Island (Gibson et al., 2002;

Van Hove et al., 2006).

Our primary objective in this study is to characterize present-day limnological

characteristics from lakes and ponds on northern Ellesmere Island, including a large number of

sites located within a warm oasis region. Warm conditions have been linked to reduced ice cover,

longer growing seasons, higher pH and conductivity, and enhanced biological production (e.g.,

Douglas and Smol, 1999; Antoniades et al. 2005; Smol et al., 2005). However, these hypotheses

have not yet been tested from sites located on similar bedrock and at comparable latitudes. Hence

our goals are three-fold: 1) to provide baseline limnological data from sites located across

northern Ellesmere Island, both within and outside an Arctic oasis, and to compare these to other

Arctic regions; 2) to examine the hypothesis that oasis sites will have different limnological

characteristics than sites located outside the oasis; and 3) to assess differences between water

chemistry data from 1963 and 2003 from selected oasis sites.

METHODS

Site description

Our sampling regime took place on northern Ellesmere Island, largely, but not

exclusively, within Quttinirpaaq National Park (Fig. 1). Three physiographic regions exist within

the Park: the Grant Land Mountains covering 65% of the Park in the north, the Lake Hazen Basin

surrounding Lake Hazen, and the Hazen Plateau located between Lake Hazen and the southern

edge of Quttinirpaaq National Park (Bednarski, 1994). Four climatic zones can also be delineated

within the Park: 1) a cool marine climate in the northern coastal areas, 2) very cool regions

characterized by high elevation ice caps, 3) a marine climate in the south-eastern portion, and 4) a

continental climate of Lake Hazen and Tanquary Fiord (Thompson, 1994). The north coast

receives the greatest amount of precipitation, and the areas near Lake Hazen receive the least

(Thompson, 1994).

The Hazen Basin region experiences anomalously warm summer conditions due to its

continental location and its placement on the leeward side of the Grant Land Mountains (Gray,

1994). While average July daily temperatures at Eureka and Alert are 5.7˚C and 3.3˚C (1971-

2000 averages), respectively (Environment Canada, 2004), temperatures at the Lake Hazen camp

during our field work in July 2003 reached an average daily maximum of 16˚C with a minimum

as high as 9.6˚C. Average annual precipitation is 75.5 mm at Eureka (1971-2000) and153.8 mm

at Alert (Environment Canada, 2004). The summer melt periods are shortest for the north coast

(~3 weeks), last around 8 weeks near Alert, and last about 10 weeks at Lake Hazen (Thompson,

1994).

When defined by bioclimatic zone, the Lake Hazen region falls in Zone 4 (Edlund and

Alt, 1989), the most diverse botanical region in the high Arctic, which is dominated by shrubs

and sedges, and includes a large number of species (>100) that are typical of more southerly

Arctic locations (Edlund and Alt, 1989). Within the Lake Hazen oasis, however, there are also

some mountain sites that we consider ‘controls’ due to their relatively high elevation and lack of

catchment vegetation. Outside the oasis, study sites are located within a broad range of

vegetation zones, from low diversity Zone 0 sites (unvegetated) to Zone 3 sites (60-100 taxa,

prostrate shrub zone, dominated by Salix arctica and/or Dryas integrifolia; Edlund and Alt,

1989).

Geology

Northern Ellesmere Island is largely underlain by sandstones, limestones, and slates

(Christie, 1957; Christie, 1964). In the most northerly regions along the north coast, Precambrian

gneisses, schists and granitic rock dominate, while volcanic and sedimentary rocks, including

sandstones and limestones, underlie the northern interior regions (Christie, 1964). The north shore

of Lake Hazen, including the Hazen oasis, is composed of Permian, Triassic, Jura-Cretaceous and

Cenozoic sandstone and shale (Christie, 1964).

Sampling techniques

In July 2003, 55 ponds (< 2 m deep) and lakes (> 2 m deep) were sampled around

northern Ellesmere Island (Fig. 1). Of these, 23 sites were located in the Arctic oasis

immediately north of Lake Hazen, hereafter referred to as “oasis sites” and given unofficial

names EP1 through EP24. It should be noted that EP19 is Lake Hazen, and is kept separate from

all analyses due to its very large size (i.e. surface area ~ 54200 ha). Three of these sites (EP22,

23, 24) were located at relatively high elevations of > 850 m above sea level. Therefore, despite

their location in the warm oasis region, they serve as cooler controls within the oasis set. The

remaining 31 sites were selected from around the northern half of Ellesmere Island, to the north,

east, south and west of Lake Hazen (hereafter referred to as the “northern sites”, and given

unofficial names EPA through EPAE).

For each site, latitude, longitude, and elevation measurements were taken using a

handheld global positioning unit, the helicopter altimeter, and topographical maps. Water

temperature was recorded with a hand-held thermometer, and samples for total phosphorus

(unfiltered, TPu), trace metals (aluminum, Al; beryllium, Be; cadmium, Cd; chromium, Cr;

cobalt, Co; copper, Cu; iron, Fe; lead, Pb;, manganese, Mn; molybdenum, Mo; nickel, Ni;

vanadium, V; zinc, Zn; and silver, Ag), and major ions (calcium, Ca; magnesium, Mg; sodium,

Na; potassium, K; chloride, Cl; sulphate SO4) were retrieved using a pre-cleaned 125 mL sample

bottles from ~15 cm depth from the near shore area of each site. We have used identical

sampling techniques and analyses, as well as a similar time frame, to our previous limnological

investigations, allowing us to make comparisons amongst regions (Douglas and Smol, 1994; Lim

et al., 2001; Michelutti et al., 2002a; Michelutti et al., 2002b; Lim and Douglas, 2003; Antoniades

et al., 2003a; Antoniades et al., 2003b; Lim et al., 2005). A detailed description of the

methodologies used for water sampling are given in Appendix 1.

Additional water samples for pH, specific conductivity, filtered nutrients and related

variables (dissolved silica, SiO2; total phosphorus filtered, TPf; soluble reactive phosphorus, SRP;

nitrate-nitrogen, NO3-N; nitrate-nitrite-nitrogen, NO3-NO2-N; ammonia-nitrogen, NH3-N; total

Kjeldahl nitrogen (filtered), TKN; total dissolved nitrogen, TdN; particulate nitrogen, PON;

dissolved organic carbon, DOC; dissolved inorganic carbon, DIC; particulate carbon, POC; and

chlorophyll a, Chla) were taken with 1 L plastic Nalgene® bottles, rinsed three times with

pond/lake water. At base camp, pH and specific conductivity were measured the same day the

samples were obtained using a handheld Hanna pHep 3 meter and a YSI model 33 conductivity

meter, respectively. The dissolved and particulate fractions of the variables described above were

filtered on site following Environment Canada (1994) and details are given in Appendix 1. All

other analyses were performed at the National Water Research Institute (NWRI) in Burlington,

Ontario (Environment Canada), using protocols described in Environment Canada (1994).

Statistical analyses

Data were visually screened to assess normality of distribution using CALIBRATE 1.0

(Juggins and ter Braak, 1992). Any variables that were not normally distributed were

transformed using mostly logX, logX+1 or square root transformations. Variables whose

distributions could not be normalised were run passively in statistical analyses (i.e. they were

plotted onto the biplot after it was produced, and thus did not affect the results). A Pearson

correlation matrix with Bonferroni-adjusted probabilities was performed on the full dataset to

remove those variables which were highly correlated with each other, thereby reducing the

dataset to a more manageable size for ordination analyses.

A Principal Components Analysis (PCA) was run on the reduced dataset (by removing

highly correlated variables) to assess the important limnological gradients in the dataset using the

ordination program CANOCO 4.5 (ter Braak and Šmilauer, 2002).

Canonical Variates Analysis (CVA, also known as linear discriminant analysis), was used

to identify environmental variables that significantly discriminate between clusters of samples (in

this case, our oasis and northern sites) (Lepš and Šmilauer, 2003). Initially, a CVA was run for

each individual variable to assess whether it explained a significant portion of the variation

distinguishing the two groups. Any significant variables were retained. Next, highly correlated

variables were removed, according to the same variables that were kept in the PCA, and another

CVA was performed using forward selection to sequentially choose the most important

explanatory variables, given the presence of the other variables.

Comparison to historical data

The DRB water sampling of sites around Lake Hazen (Oliver and Corbet, 1966),

provides the earliest historical limnological survey data available in the Canadian high Arctic, and

thus provides a unique opportunity to assess changes in water chemistry on a regional scale over

40 years. Using site descriptions and locations from the DRB map, a subset of sites common to

both our study and the DRB study were identified. While we acknowledge that direct

comparisons of pH, specific conductivity and major ion concentrations are difficult due to

differences in both measurement techniques and seasonal sampling dates, we nonetheless make

use of this valuable historical dataset by providing a brief comparison.

RESULTS & DISCUSSION

Physical characteristics

The oasis sites consisted of 19 ponds and 4 small lakes (EP1, EP2, EP3, EP24; median

surface area (SA)oasis = 0.13 hectares). In contrast, less than one-third of the northern sites were

ponds (9 out of 31, median SAnorthern = 6 hectares). As would be expected based on their location

in the oasis and their smaller sizes, the oasis sites were much warmer (mean temp. = 15.7°C) than

the northern sites (mean temp. = 9.1°C). There was no significant elevational differences

between the two groups (meanoasis= 318 m, meannorthern = 289 m).

pH, specific conductivity, and major ions

The oasis and northern site were not significantly different with respect to pH values

(meanoasis = 8.23, meannorthern = 8.20, Tables 1, 2), and were similar to mean pH values elsewhere

in the Canadian Arctic including Devon (Lim and Douglas, 2003) and Bathurst (Lim et al., 2001)

islands, as well as Alert, Ellesmere Island (Antoniades et al., 2003a). The similar pH between

both our two groups of sites and between our study and those from previous surveys (Lim et al.,

2001; Antoniades et al. 2003a; Lim and Douglas, 2003) likely reflects the broadly similar bedrock

common to most of the sites.

Specific conductivity was significantly higher in the oasis sites (mean = 490 μS/cm) than

in the northern sites (mean = 245 μS/cm) (p = 0.022, Table 1, 2). Previous high Arctic

limnological surveys have reported mean specific conductivity ranging from ~100 μS/cm

(Victoria Island, Michelutti et al., 2002a; Bathurst Island, Lim and Douglas, 2003) to up to 405

μS/cm (Ellef Ringnes Island, Antoniades et al., 2003b), although specific conductivities >300

μS/cm generally reflect the influence of sea spray on coastal lakes and ponds (Michelutti et al.,

2002b, Antoniades et al., 2003b). While some of our northern sites included coastal ponds, all

our oasis sites are located inland and thus sea spray cannot be a factor for these elevated specific

conductivity values. In some sites, very high SO4 values contribute to high conductivity both in

the oasis (EP9, a very shallow site) and the northern (EPY, a small coastal site with gypsum

precipitates) datasets (Table 1a, b). Both these sites also had high Ca concentrations, suggesting

that local bedrock may have been important in influencing these values, as Ca and SO4 are known

to be very high in gypsiferous shale (McNeely et al., 1979).

Higher specific conductivities would, however, also be expected in smaller water bodies

under warmer conditions as increased evaporation would cause greater concentration of solutes in

the water column. During the summer months, prolonged solar radiation combined with clear

skies and warm temperatures characteristic of the Lake Hazen basin could result in enhanced

evaporation, further concentrating the solutes within the lakes and ponds. Although we do not

have seasonal data from our field season, previous work at Lake Hazen documented an average

drop in water levels by ~0.4 cm/day throughout the ice-free season (Oliver and Corbet, 1966).

This appears to be the case in our oasis sites; indeed, the subset of cool, poorly vegetated, high

elevation sites within the oasis region (EP22, 23, 24) had much lower conductivities (mean = 84

μS/cm) than the remaining low elevation oasis sites.

Concentrations of major ions (Ca, K, Mg, Na, SO4) were typically greater in oasis sites,

with K being significantly higher (meanoasis = 7.0 mg/L, meannorthern = 1.6 mg/L). Average K

concentrations elsewhere in the high Arctic range from 0.24 mg/L (Victoria Island; Michelutti et

al., 2002a) to 4.6 mg/L (Axel Heiberg Island; Michelutti et al., 2002b). Non-marine derived K is

often associated with exudates from plants (Prentki et al., 1980). As previously discussed, the

inland location of the oasis points to a terrestrial source of K; thus the relatively high

concentrations of K are likely indicative of the more highly vegetated catchments common in the

oasis. Indeed, our high elevation sites were distinctive in that they had an average K

concentration of 0.48 mg/L, less than 6% of that found in the oasis sites. More specifically, Na:K

ratios less than 2:1 may reflect enhanced terrestrial production (McNeely et al., 1979). Therefore,

the low ratio of Na:K (Na:Koasis = 1.5, Na:Knorthern = 3.8) in the oasis sites are likely indicative of

the more developed catchment vegetation. Once again, our high elevation oasis sites had

relatively higher Na:K (1.9) than the other oasis sites, reflecting the sparseness of catchment

vegetation. Average Na:K previously reported from across the Canadian Arctic range from 1.8

(Victoria Island; Michelutti et al. 2002a) to 18.4 (Alert; Antoniades et al., 2003a).

Nutrients and related variables

As expected, nutrients (TPu, TPf, SRP, TdN, TKN) and related variables (DOC, POC,

PON, SiO2) were significantly higher (p<0.05) in the oasis sites than in the northern sites (Table

1). When we compared only the ponds, most nitrogen fractions, as well as DOC and SiO2 were

significantly higher in the oasis. These high concentrations of TP and TdN in the oasis sites

indeed suggest that warmer conditions enhance nutrient export from the catchment into the lake

or pond. NH3 and Chla were not significantly different between zones.

TPu values for oasis sites (meanoasis = 11.3 μg/L) were most similar to those reported

from more southerly locations including Banks Island (18 μg/L, Lim et al., 2005), Bathurst Island

(12.7 μg/L, Lim et al., 2001), and Mould Bay, Prince Patrick Island (16.1 μg/L, Antoniades et al.,

2003a), the latter two of which include sites identified by Aiken et al. (1999 onwards) as potential

polar oases. Banks Island includes freshwater environments that occur in low, mid and high

Arctic ecozones, and Banks Island itself is one of the lushest islands in the Arctic Archipelago

(Lim et al., 2005). These relatively high TPu concentrations for the oasis sites once again are

indicative of their shared characteristics with other relatively warm, productive Arctic regions. It

should be noted, however, that while much higher TPu concentrations have been reported from

Arctic ponds and lakes than those we report for the oasis sites (see Lim et al., 2005 for a

summary), these have been attributed to sediment re-suspension rather than indicating high

production (Antoniades et al., 2003b).

When classified to trophic status based on TPu values (Wetzel, 1983), 48% of the oasis

sites were considered mesotrophic (i.e. TPu 10-30 μg/L, Table 2). TPu concentrations of the

northern sites (meannorthern = 7 μg/L) were more typical of aquatic habitats in the polar desert at

Axel Heiberg Island (mean = 4 μg/L, Michelutti et al., 2002b), Victoria Island (mean = 1.3 μg/L,

Michelutti et al., 2002a) and the Haughton Crater, Devon Island (mean = 3.7 μg/L, Lim and

Douglas, 2003). Only 19% of northern sites were mesotrophic or above (Table 1). The TPu

concentrations of the high elevation oasis sites (mean = 6 μg/L) were, once again, much lower

than those of the oasis area as a whole and even lower than the mean of the northern sites. Based

on the Wetzel (1983) TPu classification, the high elevation northern sites fall into the ultra-

oligotrophic (i.e. TPu < 5 μg/L, EP23, EP24) or oligo-mesotrophic (i.e. TPu 5-10 μg/L, EP22)

category.

Likewise, total N (TN) values for the oasis sites (meanoasis= 1.14 mg/L) exceed the

previously reported averages for Arctic islands (see summary in Lim et al., 2005), but are closest

to those reported from the lush regions of Mould Bay (0.616 mg/L, Antoniades et al., 2003a) and

Banks Island (0.499 mg/L, Lim et al. 2005). The high elevation oasis sites have a mean TN

concentration of 0.206 mg/L; suggesting that these high elevation sites are more similar to the

northern sites (meannorthern = 0.330 mg/L) than to those located within the oasis.

Interestingly, the TN:TPu ratios of the two groups of sites do not differ greatly (TN:TPu

meanoasis = 98, meannorthern = 67) and both groups are clearly limited by P (Downing and

McCauley, 1992). However, when we examine TPu versus TN graphically, we see that there is

little relationship between the two variables in either the full dataset (graph not shown) or in the

northern sites alone (Fig. 2b), but a positive linear relationship between them in the oasis sites

(Fig. 2a). This finding suggests that, in the northern sites, different mechanisms control nitrogen

and phosphorus delivery to the aquatic ecosystems, but that in the oasis sites the cycles of these

nutrients are linked. It is probable that autochthonous production is higher in the oasis sites (e.g.

Quesada et al. 1999). Similar to other High Arctic limnological surveys (see below), there is no

relationship between either TN or TPu to Chla.

Concentrations of DOC in the oasis sites (meanoasis = 17.3 mg/L) exceed the highest

previously reported mean values which were 6.7 mg/L for Mould Bay and 6.1 mg/L for Banks

Island by more than two-fold. DOC concentrations for the northern sites are similar to averages

for most other Arctic limnological surveys (meannorthern = 3.4 mg/L). The subset of the high

elevation oasis sites had even lower DOC than the northern sites (mean = 2.3 mg/L). As most

DOC is derived from catchment vegetation and aquatic mosses, and as the vegetation is much

richer in the oasis than outside, it is not a surprising result. What is especially noteworthy,

however, is the unprecedentedly high DOC concentrations from the oasis sites. This likely

reflects a few ponds that could possibly be considered wetlands due to their very shallow depths

and the mosses, grasses, and sedges growing throughout them.

Previously reported mean SiO2 concentrations for High Arctic lakes and ponds have

ranged from 0.41 mg/L (Melville Island, Keatley et al., [2]; and Mould Bay, Antoniades et al.,

2003a) to 1.69 mg/L (Axel Heiberg Island, Michelutti et al. 2002b). In our study, the oasis sites

had average SiO2 concentrations of 5.35 mg/L, while the northern sites had an average of 1.47

mg/L. While both our zones have high SiO2 concentrations, likely reflective of the bedrock

geology, the oasis sites greatly exceed previously reported Canadian High Arctic SiO2

concentrations. This may be attributable to the increased action of weathering due to enhanced

run-off during late spring snowmelt under these warmer oasis conditions. In addition, because

our SiO2 measurements were taken from unfiltered water samples, the high SiO2 values may also

reflect increased abundance of siliceous algae within the water samples of the more productive

oasis sites.

It is hypothesized that warmer conditions will result in higher nutrients and related

variables (e.g., Douglas and Smol, 1999), and consequently higher biological production. While

terrestrial production was indeed high in the oasis sites, there was no significant difference

between the two zones with respect to Chla, our proxy for autochthonous phytoplanktonic

production (meanoasis = 0.6 μg/L, meannorthern = 0.5 μg/L). Likewise, there was no relationship

between Chla and either TN or TPu, regardless of whether we examine the two zones together or

separately, or whether we examine the oasis sites with or without the high elevation sites. Chla

concentrations have similarly borne little resemblance to other typical indicators of high

production (such as high P and N concentrations) in other Canadian High Arctic limnological

surveys (Michelutti et al., 2002a; Michelutti et al., 2002b; Antoniades et al., 2003a; Antoniades et

al., 2003b; Lim et al. 2005). This has been attributed to discrepancies between measuring Chla in

the water column, whereas most of the primary production occurs in the periphytic habitat

(Vezina and Vincent, 1997; Villeneuve et al., 2001, Bonilla et al. 2005). Indeed, Quesada et al.

(1999), working on some of the same lakes (e.g. Skeleton Lake), have documented some of the

highest standing stocks of phytobenthos yet measured in polar regions.

Statistical results

The PCA ordination biplot of all sites (Fig. 3) indicates two main directions of variation

in the measured environmental data: axis 1 includes nutrients and related variables (TPu, TPf,

DOC, TdN, SiO2) as well as conductivity and major ions, and explains 52.9% of the variation in

the sites. Meanwhile, axis 2 represents a trace metal gradient and explains 16% of the variation

(Fig. 3). For the sake of clarity in the ordination plot (Fig. 3), we have chosen to remove some

highly correlated variables based on the Pearson correlation matrix (Table 3). For example, SiO2

has replaced the highly correlated variables of POC and PON, TdN represents both TKN and

TdN, and the metals U, V, Zn, Co, Cr, Be, Mg, Mn have been removed. The following

ecologically important variables could not be normalized and thus were plotted passively in the

ordination (Chla, DIC, K, SO4, Cl), along with the geograpahical variables (elevation, latitude,

longitude, temperature).

As expected, the oasis sites plot closer to each other than to the northern sites (Fig. 3),

and most of these lie along the left end of axis 1. This once again indicates that conductivity and

nutrients and related variables seem to distinguish the oasis sites even in the presence of all other

measured limnological variables. Some exceptions to this general trend include high elevation

oasis sites (EP22, 23, and 24) that were more dilute and less nutrient rich than most other oasis

sites (see above). These high elevation sites also had persistent ice cover and very little

vegetation in their catchments. The northern sites that plotted closest to our oasis sites on the

PCA (AE, E, H) tended to be small ponds with relatively rich vegetation when compared to the

rest of the northern sites. The two sites that plotted at the low end of axis 1 were the least

nutrient-rich and most dilute in the entire dataset were Lake Hazen, a very large lake, and EPO, a

pond located on top of a high mountain glacier with no vegetation, soil, or even rock in its

watershed.

In an attempt to quantitatively determine the main environmental gradients defining the

oasis and northern zones, a Canonical Variates Analysis (CVA) was performed to identify

environmental variables that could significantly discriminate between clusters of samples. Using

this method, only DOC explained a significant portion of the variation between the oasis and

northern sites (p = 0.001). However, DOC was also highly correlated to many nutrients and

related variables (including TPu, TPf, TdN, TKN, POC, and PON, Table 3, Fig. 3), and thus,

while DOC was the only significant variable retained in the analysis, it represents a number of

correlated water chemistry variables.

Historical data

Some of the sites we sampled at Lake Hazen had been part of a Defence Research Board

limnological study in 1963 (Oliver and Corbet, 1966). These historical data represent the earliest

available quantitative limnological data for the Canadian High Arctic. Instrumental temperature

records from Alert and Eureka, as well as proxy climate indicators from Alexandra Fiord

(Rayback and Henry, 2006) and glacier mass balance records from around north central

Ellesmere Island (Braun et al., 2004), indicate a relatively cool period in the 1960s compared to

the late-1990s and the early 21st century. Temperature records from the DND study indicate

average July 1963 temperatures of 6.6°C (Oliver and Corbet, 1966), compared to an average

temperature of 12.8°C during our field season in July 2003. Since limnological characteristics

such as pH and specific conductivity also change over the course of a growing season in High

Arctic lakes and ponds (Douglas and Smol, 1994), comparisons between the two datasets must be

made with caution. Nevertheless, there are no other Arctic regions with available water chemistry

data from the 1960s, and so a comparison, even at a basic level, is warranted.

Interestingly, in almost all sites, we see a slight increase in pH (Fig. 4a) in 2003 relative

to 1963. By examining the identical sites 40 years apart, we have removed any influence of

differences in geology. In our modern survey, recall that we did not record significant differences

between our pH values in the oasis and northern sites, and that this was likely because of the

over-riding influence of geology. By removing the influence of geology (i.e. resampling the

same sites), we may be more directly tracking limnological differences related to a longer

growing season that would be reflected in the warmer temperatures.

Specific conductivity showed no clear pattern between 2003 and 1963, but instead

appears to be related to sampling date (Fig. 4b). In general, specific conductivity is, not

surprisingly, much higher later in the growing season (Fig. 4b), although this pattern is not

without exception (see EP17, for example). Indeed, seasonal studies both elsewhere in the High

Arctic (Douglas and Smol, 1994), and during 1963 at the Hazen Camp (Oliver and Corbet, 1966)

have noted that specific conductivity increased in the majority of sites over the course of the

summer season due to evaporation. During the 1963 study at Hazen Camp, conductivity

fluctuated over an average range of 500 μS/cm, with some ponds drying up completely over the

course of the ice-free season (Oliver and Corbet, 1966). In our modern comparison, we

conducted our field sampling within a short time window of less than two weeks, and thus we

largely removed the seasonal effect of changes in conductivity. Changes in precipitation regime

would also influence conductivity. Although there has been a significant increase in total annual

precipitation at Eureka, there has been no clear trend in annual precipitation at Alert, the closest

meteorological station, over the last 50 years (Environment Canada, 2007). Concentrations of K

and SiO2 are both higher in most sites in 2003 compared to 1963 (Fig. 4c,d), but Ca, Mg, Na, Cl,

and SO4 all show complex patterns that are similar to those found with conductivity (data not

shown).

SUMMARY & CONCLUSIONS

We provide a limnological survey of aquatic habitats located throughout the diverse landscape of

northern Ellesmere Island and compare these to other High Arctic limnological surveys. The

concentrations of nutrients and DOC reported from the oasis ponds and lakes are amongst, and in

some cases are, the highest yet reported from the Canadian High Arctic. The oasis sites at Hazen

Camp are more similar to oasis sites located at Mould Bay, Prince Patrick Island, Banks Island,

and Bathurst Island (many hundreds of kilometres to the southwest) than to those located within a

few hundred kilometres on Ellesmere Island. Meanwhile, the northern Ellesmere lakes and ponds

from our dataset are more similar to those located within the polar deserts of Alert, Axel Heiberg

Island, and Devon Island.

We compared point samples of limnological characteristics between aquatic habitats

located within an Arctic oasis at Hazen Camp to those located outside this oasis area to determine

if these smaller, warmer water bodies had higher specific conductivity and increased nutrient

concentrations. Our comparisons indicate that smaller sites located in warmer and more lushly

vegetated Arctic regions have distinctive water chemistry, particularly with respect to nutrients

and related variables. In our dataset, these higher concentrations of nutrients and related variables

(particularly DOC and correlated variables) were significant despite differences in latitude,

elevation, and surface area between the oasis and northern sites. Interestingly, the three high

elevation oasis ponds were more similar to the polar desert sites than to the other Arctic oasis

ponds with respect to specific conductivity, and nutrients and related variables.

A comparison of water chemistry from a subset of the oasis sites that were first examined

in 1963 to data we collected in 2003 showed that most sites had higher pH in 2003 than they did

in 1963, consistent with documented warming temperatures. Comparisons of specific

conductivity, however, appear to be more related to sampling date.

In summary, aquatic ecosystems in this Arctic oasis have distinct water chemistry from

those located in the near-by polar desert. We associate this difference to increased catchment

vegetation, greater run-off from the watershed, and enhanced evaporation, all of which can be

linked to the warmer temperatures of the oasis. Under a continued Arctic warming scenario, our

results may represent a preview of how other Arctic freshwater systems might change.

ACKNOWLEDGEMENTS

This project was supported by NSERC grants to BEK, MSVD, and JPS. We thank PCSP for

logistical and field support, NSTP for a field research grant to BEK, and Parks Canada for

allowing us to use the Parks Canada base camp at Lake Hazen. Field sampling assistance was

also provided by S. Arnott. We thank A. Poulain, K. Rühland, N. Michelutti, and W. Vincent, as

well as two anonymous journal reviewers for comments on the manuscript. This is PCSP

contribution number 012-07.

REFERENCES AIKEN, S.G., DALLWITZ, M.G., CONSAUL, L.L., MCJANNET, C.L., BOLES, R.L., ARGUS, G.W., GILLETT, J.M., SCOTT, P.J., ELVEN, R., LEBLANC, M.C., GILLESPIE, L.J., BRYSTING, A.K., SOLSTAD, H., and HARRIS, J.G. 1999 onwards. Flora of the Canadian Arctic Archipelago: Descriptions, illustrations, identification, and information retrieval. Version: 1 October 2005. http://www.mun.ca/biology/delta/arcticf/. Accessed 10 March 2006. ANTONIADES, D., DOUGLAS, M.S.V., and SMOL, J.P. 2003a. Comparative limnology of two Canadian High Arctic regions: Alert (Ellesmere Island, NU) and Mould Bay (Prince Patrick Island, NWT). Archiv für Hydrobiologie 158: 485-516. ANTONIADES, D., DOUGLAS, M.S.V., and SMOL, J.P. 2003b. Physical and chemical limnology of 24 ponds and one lake from Isachsen, Ellef Ringnes Island, Canadian High Arctic. International Review of Hydrobiology 88: 519-538. ANTONIADES, D., DOUGLAS, M.S.V., and SMOL, J.P. 2005. Quantitative estimates of environmental changes in the Canadian High Arctic inferred from diatoms in lake and pond sediments. Journal of Paleolimnology 33: 349-360. BEDNARSKI, J. 1994. Geomorphology. In: Resource description and analysis: Ellesmere Island, National Park Reserve. National Resource Conservation Section, Prairie and Northern Region, Parks Canada, Department of Canadian Heritage, Winnipeg. 124 pp. BLISS, L.C. 1977a. Introduction. In: Bliss, L.C., ed. Truelove Lowland, Devon Island, Canada: A High Arctic ecosystem. Edmonton: University of Alberta Press. 1-12. BLISS, L.C. 1977b. Truelove Lowland, Devon Island, Canada: A High Arctic ecosystem. Edmonton: University of Alberta Press. 714 pp. BONILLA, S., VILLENEUVE, V., and VINCENT, W.F. 2005. Benthic and planktonic algal communities in a high Arctic lake: pigment structure and contrasting responses to nutrient enrichment. Journal of Phycology 41: 1120-1130. BRAUN, C., HARDY, D.R., and BRADLEY, R.S. 2004. Mass balance and area changes of four high Arctic plateau ice caps, 1959-2002. Geografiska Annaler Series A-Physical Geography 86A: 43-52. CHRISTIE, R.L. 1957. Geological reconnaissance of the north coast of Ellesmere Island, District of Franklin, Northwest Territories. Paper 56-9. Ottawa: Department of Mines and Technical Surveys. 40 pp. CHRISTIE, R.L. 1964. Geological reconnaissance of northeastern Ellesmere Island, District of Franklin. Memoir 331. Ottawa: Geological Survey of Canada. 79 pp. CROASDALE, H. 1973. Freshwater algae of Ellesmere Island, N.W.T. Ottawa: National Museums of Canada. 131 pp. DOUGLAS, M.S.V., and SMOL, J.P. 1994. Limnology of high Arctic ponds (Cape Herschel, Ellesmere Island, N.W.T.). Archiv für Hydrobiologie 131: 410-434.

DOUGLAS, M.S.V., and SMOL, J.P. 1999. Freshwater diatoms as indicators of environmental change in the high Arctic. In: Stoermer, E.F. and Smol, J.P., eds. The diatoms: Applications for the environmental and earth sciences. Cambridge: Cambridge University Press. 227-244. DOWNING, J.A., and MCCAULEY, E. 1992. The nitrogen-phosphorus relationship in lakes. Limnology and Oceanography 37: 936-945. EDLUND, S.A., and ALT, B.T. 1989. Regional congruence of vegetation and summer climate patterns in the Queen Elizabeth Islands, Northwest Territories, Canada. Arctic 42: 3-23. ENVIRONMENT CANADA 1994. Manual of analytical methods – National Laboratory for Environmental Testing, Canadian Centre for Inland Waters. Burlington, Canada. 2 vols, 1097 pp. ENVIRONMENT CANADA, 2004. Canadian Climate Normals 1971-2000. Environment Canada. http://www.climate.weatheroffice.ec.gc.ca/climate_normals/index_e.html, Accessed 2 April 2004. ENVIRONMENT CANADA, 2007. Climate data online. http://www.climate.weatheroffice.ec.gc.ca/climateData/canada_e.html Accessed 23 February 2007. FRANCE, R.L. 1993. The Lake Hazen trough - A late winter oasis in a polar desert. Biological Conservation 63: 149-151. FREEDMAN, B., SVOBODA, J., and HENRY, G.H.R. 1994. Alexandra Fiord - An ecological oasis in the polar desert. In: Svoboda, J. and Freedman, B. eds. Ecology of a polar oasis: Alexandra Fiord, Ellesmere Island, Canada. Toronto: Captus University Publications: 1-12. GIBSON, J.A.E., VINCENT, W.F., VAN HOVE, P., BELZILE, C., WANG, X., and MUIR, D. 2002. Geochemistry of ice-covered, meromictic Lake A in the Canadian High Arctic. Aquatic Geochemistry 8: 97-119. GRAY, D.R. 1994. Ecology. In: Resource description and analysis: Ellesmere Island, National Park Reserve. National Resource Conservation Section, Prairie and Northern Region, Parks Canada, Department of Canadian Heritage, Winnipeg. 72 pp. HAMILTON, P.B., LEAN, D.R.S., and POULIN, M. 1994. The physicochemical characteristics of lakes and ponds from the northern regions of Ellesmere Island. In: Hamilton, P.B. ed. Proceedings of the Fourth Arctic-Antarctic Diatom Workshop, Canadian Museum of Nature. Canadian Technical Report of Fisheries and Aquatic Sciences 1957: 57-63. HAMILTON, P.B., GAJEWSKI, K., ATKINSON, D.E., and LEAN, D.R.S. 2001. Physical and chemical limnology of 204 lakes from the Canadian Arctic Archipelago. Hydrobiologia 457: 133-148. HENRY, G.H.R., SVOBODA, J., and FREEDMAN, B. 1990. Standing crop and net production of sedge meadows of an ungrazed polar desert oasis. Canadian Journal of Botany 68: 2660-2667. JUGGINS, S. and TER BRAAK, C.J.F., 1992. CALIBRATE - a program for species-environment calibration by [weighted averaging] partial least squares regression. Environmental Change Research Centre, University College, London.

KEATLEY, B.E., DOUGLAS, M.S.V., and SMOL, J.P. [2]. Physical and chemical limnological characteristics across environmental gradients on Melville Island, Nunavut/N.W.T., High Arctic Canada. Fundamental and Applied Limnology. 168: 355-376. LAURION, I., VINCENT, W.F., and LEAN, D.R.S., 1997. Underwater ultraviolet radiation: development of spectral models for northern high latitude lakes. Photochemistry and Photobiology 65: 107-114. LEPŠ, J. and ŠMILAUER, P. 2003. Multivariate analysis of ecological data using CANOCO. London: Cambridge University Press. 282 pp. LIM, D.S.S., and DOUGLAS, M.S.V. 2003. Limnological characteristics of 22 lakes and ponds in the Haughton Crater region of Devon Island, Nunavut, Canadian High Arctic. Arctic Antarctic and Alpine Research 35: 509-519. LIM, D.S.S., DOUGLAS, M.S.V., SMOL, J.P., and LEAN, D.R.S. 2001. Physical and chemical limnological characteristics of 38 lakes and ponds on Bathurst Island, Nunavut, Canadian High Arctic. International Review of Hydrobiology 86: 1-22. LIM, D.S.S., DOUGLAS, M.S.V., and SMOL, J.P. 2005. Limnology of 46 lakes and ponds on Banks Island, NWT, Canadian Arctic Archipelago. Hydrobiologia 545: 11-32. MCLAREN, I. A. 1964. Zooplankton of Lake Hazen Ellesmere Island and a nearby pond with special reference to copepod Cyclops scutifer Sars. Canadian Journal of Zoology 42: 613-629. MCNEELY, R.N., NEIMANIS, V.P., and DWYER, L. 1979. Water quality sourcebook: A guide to water quality parameters. Ottawa: Minister of Supply and Services Canada. 88 pp. MICHELUTTI, N., DOUGLAS, M.S.V., LEAN, D.R.S., and SMOL, J.P. 2002a. Physical and chemical limnology of 34 ultra-oligotrophic lakes and ponds near Wynniatt Bay, Victoria Island, Arctic Canada. Hydrobiologia 482: 1-13. MICHELUTTI, N., DOUGLAS, M.S.V., MUIR, D.C.G., WANG, X., and SMOL, J.P. 2002b. Limnological characteristics of 38 lakes and ponds on Axel Heiberg Island, High Arctic Canada. International Review of Hydrobiology 87: 385-399. MINNS, C. K. 1977. Limnology of some lakes on Truelove Lowland. In: Bliss, L.C. ed. Truelove Lowland, Devon Island, Canada: A High Arctic ecosystem. Edmonton: University of Alberta Press. 567-586. MUC, M., and BLISS, L.C. 1977. Plant communities of Truelove Lowland. In: Bliss, L.C. ed. Truelove Lowland, Devon Island, Canada: A High Arctic ecosystem. Edmonton: University of Alberta Press. 143-154. OLIVER, D. R., and CORBET, P.S. 1966. Aquatic habitats in a high arctic locality: the Hazen Camp study area, Ellesmere Island, N.W.T. Ottawa: Department of National Defence. 115 pp. PRENTKI, R.T., MILLER, M.C., BARSDATE, R.J., ALEXANDER, V., KELLY, J., and COYNE, P. 1980. Chemistry. In: Hobbie, J.E. ed. Limnology of tundra ponds, Barrow Alaska. Stroudsburg: Dowden, Hutchinson and Ross, Inc., 76-178.

QUESADA, A., VINCENT, W.F., and LEAN, D.R.S. 1999. Community and pigment structure of Arctic cyanobacterial assemblages: the occurrence and distribution of UV-absorbing compounds. FEMS Microbiology Ecology 28: 315-323. RAYBACK, S.A., and HENRY, G.H.R. 2006. Reconstruction of summer temperature for a Canadian High Arctic site from retrospective analysis of the dwarf shrub, Cassiope tetragona. Arctic Antarctic and Alpine Research 38: 228-238. SMITH, I.R. 2002. Diatom-based holocene paleoenvironmental records from continental sites on northeastern Ellesmere Island, high Arctic, Canada. Journal of Paleolimnology 27: 9-28. SMOL, J.P., WOLFE, A.P., BIRKS, H.J.B., DOUGLAS, M.S.V., JONES, V. J., KORHOLA, A., PIENITZ, R., RÜHLAND, K., SORVARI, S., ANTONIADES, D., BROOKS, S.J., FALLU, M.-A., HUGHES, M., KEATLEY, B.E., LAING, T.E., MICHELUTTI, N., NAZAROVA, L., NYMAN, M., PATERSON, A.M., PERREN, B., QUINLAN, R., RAUTIO, M., SAULNIER-TALBOT, E., SIITONENI, S., SOLOVIEVA, N., and WECKSTRÖM, J. 2005. Climate-driven regime shifts in the biological communities of arctic lakes. Proceedings of the National Academy of Sciences of the United States of America 102: 4397-4402. SOPER, J. H., and POWELL, J.M. 1985. Botanical studies in the Lake Hazen Region, northern Ellesmere Island, Northwest Territories, Canada. Ottawa: National Museums of Canada. 67 pp. TER BRAAK, C. J. F., and ŠMILAUER, P. 2002. CANOCO reference manual and CANOdraw for Windows user guide: Software for Canonical Community Ordination (Version 4.5). THOMPSON, W. 1994. Climate. In: Resource description and analysis: Ellesmere Island, National Park Reserve. National Resource Conservation Section, Prairie and Northern Region, Parks Canada, Department of Canadian Heritage, Winnipeg. 78 pp. VAN HOVE, P., BELZILE, C., GIBSON, J.A.E., and VINCENT, W.F. 2006. Coupled landscape-lake evolution in high Arctic Canada. Canadian Journal of Earth Sciences 43: 533-546. VEZINA, S., and VINCENT, W.F. 1997. Arctic cyanobacteria and limnological properties of their environment: Bylot Island, Northwest Territories, Canada (73 degrees N, 80 degrees W). Polar Biology 17: 523-534. VILLENEUVE, V., VINCENT, W.F., and KOMÁREK, J. 2001. Community structure and microhabitat characteristics of cyanobacterial mats in an extreme high Arctic environment: Ward Hunt Lake. Nova Hedwigia 123: 199-224. WETZEL, R. G., 1983. Limnology 2nd Edition. Philadephia: Saunders Publishing, Philadephia. 767 pp.

Figure captions. Figure 1. Location map of northern Ellesmere Island. Inset a) indicates Ellesmere Island within Canada. Inset b) shows the northern sites around Ellesmere Island. The dashed black line denotes the boundary of Quttinirpaaq National Park and the patterned areas within this boundary represent different climate regions based on Thompson (1994). The black star indicates the location of the oasis sites detailed in inset c). Inset c) details the oasis sites just north of Lake Hazen. Figure 2. Plot of total phosphorus unfiltered (TPu) versus total nitrogen (TN) in a) the oasis sites, and b) the northern sites. While there is little relationship between TPu and TN in the northern region, there is a clear positive relationship with TPu and TN in the oasis sites, suggesting that different factors control nutrient cycling within the two regions. Figure 3. Biplot of a principal components analysis (PCA) of measured limnological variables for all sites. Oasis sites are represented by filled circles and northern sites are represented by open circles. Lake Hazen is kept separate due to its extremely large size and is represented by a star. Axis 1 most closely represents nutrients and related variables, pH, and conductivity, and explains 52.9% of the variance in the dataset. Axis 2 most closely represents a gradient of metals and explains 16% of the variance in the dataset. The dashed lines represent variables that were run passively in the ordination. Figure 4. Histograms indicating the change in the values of selected limnological variables in 2003 relative to 1963 for a) pH, b) specific conductivity, c) K, and d) SiO2. Site names with * indicate sites for which identification was approximate.

Figure 1.

Figure 2.

Figure 3.

Figure 4.

Table 1. Summary of selected limnological variables for the northern sites and Lake Hazen. Lake Hazen is isolated at the bottom due to its extremely large size. All sites were sampled in July 2003, specific days are given in the “Date” column. Full details for other limnological parameters (e.g. metals, rare earth elements) are available in Appendix 6. Non-standard abbreviations are as follows: elevation (elev, given in meters above sea level) and surface area (SA). Abbreviations for other parameters are given in the text.

ID Name Latitude Longitude Date elev SA Pond or pH condN W m ha Lake μS/cm

EPA Craig 81˚50.34’ 68˚51.13’ 12 152 1455.0 L 8.47 296EPB Appleby 81˚50.90’ 68˚15.49’ 12 366 48.8 L 8.23 232EPC Brainard 81˚45.81’ 68˚10.32’ 12 640 18.0 L 8.03 169EPD 81˚46.78’ 64˚32.96’ 12 274 49.1 L? 8.37 172EPE 81˚42.84’ 70˚05.33’ 12 259 1.1 P 9.00 379EPF 82˚25.20’ 68˚12.77’ 12 91 125.0 L 8.13 110EPG 82˚36.21’ 68˚12.25’ 12 91 14.7 L 8.20 122EPH 82˚36.23’ 68˚13.03’ 12 91 0.4 P 8.63 322EPI 82˚55.11’ 66˚51.88’ 12 15 2.4 P 8.53 75EPJ Ward Hunt 83˚05.30’ 74˚09.87’ 13 61 35.0 L 8.30 45

EPK Lake A 83˚00.74’ 75˚23.28’ 13 30 490.0 L 7.90 69EPL 82˚58.60’ 75˚24.70’ 13 46 99.0 L 8.27 56

EPM 82˚58.54’ 75˚11.44’ 13 213 0.3 P 8.63 102EPN Lake C2 82˚49.59’ 77˚56.24’ 13 30 165.0 L 7.73 30EPO 82˚16.00’ 77˚53.32’ 13 1006 6.3 L 7.47 45EPP 81˚36.17’ 73˚53.11’ 13 686 6.0 L? 8.17 102EPQ 81˚27.28’ 67˚22.21’ 17 579 2.4 P 8.87 138EPR 81˚18.71’ 65˚34.92’ 17 183 14.7 L 8.30 118EPS 81˚19.41’ 66˚25.08’ 17 335 2.6 L 8.33 131EPT 81˚47.53’ 70˚26.77’ 17 457 0.2 P 8.37 135EPU Carolyn 81˚17.96’ 70˚43.48’ 17 305 161.0 L 8.30 119EPV Nan 81˚13.15’ 72˚19.46’ 17 305 6.7 L 8.77 409EPW 81˚04.89’ 74˚20.67’ 17 518 114.0 L 8.30 109EPX 80˚55.91’ 76˚32.69’ 17 396 61.4 L 8.30 154EPY 80˚36.43’ 79˚42.96’ 17 1 0.1 P 8.23 1500EPZ 81˚00.50’ 78˚13.03’ 17 244 4.1 L 8.47 160

EPAA Kettle 81˚23.80’ 76˚47.14’ 18 200 7.7 L 8.73 500EPAB 81˚58.49’ 80˚04.14’ 18 122 0.6 L? 7.90 58EPAC 82˚05.44’ 81˚50.39’ 18 76 98.2 L 8.20 200EPAD 82˚05.79’ 82˚34.70’ 18 75 0.1 P 8.73 325EPAE 81˚42.78’ 82˚17.04’ 18 808 0.5 P 8.30 1200mean 279 96.5 8.20 245

median 213 7.7 8.30 135max 1006 1455.0 9.00 1500min 1 0.1 7.47 30

EP19 Lake Hazen 81˚49.37’ 71˚20.18’ 15 154 54200 L 7.73 68

Table 1. Continued.

ID Chla DOC DIC POC PON NH3 TKN TdN TN TPu TPfμg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L μg/L μg/L

EPA 0.05 3.1 31.4 0.204 0.030 0.016 0.179 0.191 0.216 13.7 2.3EPB 0.05 4.7 36.3 0.246 0.030 0.021 0.383 0.361 0.4155 6.7 2.2EPC 0.1 3.6 27.4 0.256 0.030 0.021 0.315 0.327 0.35 7.4 3.3EPD 0.7 3.3 24.4 0.231 0.029 0.036 0.287 0.293 0.3185 19.0 2.9EPE 0.6 12.6 47.2 0.509 0.047 0.031 0.774 0.752 0.826 8.5 5.8EPF 0.1 1.5 7 0.208 0.024 0.013 0.044 0.077 0.097 6.3 1.6EPG 0.1 0.8 14 0.403 0.039 0.008 0.054 0.095 0.14 2.9 2.1EPH 0.6 5.9 23.5 0.438 0.050 0.043 0.467 0.469 0.5195 10.9 9.8EPI 0.7 2.6 10.8 0.389 0.052 0.008 0.133 0.183 0.233 6.7 2.3EPJ 0.6 1 4.4 0.163 0.017 0.067 0.148 0.217 0.187 4.1 3.8

EPK 0.6 1.3 3.6 0.180 0.027 0.016 0.078 0.089 0.113 3.2 2.2EPL 0.6 1.4 8.5 0.187 0.026 0.044 0.208 0.178 0.249 4.1 3.0

EPM 1 4.3 16.1 0.345 0.042 0.040 0.378 0.405 0.4225 10.3 5.1EPN 0.6 2.4 2.8 0.145 0.021 0.008 0.049 0.067 0.0725 7.4 2.2EPO 0.5 0.7 3.2 0.083 0.016 0.027 0.065 0.128 0.134 1.8 1.4EPP 1.4 4.7 11.9 0.415 0.043 0.023 0.398 0.412 0.452 6.5 3.8EPQ 0.7 12.3 24.4 1.670 0.137 0.048 0.997 1.010 1.143 9.5 6.8EPR 0.05 1.1 17.9 0.357 0.021 0.007 0.056 0.064 0.089 2.4 1.6EPS 0.5 2.2 21.2 0.285 0.012 <0.005 0.068 0.107 0.118 2.1 1.5EPT 0.1 0.9 19.5 0.116 0.021 0.005 0.053 0.413 0.457 1.0 1.4EPU 0.6 0.9 19.3 0.324 0.038 0.005 0.040 0.057 0.098 2.1 1.0EPV 0.1 5 59.5 0.473 0.055 0.020 0.458 0.408 0.522 5.0 3.6EPW 0.05 0.6 14 0.343 0.042 0.030 0.084 0.159 0.205 2.5 1.3EPX 0.05 0.9 25.9 0.288 0.031 0.013 0.091 0.103 0.134 2.6 2.2EPY 0.2 2.2 15.5 0.611 0.059 0.035 0.234 0.159 0.2955 33.5 3.7EPZ 0.05 2.3 26.9 0.230 0.030 0.009 0.174 0.175 0.209 3.4 1.9

EPAA 0.5 4.8 23.3 0.292 0.042 0.043 0.403 0.395 0.453 11.7 4.7EPAB 1.6 1.7 4.8 0.221 0.027 0.005 0.080 0.079 0.1095 3.8 1.0EPAC 0.1 1 25.4 0.283 0.039 0.005 0.068 0.179 0.231 4.8 1.1EPAD 0.6 4.7 36.5 0.441 0.045 0.015 0.336 0.351 0.439 5.2 3.6EPAE 1.5 11.5 22.3 0.430 0.043 0.012 0.936 0.882 0.985 8.1 8.3mean 0.5 3.4 20.3 0.347 0.038 0.022 0.259 0.283 0.33 7.0 3.1

median 0.5 2.3 19.5 0.288 0.031 0.016 0.174 0.183 0.23 5.2 2.3max 1.6 12.6 59.5 1.670 0.137 0.067 0.997 1.010 1.14 33.5 9.8min 0.05 0.6 2.8 0.083 0.012 <0.005 0.040 0.057 0.07 1.0 1.0

EP19 0.05 1.1 9 0.156 0.016 0.0025 0.040 0.060 0.047 2.0 2.1

Table 1. Continued.

ID SRP SiO2 Ca K Na Mg Cl SO4 Al Feμg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L μg/L μg/L

EPA 0.5 2.61 44.70 2.44 4.28 17.90 2.39 56.20 25.1 56.0EPB 0.2 0.37 30.10 3.53 5.34 17.60 3.20 6.31 1.6 12.4EPC 0.6 0.73 26.50 1.03 1.61 8.16 0.96 5.14 3.0 17.3EPD 0.7 1.15 31.10 0.46 3.18 3.89 7.11 1.58 3.7 25.9EPE 0.8 3.21 19.10 9.75 11.20 46.70 15.10 52.40 7.3 36.1EPF 1.7 0.49 18.30 0.32 0.69 3.04 1.05 39.70 177.0 374.0EPG 0.6 0.88 20.50 0.26 0.20 5.76 0.29 20.70 30.3 48.2EPH 0.9 2.21 52.90 0.78 0.78 19.60 0.54 120.00 6.9 31.8EPI 1.9 0.47 12.30 0.11 0.42 3.23 1.00 4.29 2.7 18.9EPJ 0.7 0.16 5.53 0.07 0.23 0.73 0.50 0.27 3.8 5.8

EPK 0.1 0.21 4.89 0.41 7.83 1.99 14.70 2.86 12.8 23.2EPL 0.5 0.45 10.30 0.53 8.33 2.91 15.70 4.17 10.6 11.6

EPM 0.2 1.33 21.80 0.36 0.43 3.05 0.29 3.83 12.2 236.0EPN 1.0 0.20 3.21 0.07 0.72 0.69 1.24 1.51 6.9 12.1EPO 0.1 0.13 4.11 0.07 0.08 1.93 0.15 5.62 8.6 21.8EPP 0.6 3.96 19.40 0.69 0.92 1.95 1.50 3.72 69.4 528.0EPQ 1.6 4.80 30.00 0.12 0.52 5.60 0.36 0.30 20.3 486.0EPR 0.5 0.35 24.30 0.20 1.74 3.95 3.71 2.06 20.0 25.6EPS 0.5 0.49 28.30 0.17 0.38 4.85 1.22 5.38 13.6 12.7EPT 0.4 0.61 26.70 0.15 0.83 5.79 3.07 5.95 11.8 6.6EPU 0.2 1.12 19.90 0.30 0.60 7.44 0.74 3.54 35.2 60.1EPV 0.6 1.66 25.90 5.34 7.18 57.30 14.20 50.90 6.0 16.8EPW 0.2 0.57 18.70 0.18 0.16 4.99 0.30 7.50 52.1 90.0EPX 0.5 1.67 28.40 0.36 0.30 8.94 0.57 3.49 8.4 12.9EPY 3.1 3.62 451.00 1.80 17.30 27.90 29.40 1160.00 2.2 4.2EPZ 0.7 1.36 27.30 0.65 0.76 9.65 1.72 0.77 2.3 27.9

EPAA 0.7 0.18 30.00 8.48 25.60 41.70 32.40 166.00 14.0 45.8EPAB 0.6 0.36 6.36 0.16 0.17 0.94 0.17 1.29 4.0 12.9EPAC 0.4 1.26 33.90 0.48 1.54 12.00 2.06 26.00 26.5 47.9EPAD 1.1 3.32 38.70 0.68 1.17 27.80 1.03 68.10 73.1 201.0EPAE 2.5 5.69 248.00 9.97 4.26 76.00 13.20 827.00 18.4 62.2mean 0.8 1.47 43.94 1.61 3.51 14.00 5.48 85.70 22.3 83.0

median 0.6 0.88 25.90 0.41 0.83 5.76 1.24 5.38 11.8 25.9max 3.1 5.69 451.00 9.97 25.60 76.00 32.40 1160.00 177.0 528.0min 0.1 0.13 3.21 0.07 0.08 0.69 0.15 0.27 1.6 4.2

EP19 0.1 0.60 13.30 0.25 0.30 1.38 0.17 6.32 17.5 37.5

Table 1. Continued.

ID TN:TPU Na:K Field observations

EPA 15.77 1.75 75% ice coveredEPB 62.01 1.51 75% ice coveredEPC 47.30 1.56 80-90% ice covered, Nostoc ballsEPD 16.76 6.91 80% ice covered, EPE 97.18 1.15 organic crust, Nostoc sheetsEPF 15.40 2.16 glaciers at both ends, mostly ice covered, supsended silt,EPG 48.28 0.77 80-90% ice covered, EPH 47.66 1.00 no iceEPI 34.78 3.82 60% ice covered, tufts of green algae, Nostoc ballsEPJ 45.61 3.29 ~100% ice covered,

EPK 35.31 19.10 99% ice covered, Nostoc near shore, EPL 60.73 15.72 flows into EPK, ~100% ice covered,

EPM 41.02 1.19 light brownish algal mats floatingEPN 9.80 10.29 some bright green algal filaments, some NostocEPO 74.44 1.14 on top of a glacierEPP 69.54 1.33 very grassy catchmentEPQ 120.32 4.33 lots of evidence of animal activityEPR 37.08 8.70 75% ice coveredEPS 56.19 2.24 in valley between 2 mountains, green filamentous algae, inflow, desolateEPT 457.00 5.53 probably had a stream flowing through, but not when we sampled, EPU 46.67 2.00 90% ice coveredEPV 104.40 1.34 some foam on water, ice freeEPW 82.00 0.89 very desolate, 95% ice coveredEPX 51.54 0.83 some floating pancakes of ice, but mostly ice-free, copepods with eggsEPY 8.82 9.61 ~200 m from sea, small pond with white precipitate (gypsum?)EPZ 61.47 1.17 foam, mosses, some green algal filaments

EPAA 38.72 3.02 lots of zoops, flocculant sedsEPAB 28.82 1.06 50% ice covered, EPAC 48.13 3.21 30-40% ice covered, EPAD 84.42 1.72 seems to have had streams flowing in, but not when we sampledEPAE 121.60 0.43 loads of fairy shrimpmean 66.73 3.83

median 48.13 1.75max 457.00 19.10min 8.82 0.43

EP19 23.50 1.20

Tabl

e 2.

Sum

mar

y of

sele

cted

lim

nolo

gica

l var

iabl

es fo

r the

oas

is si

tes,

with

abb

revi

atio

ns a

s des

crib

e in

Tab

le 1

. P

valu

es re

pres

ent

the

resu

lts o

f t-te

sts (

assu

min

g un

equa

l var

ianc

e) b

etw

een

the

oasi

s and

nor

ther

n si

tes.

“N

S” in

dica

tes n

o si

gnifi

cant

diff

eren

ce.

Full

deta

ils fo

r oth

er li

mno

logi

cal p

aram

eter

s (e.

g. m

etal

s, ra

re e

arth

ele

men

ts) a

re a

vaila

ble

in A

ppen

dix

7.

ID

Nam

eL

atitu

deL

ongi

tude

Dat

eel

evSA

Pond

/pH

cond

Chl

aD

ICD

OC

POC

PON

NH

3

NW

mha

Lak

eμS

/cm

μg/L

mg/

Lm

g/L

mg/

Lm

g/L

mg/

LEP

1Sk

elet

on81˚4

9.79

8’71˚2

8.48

3’8

296

1.84

L8.

2017

50.

620

.75.

80.

529

0.08

40.

014

EP2

81˚4

9.84

5’71˚2

8.35

2’8

296

1.00

L8.

1718

00.

622

.84.

61.

240

0.25

90.

032

EP3

81˚4

9.88

4’71˚2

8.05

2’8

296

0.24

L8.

0718

70.

925

.94.

11.

490

0.33

40.

010

EP4

81˚5

0.72

1’71˚2

3.92

8’9

300

0.06

P8.

9323

51.

122

31.1

1.37

00.

131

0.14

5EP

581˚5

0.75

2’71˚2

3.84

9’9

297

0.16

P8.

7044

30.

132

.429

.62.

250

0.48

60.

063

EP6

81˚5

0.77

3’71˚2

4.81

5’9

250

0.06

P7.

9090

0.5

7.3

5.6

0.90

00.

168

0.00

9EP

781˚5

0.33

7’71˚2

0.06

0’9

230

1.40

P8.

4070

01.

531

.412

.50.

574

0.08

60.

054

EP8

81˚5

0.47

4’71˚1

9.18

7’9

220

0.02

P8.

4365

00.

742

.740

.11.

000

0.17

40.

034

EP9

81˚5

0.09

6’71˚1

8.53

9’9

210

0.34

P8.

4016

500.

0524

.535

.90.

485

0.04

20.

063

EP10

81˚4

8.96

5’71˚2

4.99

5’10

170

0.05

P8.

5036

20.

620

.38.

10.

449

0.05

60.

003

EP11

81˚4

8.65

0’71˚2

6.92

2’10

170

0.76

P8.

9017

21.

113

.79.

10.

511

0.05

70.

057

EP12

81˚4

8.59

0’71˚3

4.73

2’10

190

0.13

P8.

1713

000.

555

360.

693

0.07

50.

011

EP13

81˚4

8.71

0’71˚3

3.61

4’10

200

0.02

P8.

3735

50.

0542

.427

.21.

550

0.31

30.

010

EP14

81˚4

9.23

6’71˚3

2.27

9’10

210

0.01

P8.

8739

00.

0531

.331

.80.

847

0.08

20.

049

EP15

81˚4

9.89

9’71˚3

1.62

9’15

300

0.14

P8.

5356

00.

112

.29.

20.

769

0.14

50.

044

EP16

81˚4

9.73

7’71˚3

2.25

1’15

300

0.50

P8.

7310

001

2518

.20.

854

0.08

80.

027

EP17

81˚4

9.78

1’71˚3

0.85

5’15

300

0.05

P8.

5352

00.

0524

.824

.10.

538

0.04

90.

044

EP18

81˚4

9.30

6’71˚2

1.04

5’15

160

0.05

P8.

3347

00.

527

.413

.40.

482

0.05

50.

047

EP20

81˚4

9.53

3’71˚1

9.99

9’15

170

0.09

P8.

0388

01

24.5

21.6

0.62

60.

061

0.00

7EP

2181˚4

9.53

3’71˚1

9.99

9’15

170

0.07

P8.

3369

01.

126

.323

.30.

387

0.03

50.

023

EP22

81˚4

9.26

’71˚4

5.04

’16

853

1.60

P8.

4711

50.

69.

93.

90.

317

0.05

20.

003

EP23

81˚4

9.29

8’71

.45.

162’

1686

00.

11P

7.97

910.

058.

92.

30.

185

0.02

90.

018

EP24

81˚4

9.04

1’71˚4

6.88

6’16

870

2.80

L7.

5345

0.5

4.8

0.7

0.42

50.

077

0.00

9m

ean

318

0.50

8.23

490

0.6

24.2

17.3

0.80

30.

128

0.03

4m

edia

n25

00.

138.

4039

00.

624

.513

.40.

626

0.08

20.

027

max

870

2.80

8.93

1650

1.5

5540

.12.

250

0.48

60.

145

min

160

0.01

7.53

450.

054.

80.

70.

185

0.02

90.

003

P va

lue

NS

0.02

2N

SN

S0

00.

001

NS

Tabl

e 2.

Con

tinue

d.

ID

TK

NT

dNT

NT

PuT

PfSR

PSi

O2

Ca

KN

aM

gC

lSO

4A

lFe

mg/

Lm

g/L

mg/

Lμg

/Lμg

/Lμg

/Lm

g/L

mg/

Lm

g/L

mg/

Lm

g/L

mg/

Lm

g/L

μg/L

μg/L

EP1

0.31

40.

354

0.40

39.

55.

21.

34.

6138

.80

1.09

1.42

6.53

0.61

41.7

022

.729

.4EP

20.

296

0.34

60.

561

7.5

3.6

1.0

5.21

41.7

00.

851.

265.

590.

5830

.30

1.3

72.3

EP3

0.29

80.

275

0.63

455.

63.

20.

85.

7044

.30

0.98

1.41

6.04

0.56

33.5

020

.545

.4EP

42.

270

1.94

02.

4035

15.7

9.3

2.3

4.99

35.0

05.

702.

1317

.20

2.27

65.4

08.

267

.7EP

52.

100

1.82

02.

592

11.7

7.1

2.1

5.91

51.2

08.

073.

2932

.80

3.45

132.

004.

414

9.0

EP6

0.28

90.

316

0.45

959.

25.

51.

03.

1615

.00

0.08

0.73

1.83

0.12

10.1

04.

014

3.0

EP7

0.77

80.

746

0.86

911

.85.

20.

90.

9829

.60

13.2

068

.40

40.7

021

.60

260.

009.

343

.6EP

82.

040

1.91

02.

2165

24.1

8.5

2.5

7.21

83.2

010

.30

5.71

41.1

05.

3121

1.00

2.0

206.

0EP

92.

120

1.99

02.

167

16.0

9.6

4.2

12.3

034

4.00

21.8

029

.60

97.7

016

.70

1150

.00

13.8

359.

0EP

100.

318

0.32

80.

3765

9.5

5.6

1.0

2.11

60.3

03.

391.

9514

.90

1.04

131.

0042

.186

.5EP

110.

584

0.56

70.

6435

13.5

6.4

0.7

2.49

29.7

02.

171.

094.

120.

8432

.90

19.6

231.

0EP

121.

610

1.47

01.

6875

13.0

10.5

4.5

12.2

023

4.00

32.0

028

.70

87.1

015

.20

751.

002.

243

0.0

EP13

1.15

01.

380

1.46

559.

75.

83.

76.

5159

.80

5.37

3.81

17.2

03.

5451

.60

9.6

178.

0EP

141.

840

1.61

01.

9245

20.6

7.1

2.2

5.33

67.2

04.

834.

4722

.40

2.92

130.

004.

218

3.0

EP15

0.59

40.

645

0.74

1510

.44.

60.

31.

1441

.00

2.86

3.22

18.9

01.

1912

7.00

6.9

68.2

EP16

1.38

01.

210

1.51

412

.06.

91.

52.

6610

7.00

19.9

013

.70

117.

005.

3468

0.00

4.7

92.4

EP17

1.50

01.

430

1.55

87.

66.

11.

913

.90

90.5

03.

945.

7434

.20

1.56

262.

0012

.773

.5EP

180.

961

0.92

71.

036

17.1

7.0

0.8

2.30

64.9

06.

321.

9234

.50

1.42

189.

005.

911

7.0

EP20

1.32

00.

030

1.38

359.

46.

10.

88.

1919

1.00

8.32

4.79

38.1

04.

7649

4.00

8.5

610.

0EP

210.

992

0.90

71.

0295

7.8

6.4

0.9

11.6

014

8.00

7.49

2.77

25.4

01.

8932

9.00

3.0

274.

0EP

220.

156

0.17

90.

2105

8.5

2.4

0.3

1.62

22.4

00.

641.

211.

940.

6418

.90

23.7

41.5

EP23

0.16

70.

207

0.22

74.

92.

40.

11.

4218

.30

0.55

1.06

1.76

0.37

16.6

031

.452

.1EP

240.

087

0.10

90.

182

4.8

2.0

0.1

1.56

8.87

0.25

0.52

0.83

0.36

3.25

37.1

53.4

mea

n1.

007

0.90

01.

1411

.35.

91.

55.

3579

.38

6.96

8.21

29.0

44.

0122

3.92

12.9

156.

8m

edia

n0.

961

0.74

61.

039.

76.

11.

04.

9951

.20

4.83

2.77

18.9

01.

5613

0.00

8.5

92.4

max

2.27

01.

990

2.59

24.1

10.5

4.5

13.9

034

4.00

32.0

068

.40

117.

0021

.60

1150

.00

42.1

610.

0m

in0.

087

0.03

00.

184.

82.

00.

10.

988.

870.

080.

520.

830.

123.

251.

329

.4P

valu

e0

00

0.00

70

0.01

80

NS

0.00

5N

SN

SN

SN

SN

SN

S

Tabl

e 2.

Con

tinue

d.

ID

TN

:TPU

Na:

KFi

eld

Obs

erva

tions

EP1

42.4

21.

30m

oss b

ank,

flow

s int

o EP

2EP

274

.80

1.48

gras

sy m

oss b

ank,

flow

s int

o EP

3EP

311

3.30

1.44

som

e N

osto

cEP

415

3.09

0.37

lots

of N

osto

c an

d fa

iry sh

rimp

EP5

221.

540.

41so

me

Nos

toc,

lots

of f

airy

shrim

p an

d D

aphn

iaEP

649

.95

9.13

wet

land

in a

col

, EP

773

.64

5.18

full

of c

otto

n gr

ass

EP8

91.9

70.

55gr

assy

are

a, m

oss t

hrou

ghou

t, w

ater

slig

htly

col

oure

dEP

913

5.44

1.36

flood

ed g

rass

y m

eado

w, b

ig fa

iry sh

rimp,

scat

in w

ater

EP10

39.6

30.

58co

nnec

ted

to a

noth

er b

asin

, old

shee

ts o

f Nos

toc,

lots

of D

aphn

ia a

nd fa

iry sh

rimp

EP11

47.6

70.

50re

d th

roat

ed lo

on n

estin

g he

reEP

1212

9.81

0.90

som

e N

osto

c,

EP13

151.

080.

71N

osto

c flo

atin

g, b

ig fa

iry sh

rimp,

pro

babl

y co

uld

take

cor

e he

reEP

1493

.42

0.93

near

Blis

ter C

reek

, ver

y lu

sh, E

quis

etum

nea

rby,

cou

ld b

e ph

ytol

iths

EP15

71.3

01.

13lu

sh, b

eige

fila

men

tous

mat

eria

l on

aqua

tic m

osse

s, ca

ribou

and

mus

k ox

act

ivity

, fai

ry sh

rimp

and

Dap

hnia

EP16

126.

170.

69lo

ts o

f Dap

hnia

and

fairy

shrim

p, w

olf t

rack

s and

ani

mal

act

ivity

nea

rby

EP17

205.

001.

46lo

ts o

f tad

pole

shrim

p he

re, f

airy

shrim

p al

so h

ad e

ggs

EP18

60.5

80.

30lo

ts o

f tad

pole

shrim

p, fa

iry sh

rimp,

Dap

hnia

EP20

147.

180.

58ne

xt to

EP2

1 (b

ut n

ot a

ttach

ed),

big

fairy

shrim

pEP

2113

1.99

0.37

pond

bot

tom

mos

tly m

oss,

som

e se

dEP

2224

.76

1.89

high

ele

v, a

rctic

har

e an

d pt

arm

igan

, EP

2346

.33

1.93

lots

of l

ight

gre

en fi

lam

ento

us a

lgae

, muc

h sh

allo

wer

than

EP2

2,

EP24

37.9

22.

0880

% ic

e co

vere

d, ro

cky

botto

m,

mea

n98

.65

1.53

med

ian

91.9

70.

93m

ax22

1.54

9.13

min

24.7

60.

30P

valu

e0.

041

0.00

7

Table 3. Pearson correlation matrix with Bonferroni-adjusted probabilities. Significantly correlated variables are shown in bold (p<0.01) or italics (p<0.05).

pH COND SiO2 Ca POC PON DOC TdN TPu TPf Mg NaCOND 0.467 1SiO2 0.347 0.67 1Ca 0.366 0.92 0.772 1POC 0.454 0.519 0.729 0.525 1PON 0.292 0.391 0.631 0.371 0.92 1DOC 0.516 0.734 0.749 0.659 0.698 0.593 1TdN 0.593 0.599 0.606 0.512 0.597 0.554 0.8 1TPu 0.416 0.613 0.557 0.569 0.566 0.523 0.742 0.595 1TPf 0.494 0.665 0.688 0.612 0.649 0.55 0.879 0.775 0.751 1Mg 0.557 0.943 0.586 0.811 0.48 0.354 0.694 0.616 0.495 0.61 1Na 0.349 0.75 0.39 0.586 0.336 0.292 0.616 0.5 0.595 0.55 0.723 1TKN 0.563 0.74 0.734 0.66 0.708 0.616 0.943 0.842 0.753 0.9 0.718 0.624Al -0.06 -0.243 -0.089 -0.184 -0.18 -0.201 -0.324 -0.29 -0.315 -0.284 -0.23 -0.259Fe 0.228 0.372 0.626 0.414 0.518 0.44 0.593 0.395 0.398 0.514 0.316 0.164Chla 0.044 0.036 0.116 -0.008 0.147 0.111 0.237 0.096 0.203 0.262 -0.036 0.047DIC 0.511 0.658 0.525 0.541 0.448 0.357 0.576 0.545 0.337 0.44 0.772 0.547Cl 0.175 0.497 -0.001 0.327 0.035 0.002 0.187 0.18 0.331 0.259 0.469 0.76SO4 0.077 0.764 0.475 0.784 0.251 0.125 0.454 0.331 0.477 0.486 0.626 0.591K 0.232 0.728 0.469 0.608 0.351 0.262 0.667 0.567 0.445 0.603 0.719 0.695SRP 0.241 0.607 0.594 0.646 0.527 0.426 0.65 0.59 0.558 0.601 0.535 0.481

Table 3. Continued.

TKN Al Fe Chla DIC Cl SO4 KTKN 1Al -0.354 1Fe 0.533 0.268 1Chla 0.242 -0.023 0.19 1DIC 0.581 -0.295 0.233 -0.137 1Cl 0.253 -0.183 -0.139 0.075 0.272 1SO4 0.468 -0.209 0.208 0.103 0.223 0.568 1K 0.638 -0.275 0.405 0.147 0.555 0.477 0.702 1SRP 0.599 -0.251 0.367 -0.017 0.417 0.342 0.685 0.663

CHAPTER 5

EARLY-20TH CENTURY ENVIRONMENTAL CHANGES INFERRED USING SUB-FOSSIL DIATOMS FROM A SMALL POND ON MELVILLE ISLAND, N.W.T., CANADIAN HIGH ARCTIC

BRONWYN E. KEATLEY, MARIANNE S. V. DOUGLAS, AND JOHN P. SMOL

This chapter has been published separately: Keatley, B.E., M.S.V. Douglas, and J.P. Smol. 2006. Early-20th century environmental change inferred using sub-fossil diatoms from a small pond on Melville Island, N.W.T., Canadian High Arctic. Hydrobiologia. 533:15-26.

ABSTRACT Diatom-based paleolimnological studies are being increasingly used to track long-term

environmental change in arctic regions. Little is known, however, about the direction and nature

of such environmental changes in the western Canadian high Arctic. In this study, shifts in diatom

assemblages preserved in a 210Pb-dated sediment core collected from a small pond on Melville

Island, N.W.T., were interpreted to record marked environmental changes that had taken place

since the early 20th century. For most of the history of the pond recorded in this core, the diatom

assemblage remained relatively stable and was dominated by Fragilaria capucina. A major shift

in species composition began in the early-20th century, with a sharp decline in F. capucina and a

concurrent increase in Achnanthes minutissima. In the last ~20 years, further changes in the

diatom assemblage occurred, with a notable increase in the Nitzschia perminuta complex. The

assemblage shifts recorded at this site appear to be consistent with environmental changes

triggered by recent climatic warming.

INTRODUCTION

The Arctic is well-recognized to be especially sensitive to environmental change (Serreze

et al. 2000; Houghton et al. 2001). Due to a number of feedback mechanisms, such as snow

cover-albedo, proposed temperature increases are likely to be maximized in high-latitude regions.

Thus the Arctic comprises a critical reference area for environmental change. In the Canadian

high Arctic, logistical constraints and the short duration and poor spatial coverage of the few

meteorological records makes monitoring of this vast area particularly difficult. The lack of long-

term instrumental data in many ways precludes an accurate assessment of long-term

environmental change. Paleolimnological techniques, however, may provide an effective

alternative method of gathering records of past environmental conditions when traditional

monitoring data are not available (Smol 2002).

Unlike many other arctic proxy records, lakes and ponds are abundant throughout the

Canadian high Arctic, and thus offer the potential of greater regional synthesis. Diatoms (class

Bacillariophyceae), siliceous unicellular algae, are particularly useful paleoenvironmental

indicators because they are ubiquitous, they respond rapidly to changing conditions, and different

species often have distinct optima to given environmental variables (Stoermer & Smol 1999).

Observational data (Serreze et al. 2000) and proxy records (e.g., Kaufman et al. 2004)

indicate that the timing and nature of environmental changes are not synchronous across the

Arctic. For example, divergences in stable oxygen isotope, atmospheric dust, and glaciochemical

records between ice core records from the Devon Ice Cap (Devon Island), Penny Ice Cap (Baffin

Island), and Greenland (Camp Century and GISP2) suggest increasingly regional climatic

influences during the Holocene (Paterson et al. 1977; Fisher 1979; O’Brien et al. 1995;

Zdanowicz et al. 2000; Grumet et al. 2001).

To date, most paleolimnological records from the Canadian high Arctic are from eastern

regions, and have illustrated differences in the timing and magnitude of environmental change

(e.g. Douglas et al. 1994; Doubleday et al. 1995; Perren et al. 2003; Michelutti et al. 2003a;

Antoniades et al. 2005). For example, diatom assemblages from shallow ponds on Cape

Herschel, east-central Ellesmere Island (Figure 1, site G), remained relatively static and were

interpreted to record cool temperatures for several millennia up until the mid-19th century, when

the diatom assemblages underwent substantial changes that were attributed to climatic warming

(Douglas et al. 1994). Since this initial study, other high arctic paleolimnological investigations

have shown similar changes in diatom community structure since ~1850 (e.g., Doubleday et al.

1995; Gajewski et al. 1997; Wolfe 2000; Perren et al. 2003; Michelutti et al. 2003a; Antoniades et

al. 2005). These shifts have not been simultaneous, but rather appear to be at least partly related

to the local limnological conditions, such as lake size (e.g., Doubleday et al. 1995; Michelutti et

al. 2003a) and other variables. For example, diatom assemblages from small water bodies near

Isachsen, Ellef Ringnes Island (Figure 1, site B) experienced species turnover starting in the mid-

19th century, while those from a larger lake at Alert, Ellesmere Island (Figure 1, site C), only

began to shift after the mid-to late 20th century (Antoniades et al. 2005). Likewise, subtle diatom

assemblage changes at a large and deep high arctic lake (Char Lake, Cornwallis Island, Figure 1,

site K) only began to occur in the late-1980s (Michelutti et al. 2003a). These diatom community

shifts were correlated to warmer temperatures documented by nearby instrumental meteorological

records. Warming conditions are expected to result in a lengthening of the growing season and

enhanced autochthonous production due to a reduction in duration of ice cover; in turn, this is

expected to affect limnological variables such as pH, nutrients, and specific conductivity

(Douglas & Smol 1999). These studies suggest that larger water bodies of the high Arctic had a

greater thermal inertia which acted as a buffer against the early-onset of changes in diatom

communities (e.g., Doubleday et al. 1995; Michelutti et al. 2003a).

Paleolimnological research using abiotic proxies, such as varves, have also indicated

relatively recent and marked environmental changes in Canadian high Arctic lakes and have

similarly been correlated to climatic warming (Smith et al. 2004). Likewise, analyses of sediment

cores from five lakes on Svalbard suggested that climatic warming was a contributing factor to

changing diatom assemblages over the last ~150 years (Birks et al. 2004; Jones & Birks 2004).

Similarly, research programs in subarctic regions have recorded recent environmental changes

consistent with warming (e.g., Sorvari et al. 2002; Rühland et al. 2003a).

Despite a growing body of arctic paleoenvironmental literature, no detailed

paleolimnological investigations have yet been published from the vast western Canadian high

Arctic. Instrumental meteorological data are from this region are sparse and of short duration; the

nearest weather station is located at Mould Bay, Prince Patrick Island (Figure 1, site L), which

began collecting data in 1948. This lack of instrumental meteorological data hampers our ability

to assess past climatic and associated environmental changes from the western Canadian high

Arctic. Given the sensitivity of small ponds in the eastern high Arctic (Douglas et al. 1994), a

small pond on Melville Island was chosen to be the focus of a high-resolution paleoenvironmental

investigation. The goals of this study are to assess diatom-based paleolimnological changes from

a small pond on central Melville Island, to examine whether diatom assemblages changed in

composition and, if so, when and why these changes occurred.

STUDY SITE

With a surface area of 42 149 km2, Melville Island (Figure 1) is the 4th largest of the

Queen Elizabeth Islands, and the 7th largest of all Canadian Arctic Islands, yet it is uninhabited.

Melville Island is the only island in the western high Arctic that retains small permanent ice caps.

In the absence of local meteorological records, data from Mould Bay (76˚13’N, 119˚19’W),

Prince Patrick Island (Figure 1), located approximately 230 km to the northwest, was used to

estimate average February and July temperature of -34.0˚C and 4.0˚C, respectively, and a mean

annual precipitation of 111.0 mm (Meteorological Service of Canada 2002).

Pond MV-AT (unofficial name, 75˚19’N, 111˚25’W) is a small, shallow (maximum

depth = 0.40 m), alkaline (pH = 8.1), and very dilute (specific conductivity = 39 μS/cm) pond

located on central Melville Island, Northwest Territories (Figure 1, site A; Table 1). The pond

has moderately low nutrients (e.g., total phosphorus = 12.7 μg/L) and relatively high dissolved

organic carbon (8.2 mg/L) (Table 1). Surficial geology is composed of a veneer of weathered

bedrock (sandstone, siltstone, and shale) of the Griper Bay Formation (Hodgson & Vincent

1984). Pond MV-AT is located on an interfluve approximately 15 m above two braided rivers,

with a catchment dominated by grasses, sedges, and mosses.

MATERIALS AND METHODS

A modified Glew (1989) gravity corer (diameter = 7.82 cm) was used to collect a 21 cm

long core from the centre of the pond on July 24, 2002. The sediment core was sectioned into 0.5

cm intervals from 0-13 cm using a Glew (1988) extruder immediately after retrieval. This section

of the core was rich in organic matter and contained mosses. Below 13 cm, the sediment was

composed predominantly of fine minerogenic sediment, and thus the remainder of the core was

sectioned into 1 cm intervals.

Temperature, pH, and specific conductivity were measured in the field, and water

samples were sent to the Canadian Centre for Inland Waters (Burlington, Ontario, Canada) for

analysis of nutrients, trace metals, and major ions (Environment Canada 1994). Methodological

details of sampling are provided in Appendix 1.

210Pb dating was performed at the Paleoecological Environmental Assessment and

Research Lab (PEARL), Queen’s University, using gamma spectrometry (Appleby 2001).

Activity levels were converted to dates using the Constant Rate of Supply (CRS) method (Binford

1990).

Preparation of diatom samples followed standard techniques (Battarbee et al. 2001). A

minimum of 300 diatom valves were identified and enumerated at each interval. Diatom

identification followed Krammer & Lange-Bertalot (1991), Krammer (2000), Krammer (2002),

Lange-Bertalot (2001), and Antoniades (2004).

Diatom results were converted to percent relative abundance measures and plotted

against depth using the program C2 version 1.4 beta (Juggins 2003). To identify zones which

may be considered to have some statistical validity, optimal splitting combined with broken-stick

analysis was performed (Bennett 1996), using the program psimpoll 4.10 (Bennett 2002). A

detrended correspondence analysis (DCA) was applied down-core to estimate species turnover

(Birks 1998). DCA was performed using CANOCO version 4.0 (ter Braak & Šmilauer 1998).

Loss-on-ignition (LOI) analysis followed Dean (1974). Sediments were freeze-dried and

pre-weighed before combustion at 550˚C for 2 hours to provide a proxy for organic matter

content of the sediment (Heiri et al. 2001). Further combustion at 1000˚C for two hours provided

an estimate of carbonate content (Dean 1974).

RESULTS

Core Chronology

While all 35 sediment intervals were dated, the unsupported 210Pb profile was only

detected within the upper 7 cm of the sediment core. Gamma emissions were low, as is common

in high arctic lake sediments (Pienitz et al. 2004), but nonetheless the profile exhibited a roughly

exponential decay, indicating that a reliable geochronological profile could be established for the

pond’s recent history (ca. 200 years). Three intervals had higher 210Pb counts relative to the

intervals immediately above them, and thus were not used in the date calculations; such reversals

are not uncommon when emissions are very low.

Based on the CRS model, the calculated sedimentation rate for the upper ~2.5 cm was

high (0.0322 cm/yr), relative to other high arctic lakes (e.g., Douglas et al. 1994). Below this

level, the rate dropped off exponentially and between 5 - 6.5 cm, the sedimentation rate reached a

plateau at 0.0106 cm/yr (Figure 2).

Diatoms and LOI

A total of 81 diatom taxa were identified, but only 7 were considered common (i.e.,

present at >5% relative abundance in at least one interval). Broken-stick analysis based on

optimal splitting identified one split of significant note occurring between 5-5.5 cm (Figure 3).

From 21 cm to 6 cm (Zone 1), the diatom assemblage remained relatively constant (Figure 3), and

was dominated by Fragilaria capucina Desmazières (~40-50%), and secondarily by Achnanthes

minutissima Kützing (~20%).

Zone 2 begins between 5-5.5 cm (Figure 3). Between 5.5-1.0 cm, F. capucina underwent

a marked decrease to about 10%, whereas A. minutissima increased to ~50% (Figure 3). Between

3-0 cm, Nitzschia perminuta (Grunow) M. Peragallo increased from ~10% to 15% relative

abundance. At the top of the core, A. minutissima decreased to about 43% relative abundance,

with a concordant increase in N. perminuta.

Species turnover was low for most of the sediment profile, as the DCA axis 1 species

scores remained relatively constant between 1 and 1.25 standard deviation (SD) units from 21 –

5.5 cm (Figure 4). The total variance of DCA axis 1 was 1.396 SD units. Above 5.5 cm, a sharp

decline in the DCA species scores to about 0.25 SD units indicates a strong species shift (Figure

4). This trend continues until 0.5 cm, at which time there is an inflection in the opposite

direction, likely reflecting the resurgence of N. perminuta at the surface of the core (Figure 4).

This shift in direction should be treated with caution, however, since the top of the core contains

the least compacted sediment.

The percentage of LOI remained constant at about 17% between 21-13 cm, increased to

~20% between 13-5.5 cm, and increased further to ~50% beginning at 5 cm (Figure 3). There

was insufficient sediment available to determine a value for LOI at the very surface of the core

(i.e., between 0-0.5 cm). The percent carbonate content remained consistently low (~2%)

throughout the core.

DISCUSSION

The diatom assemblages from MV-AT remained stable for most of the pond’s history

captured by this sediment record, with the most striking changes occurring early in the 20th

century (~5.25 cm, Figure 3). This marks the delineation of the only clear zone based on a

broken-stick analysis (Bennett 1996). As is also evident from the DCA profile (Figure 4), a

marked shift in species turnover is underway at this time. While these changes are most dramatic

at this time, it is likely that the diatom assemblage began to change even earlier (~ 5.75 cm,

Figures 3 & 4).

Potential causes contributing to the recent shift in diatom assemblages include

atmospheric pollution, anthropogenic acidification, artificial nitrogen deposition, and/or climatic

change. Melville Island is completely uninhabited, and located far from local pollution sources.

While organic pollutants, such as polychlorinated biphenyls (PCBs), may be transported long

distances (Rose et al. 2004), it is unlikely that the shifts in diatom assemblages recorded here

would be triggered by such contamination. For example, sites that had direct and prolonged PCB

contamination in Labrador showed no diatom or chrysophyte response to this input (Paterson et

al. 2003).

Diatoms are known to be sensitive indicators of acidification (e.g., Siver et al. 2003).

However, there is no suggestion of acidification of this site as the current pH is 8.1 and the

diatom shift is away from taxa known to be acidophilic (e.g., Eunotia spp.) and towards species

commonly found in more alkaline arctic waters (Antoniades et al. 2004). The increase in

Nitzschia perminuta near the top of the core provides further evidence that pH was not declining

at this site, as N. perminuta has the highest pH optimum of the three dominant species in this

core, based on a training set from Ellef Ringnes Island, Arctic Canada (Antoniades et al. 2004).

Nitrogen deposition has been implicated as a driver of diatom assemblage change in the

Rocky Mountains of Colorado (Wolfe et al. 2001; Saros et al. 2003; Wolfe et al. 2003). As

Achnanthes minutissima is commonly found in high arctic lakes and ponds with relatively high

TN values, it is possible that some degree of anthropogenic nitrogen deposition may be affecting

MV-AT as well. It is unlikely, however, that this deposition could have initiated the shift in

diatom communities on Melville Island because the commencement of the major species shift

predates the onset of artificial N production by at least ~30 years (Vitousek et al. 1997). In

addition, pond MV-AT is currently highly P-limited (TN:TP = 50, Table 1). The presence of

Nostoc balls between 0 - 5 cm in our core also suggests that N levels have not been increasing

significantly over the 20th century, as Nostoc (cyanobacteria) are N-fixers and are especially

competitive under low N conditions. Thus, increases in N would not likely result in the marked

diatom changes recorded in this pond. Instead, these diatom changes are more likely related to

climatic warming, as discussed below.

Under conditions of warming, limnological variables such as nutrients, specific

conductivity, and pH are all expected to increase in a pond such as MV-AT, as a reduction in the

duration of ice cover allows for longer growing seasons and higher primary production (Douglas

& Smol 1999). While the three most dominant taxa in the surface of the core are not known to

have highly restrictive ecological niches (e.g., Lim et al. 2001b; Michelutti et al. 2003b),

autecological information gleaned from recent diatom calibration sets in the Arctic and Subarctic

(Lim et al. 2001a; Lim et al. 2001b; Rühland et al. 2003; Michelutti et al. 2003b; Lim 2004;

Antoniades et al. 2003) suggests that the most likely cause of these assemblage shifts are

environmental changes precipitated by climatic warming. For example, arctic diatom calibration

sets suggest that A. minutissima has a higher specific conductivity optimum than F. capucina

(Antoniades et al. 2004), and is more frequently found in sites with higher total nitrogen levels

(Lim et al. 2001a; Rühland et al. 2003) and higher dissolved organic carbon (DOC)

concentrations (Rühland et al. 2003). Likewise, N. perminuta occurs in sites with relatively high

pH and soluble reactive phosphorus levels on Devon Island (Lim 2004). Small Nitzschia taxa are

also often associated with high nutrient loadings in arctic lakes (e.g., Douglas & Smol 2000;

Michelutti et al. 2002; Douglas et al. 2004). N. perminuta, however, is not commonly associated

with any specific environmental variable on Bathurst Island (Figure 1, site M; Lim et al. 2001a;

Lim et al. 2001b), the central Canadian Subarctic (Rühland et al. 2003), or Ellef Ringnes Island

(Antoniades et al. 2004). As algal production is expected to increase with a lengthening of the

growing season, a shift toward diatom species that have affinities for higher nutrient, DOC, and

specific conductivity levels suggests that the environment experienced a warming trend at MV-

AT beginning in the early 20th century.

The other common species found in Figure 3 occur in very low abundances. They are

comprised by three Cymbella and one Eunotia spp. It is likely that the decrease in Eunotia

further supports a rising pH over time in this site, as this genus is characteristic of acidic waters.

Cymbella spp. are commonly associated with shallow water arctic habitats (Douglas & Smol

1995; Michelutti et al. 2003b). Furthermore, it should be noted that all of the 7 common taxa are

periphytic (Figure 3). F. capucina is known to be commonly associated with rocks, moss, and

sediment on Victoria Island (Figure 1, site N; Michelutti et al. 2003b), while N. perminuta and A.

minutissima are both commonly found on moss, rocks, and sediment on Victoria Island

(Michelutti et al. 2003b), Cape Hershel (Douglas & Smol 1995), and Bathurst Island (Lim et al.

2001b). Thus, the changes in the diatom assemblage do not represent a shift between planktonic

and benthic taxa.

Concurrent with the marked floristic shifts in this core, there was also a sharp rise in

%LOI near the 5 cm depth (Figure 3). Increased run-off from the catchment may result in a

greater amount of allochthonous organic matter washing-in to the pond. However, evidence for

enhanced autochthonous production is apparent in the correspondence between high %LOI values

of 30-50% and the presence of many preserved Nostoc (cyanobacteria) balls in the sediment core

(up until ~5 cm depth). The occurrence of Nostoc may be a further indicator of climatic

warming, as increased temperatures would lengthen the growing season.

Comparison to other paleoenvironmental records

The diatom profile from MV-AT exhibits species shifts that are consistent with recent

environmental change inferred from diatom-based paleolimnological records from other areas in

the Canadian high Arctic (Figure 1, Douglas et al. 1994; Doubleday et al. 1995; Michelutti et al.

2003; Perren et al. 2003; Antoniades et al. 2005). While none of the previously recorded diatom

profiles show similar assemblages to those dominating the MV-AT core prior to ~1919, the more

recent MV-AT communities share similarities to other sites. For example, pond I-F (Isachsen,

Ellef Ringnes Island) recorded increases in Nitzschia perminuta since ~1850 A.D. (Antoniades et

al. 2005). In Self Pond (Alert, Ellesmere Island), the increased relative abundance of Achnanthes

minutissima tracked increases of both pH and warming temperatures since ~1920 A.D.

(Antoniades et al. 2005). The similarities between these modern assemblages and those found in

MV-AT provide further evidence that the diatom changes in MV-AT may indicate recent

warming. Furthermore, the assemblages in other studies are different, but the nature of the

diatoms shifts are similar to that of MV-AT, and have been interpreted as a response to warming

(e.g., Douglas et al. 1994; Doubleday et al. 1995; Gajewski et al. 1997; Perren et al. 2003). The

onsets of these diatom shifts, however, have occurred at different times. The large diatom

assemblage shift in MV-AT appears to have started ~1919, which would be consistent with the

timing of diatom changes at both Sawtooth Lake (Fosheim Peninsula, Ellesmere Island, Figure 1,

site E; Perren et al. 2003), and Self Pond (Antoniades et al. 2005), but later than diatom

assemblage changes at ponds near Isachsen (Antoniades et al. 2005), and at Cape Herschel,

Ellesmere Island (Douglas et al. 1994). Assuming the dating of each of the sediment profiles is

reasonably accurate, the discrepancy in timing may be due to a regional difference in warming in

the western Canadian high Arctic.

The only available paleoenvironmental record for Melville Island is the glacier mass

balance data (Koerner 2002) from the Melville South Ice Cap (Figure 1, site I). Glaciers cover

~160 km2 of Melville Island (Ommanney 2002). Mass balance measurements from the Melville

South Ice Cap have been made annually (with some exceptions) by the Geological Survey of

Canada since 1964 (Ommanney 2002). Although these records show that there is little trend in

the mass balance data, each year since 1968 (with the exception of 1984, 1986, and 1991) has had

an average negative mass balance, indicating that melt had exceeded accumulation (Koerner

2002). This is consistent with the trend inferred from our diatom and %LOI data.

The nearest available long-term meteorological data are from Mould Bay, Prince Patrick

Island (Figures 1 & 5). Although this record extends back to only 1948, and hence does not

capture the inception of the large diatom and %LOI shifts recorded here, it may still be used to

assess more recent temperature and precipitation variations. Mean June-July-August temperature

data indicate that there is little trend in temperature between 1948-1996 (Figure 5, Meteorological

Service of Canada 2004), likewise, the diatom record does not show any major shifts after ~1950.

The increase in Nitzschia perminuta near the top of the core (Figure 3) is not captured in the

meteorological data, which ends in 1996. There does appear to be an increase in annual

precipitation values at Mould Bay (Figure 5, Meteorological Service of Canada 2004), but this is

not reflected in the diatom profile of MV-AT. A closer examination of the precipitation values

for June, July, and August (the months representing and shouldering the open water season)

indicate only a very slight increase in precipitation since 1948 at Mould Bay (data not shown).

Other abiotic proxy environmental records from the Canadian high Arctic also show 20th

century warming. For example, varved sediment records from Tuborg Lake, Ellesmere Island

(Figure 1, site D), were interpreted to show warming beginning after ~1865, and especially after

~1908 (Smith et al. 2004). Likewise, ice core melt records from the Devon Ice Cap (Figure 1,

site F) suggest warm temperatures since 1869, and particularly after 1925 (Koerner 1977).

Agassiz Ice Cap, Ellesmere Island (Figure 1, site H), melt layers also suggested that the 20th

century has been the warmest of the last millennium (Koerner & Fisher 1990), and analysis of ice

fabric, dirt and firn from Meighen Island ice cap (Figure 1, site J) suggested warm conditions

between 1884-1964 (the extent of the record) (Koerner & Paterson 1974). These records,

although all located significant distances from Melville Island (Figure 1), all suggest that regional

climatic warming accelerated into the 20th century. The dramatic diatom change at MV-AT

appears to be consistent, in both time and nature, to these paleoenvironmental records.

CONCLUSIONS Diatom species assemblages have changed markedly since the early 20th century in this

small pond on Melville Island, suggesting higher pH, specific conductivity, and nutrient levels.

These data, coupled with increases in % LOI (as a proxy for organic matter), imply that algal

production increased beginning around 1919, and are consistent with environmental changes

projected under a 20th century climate warming scenario. The direction of environmental change

in pond MV-AT is broadly similar to that found in both lakes and ponds throughout the eastern

Canadian high Arctic, the Canadian Subarctic, and Svalbard. The onset of these environmental

changes, however, is somewhat later than those of small ponds from eastern Ellesmere Island

(Douglas et al. 1994), but similarly timed to changes in Self Pond (Antoniades et al. 2005) and

Sawtooth Lake (Perren et al. 2003). As expected, the changes in diatoms from MV-AT occur

much earlier and more dramatically than the shifts apparent in relatively large lakes from the

eastern Canadian high Arctic (e.g., Michelutti et al. 2003). Future paleolimnological studies in

this region will help to refine and corroborate these interpretations.

ACKNOWLEDGEMENTS

This research was funded by Natural Sciences and Engineering Research Council (NSERC)

grants to BEK, MSVD, and JPS, and a Northern Scientific Training Program grant and a Queen’s

Graduate Award to BEK. Logistical support was provided by the Polar Continental Shelf Project

(PCSP). Assistance in the field was provided by J.R. Glew, N. Michelutti, and especially by D.

Antoniades. Input from three reviewers greatly improved this manuscript. Many thanks to K.

Rühland, D. Selbie, A. Harris, A. Strecker, and R. Bull for helpful comments. This is PCSP

contribution # 05004.

REFERENCES

Antoniades, D., 2004. The limnology, diatom autecology, and paleolimnology of lakes and ponds from Alert, Ellesmere Island, Isachsen, Ellef Ringnes Island, and Mould Bay, Prince Patrick Island, Canadian High Arctic. University of Toronto. PhD Thesis.

Antoniades, D., M. S. V. Douglas, & J. P. Smol, 2004. Diatom species-environment relationships and inference models from Isachsen, Ellef Ringnes Island, Canadian High Arctic. Hydrobiologia 529: 1-18.

Antoniades, D., M. S. V. Douglas, & J. P. Smol, 2005. Quantitative estimates of recent environmental changes in the Canadian High Arctic inferred from diatoms in lake and pond sediments. Journal of Paleolimnology 33.

Appleby, P. G., 2001. Chronostratigraphic techniques in recent sediments. In W. M. Last & J. P. Smol (eds), Tracking Environmental Change Using Lake Sediments Volume 1: Basin Analysis, Coring, and Chronological Techniques. Kluwer Academic Publishers, Dordrecht: 171-203.

Battarbee, R. W., V. J. Jones, R. J. Flower, N. G. Cameron, H. Bennion, L. Carvalho & S. Juggins, 2001. Diatoms. In J. P. Smol, H. J. B. Birks & W. M. Last (eds), Tracking Environmental Change Using Lake Sediments: Terrestrial, Algal, and Siliceous Indicators. Kluwer Academic Publishers, Dordrecht: 155-202.

Bennett, K. D., 1996. Determination of the number of zones in a biostratigraphical sequence. New Phytologist 132: 155-170.

Bennett, K. D., 2002. Documentation for psimpoll 4.10 and pscomb 1.03. Department of Earth Sciences, Uppsala Universitet, Sweden.

Binford, M. W., 1990. Calculation and uncertainty analysis of 210Pb dates for PIRLA project lake sediment cores. Journal of Paleolimnology 3: 253-267.

Birks, H. J. B., 1998. Numerical tools in palaeolimnology - Progress, potentialities, and problems. Journal of Paleolimnology 20: 307-332.

Birks, H. J. B., V. J. Jones, & N. L. Rose, 2004. Recent environmental change and atmospheric contamination on Svalbard as recorded in lake sediments - synthesis and general conclusions. Journal of Paleolimnology 31: 531-546.

Dean, Jr. W. E., 1974. Determination of carbonate and organic matter in calcareous sediments and sedimentary rocks by loss on ignition: comparison with other methods. Journal of Sediment Petrology 44: 242-248.

Doubleday, N. C., M. S. V. Douglas, & J. P. Smol, 1995. Paleoenvironmental studies of black carbon deposition in the High Arctic - a case-study from northern Ellesmere Island. Science of the Total Environment 161: 661-668.

Douglas, M. S. V., & J. P. Smol, 1995. Periphytic diatom assemblages from high arctic ponds. Journal of Phycology 31: 60-69.

Douglas, M. S. V.& J. P. Smol, 1999. Freshwater diatoms as indicators of environmental change in the High Arctic. In E. F. Stoermer & J. P. Smol (eds), The Diatoms: Applications for the Environmental and Earth Sciences. Cambridge University Press, Cambridge: 227-244.

Douglas, M. S. V., & J. P. Smol, 2000. Eutrophication and recovery in the High Arctic: Meretta lake (Cornwallis Island, Nunavut, Canada) revisited. Hydrobiologia 431: 193-204.

Douglas, M. S. V., J. P. Smol, & W. Blake Jr, 1994. Marked post-18th century environmental change in high-arctic ecosystems. Science 266: 416-419.

Douglas, M. S. V., J. P. Smol, J. M. Savelle, & J. M. Blais, 2004. Prehistoric Inuit whalers affected arctic freshwater ecosystems. Proceedings of the National Academy of Sciences of the United States of America 101: 1613-1617.

Environment Canada, 1994. Manual of Analytical Methods – National Laboratory for Environmental Testing, Canadian Centre for Inland Waters. Burlington, ON.

Fisher, D. A., 1979. Comparison of 105 years of oxygen isotope and insoluble impurity profiles from the Devon Island and Camp Century ice cores. Quaternary Research 11: 299-305.

Gajewski, K., P. B. Hamilton, & R. McNeely, 1997. A high resolution proxy-climate record from an arctic lake with annually-laminated sediments on Devon Island, Nunavut, Canada. Journal of Paleolimnology 17: 215-225.

Glew, J. R., 1988. A portable extruding device for close interval sectioning of unconsolidated core samples. Journal of Paleolimnology 2: 241-243.

Glew, J. R., 1989. A new trigger mechanism for sediment samplers. Journal of Paleolimnology 2: 241-243.

Grumet, N. S., C. P. Wake, P. A. Mayewski, G. A. Zielinski, S. I. Whitlow, R. M. Koerner, D. A. Fisher, & J. M. Woollett, 2001. Variability of sea-ice extent in Baffin Bay over the last millenium. Climatic Change 49: 129-145.

Heiri, O., A. F. Lotter, & G. Lemcke, 2001. Loss on ignition as a method for estimating organic and carbonate content in sediments: reproducibility and comparability of results. Journal of Paleolimnology 25: 101-110.

Hodgson, D.A. & J.-S. Vincent, 1984. Surficial geology, Central Melville Island, Northwest Territories; Geological Survey of Canada, Map 1583A, scale 1:250 000.

Houghton, J. T., Y. Ding, D. J. Griggs, M. Noguer, P. J. van der Linden & D. e. Xiaosu, 2001. Climate Change 2001: The Scientific Basis Contribution of Working Group I to the Third Assessment Report of the Intergovernmental Panel on Climate Change (IPCC). Cambridge University Press, Cambridge.

Jones, V. J., & H. J. B. Birks, 2004. Lake-sediment records of recent environmental change on Svalbard: results of diatom analysis. Journal of Paleolimnology 31: 445-466.

Juggins, S., 2003. C2: software for ecological and palaeocological data analysis and visualisation, version 1.3. Department of Geography, University of Newcastle.

Kaufman, D. S., T. A. Ager, N. J. Anderson, P. M. Anderson, J. T. Andrews, P. J. Bartlein, L. B. Brubaker, L. L. Coats, L. C. Cwynar, M. L. Duvall, A. S. Dyke, M. E. Edwards, W. R. Eisner, K. Gajewski, A. Geirsdóttir, F. S. Hu, A. E. Jennings, M. R. Kaplan, M. W. Kerwin, A. V. Lozhkin, G. M. MacDonald, G. H. Miller, C. J. Mock, W. W. Oswald, B. L. Otto-Bliesner, D. F. Porinchu, K. Rühland, J. P. Smol, E. J. Steig, & B. B. Wolfe, 2004. Holocene thermal maximum in the western Arctic (0-180°W). Quaternary Science Reviews 23: 529-560.

Koerner, R. M., 1977. Devon Island Ice Cap: core stratigraphy and paleoclimate. Science 196: 15-18.

Koerner, R. M., 2002. Glaciers of high arctic islands. In R. S. Williams & J. G. Ferrigno (eds), Glaciers of North America. U.S. Geological Survey Professional Paper 1386-J-1, J111-J146.

Koerner, R. M., & D. A. Fisher, 1990. A record of Holocene summer climate from a Canadian high-Arctic ice core. Nature 343: 630-631.

Koerner, R. M., & W. S. B. Paterson, 1974. Analysis of a core through the Meighen Ice Cap, Arctic Canada, and its paleoclimatic implications. Quaternary Research 4: 253-263.

Krammer, K., 2000. Diatoms of Europe: The Genus Pinnularia. A.R.G. Gantner Verlag, Ruggell.

Krammer, K., 2002. Diatoms of Europe: Cymbella. A.R.G. Gantner Verlag, Ruggell.

Krammer, K.& H. Lange-Bertalot, 1986-1991. Bacillariophyceae. In H. Ettl, J. Gerloff, H. Heynig & D. Mollenhauer (eds). Gustav Fischer Verlag, Stuttgart.

Lange-Bertalot, H., 2001. Diatoms of Europe. Navicula sensu stricto, 10 genera separated from Navicula sensu lato, Frustulia. A.R.G. Gantner Verlag, Ruggell.

Lim, D.S.S., 2004. Limnology and diatom paleoecology of diatoms on Banks Island, N.W.T., and Devon Island, Nunavut, Canadian Arctic. University of Toronto. PhD Thesis.

Lim, D. S. S., M. S. V. Douglas, & J. P. Smol, 2001a. Diatoms and thier relationship to environmental variables from lakes and ponds on Bathurst Island, Nunavut, Canadian High Arctic. Hydrobiologia 450: 215-230.

Lim, D. S. S., C. Kwan, & M. S. V. Douglas, 2001b. Periphytic diatom assemblages from Bathurst Island, Nunavut, Canadian High Arctic: An examination of community relationships and habitat preferences. Journal of Phycology 37: 379-392.

Meteorological Service of Canada 2002. Canadian Climate Normals 1971-2000. Environment Canada, Ottawa.

Meteorological Service of Canada 2004. Climate Data Online. Environment Canada, Ottawa. http://www.climate.weatheroffice.ec.gc.ca/climateData/canada_e.html, accessed December 10, 2004.

Michelutti, N., M. S. V. Douglas, & J. P. Smol, 2002. Tracking recent recovery from eutrophication in a high arctic lake (Meretta Lake, Cornwallis Island, Nunavut, Canada) using fossil diatom assemblages. Journal of Paleolimnology 28: 377-381.

Michelutti, N., M. S. V. Douglas, & J. P. Smol, 2003a. Diatom response to recent climatic change in a high arctic lake (Char Lake, Cornwallis Island, Nunavut). Global and Planetary Change 38: 257-271.

Michelutti, N., A. J. Holtham, M. S. V. Douglas, & J. P. Smol, 2003b. Periphytic diatom assemblages from ultra-oligotrophic and UV transparent lakes and ponds on Victoria Island and comparisons with other diatom surveys in the Canadian Arctic. Journal of Phycology 39: 465-480.

O'Brien, S. R., P. A. Mayewski, L. D. Meeker, D. A. Meese, M. S. Twickler, & S. I. Whitlow, 1995. Complexity of Holocene climate as reconstructed from a Greenland ice core. Science 270: 1962-1964.

Ommanney, C. S. L., 2002. Glaciers of Canada. In R. S. Williams & J. G. Ferrigno (eds), Glaciers of North America. U.S. Geological Survey Professional Paper 1386-J-1, J27-J82.

Paterson, A. M., A. A. Betts-Piper, J. P. Smol, & B. A. Zeeb, 2003. Diatom and chrysophyte algal response to long-term PCB contamination from a point-source in northern Labrador, Canada. Water Air and Soil Pollution 145: 377-393.

Paterson, W. S. B., R. M. Koerner, D. A. Fisher, S. Johnsen, H. B. Clausen, W. Dansgaard, P. Bucher, & H. Oeschger, 1977. An oxygen-isotope climate record from the Devon Island ice cap, arctic Canada. Nature 266: 508-511.

Perren, B. B., R. S. Bradley, & P. Francus, 2003. Rapid lacustrine response to recent High Arctic warming: A diatom record from Sawtooth Lake, Ellesmere Island, Nunavut. Arctic Antarctic and Alpine Research 35: 271-278.

Pienitz, R., M. S. V. Dougas & J. P. Smol, 2004. Long-Term Environmental Change in Arctic and Antarctic Lakes. Springer, Dordrecht.

Rose, N. L., C. L. Rose, J. F. Boyle, & P. G. Appleby, 2004. Lake-sediment evidence for local and remote sources of atmospherically deposited pollutants on Svalbard. Journal of Paleolimnology 31: 499-513.

Rühland, K., A. Priesnitz, & J. P. Smol, 2003a. Paleolimnological evidence from diatoms for recent environmental changes in 50 lakes across Canadian arctic treeline. Arctic Antarctic and Alpine Research 35: 110-123.

Rühland, K. M., J. P. Smol, & R. Pienitz, 2003b. Ecology and spatial distributions of surface-sediment diatoms from 77 lakes in the subarctic Canadian treeline region. Canadian Journal of Botany 81: 57-73.

Saros, J. E., S. J. Interlandi, A. P. Wolfe, & D. R. Engstrom, 2003. Recent changes in the diatom community structure of lakes in the Beartooth Mountain Range, U.S.A. Arctic, Antarctic, and Alpine Research 35: 18-23.

Serreze, M. C., J. E. Walsh, F. S. Chapin III, T. Osterkamp, M. Dyurgerov, V. Romanovsky, W. C. Oechel, J. Morison, T. Zhang, & R. G. Barry, 2000. Observational evidence of recent change in the northern high-latitude environment. Climatic Change 46: 159-207.

Siver, P. A., R. Ricard, R. Goodwin, & A. E. Giblin, 2003. Estimating historical in-lake alkalinity generation from sulfate reduction and its relationship to lake chemistry as inferred from algal microfossils. Journal of Paleolimnology 29: 179-197.

Smith, Jr. S. V., R. S. Bradley, & M. B. Abbott, 2004. A 300 year record of environmental change from Lake Tuborg, Ellesmere Island, Nunavut, Canada. Journal of Paleolimnology 32: 137-148.

Smol, J. P., 2002. Pollution of lakes and rivers: a paleoenvironmental perspective. Arnold, London.

Sorvari, S., A. Korhola, & R. Thompson, 2002. Lake diatom response to recent Arctic warming in Finnish Lapland. Global Change Biology 8: 171-181.

Stoermer, E. F.& J. P. Smol, 1999. The diatoms: applications for the environmental and earth sciences. Cambridge University Press, Cambridge.

ter Braak, C. J. F., 1998. CANOCO and CANOPost, Version 4 for Windows. Microcomputer Power, Ithaca, NY.

Vitousek, P. M., J. D. Aber, R. W. Howarth, G. E. Likens, P. A. Matson, D. W. Schindler, W. H. Schlesinger, & D. G. Tilman, 1997. Human alteration of the global nitrogen cycle. Ecological Applications 7: 737-750.

Wolfe, A. P., 2000. A 6500-year record from southwestern Fosheim Peninsula, Ellesmere Island, Nunavut. In M. Garneau & B. T. Alt (eds), Environmental Response to Climate Change in the Canadian High Arctic. Geological Survey of Canada, Ottawa: 249-256.

Wolfe, A. P., J. S. Baron, & R. J. Cornett, 2001. Anthropogenic nitrogen deposition induces rapid ecological changes in alpine lakes of the Colorado Front Range (USA). Journal of Paleolimnology 25: 1-7.

Wolfe, A. P., A. C. Van Gorp, & J. S. Baron, 2003. Recent ecological and biogeochemical changes in alpine lakes of Rocky Mountain National Park (Colorado, USA): a response to anthropogenic nitrogen deposition. Geobiology 1: 153-168.

Zdanowicz, C. M., G. A. Zielinski, C. P. Wake, D. A. Fisher, & R. M. Koerner, 2000. A Holocene record of atmospheric dust deposition on the Penny Ice Cap, Baffin Island, Canada. Quaternary Research 53: 62-69.

Figure captions.

Figure 1. Map showing location of sites discussed in this paper. The oval on the inset map shows location of Canadian High Arctic. Sites are as follows: A) MV-AT, Melville Island; B) Isachsen, Ellef Ringnes Island; C) Alert, Ellesmere Island; D) Tuborg Lake, Ellesmere Island; E) Fosheim Peninsula, Ellesmere Island; F) Devon Ice Cap, Devon Island; G) Cape Hershel, Ellesmere Island; H) Agassiz Ice Cap, Ellesmere Island; I) Melville Island ice caps, Melville Island; J) Meighen Ice Cap, Meighen Island; K) Char Lake, Resolute Bay, Cornwallis Island; L) Mould Bay, Prince Patrick Island; M) Bathurst Island; N) Victoria Island. Figure 2. Sedimentation rate for dated sections of the MV-AT core, as calculated based on the Constant Rate of Supply (CRS) method described by Appleby (2001) and Binford (1990). Figure 3. Diatom profile showing the common diatom species found in pond MV-AT. Individual species with >5% relative abundance in at least one sample were retained for the profile; “other” is a sum of all other diatoms found in each interval. Dates are based on 210Pb dating using a Constant Rate of Supply model. Percent loss-on-ignition (%LOI 550) is expressed as a % of combustion at 550˚C, and is a proxy for organic matter content of the sediment. Percent carbonates (%LOI 1000) is expressed as a percentage of dry weight combusted at 1000˚C. Zones are based on optimal splitting and broken-stick analysis. While the most marked species change occurs at ~5.25 cm, the shift in diatom assemblage appears to have begun earlier (~5.75 cm). Figure 4. Detrended Correspondance Analysis (DCA) of diatom species scores versus depth. The DCA axis 1 species scores are scaled in Standard Deviation (S.D.) units, and provide an estimate of species turnover. Figure 5. Mean June-July-August temperature data and annual precipitation data from Mould Bay, Prince Patrick Island (see Figure 1 for location) between 1948-1996 (Meteorological Service of Canada 2004). The smoothed line is a LOWESS curve with a span of 0.35.

Figure 1.

Figure 2.

Fi

gure

3

Figure 4.

Figure 5.

Table 1. Present-day physical and chemical characteristics of pond MV-AT were collected on July 24, 2002. Abbreviations are as follows: TP (total phosphorus, unfiltered), TN (total nitrogen), DOC (dissolved organic carbon), DIC (dissolved inorganic carbon), Chl a (chlorophyll a), POC (particulate organic carbon), PON (particulate organic nitrogen), SRP (soluble reactive phosphorus), TKN (total Kjeldhal nitrogen).

Parameter Value

Latitude 75˚19’N Longitude 111˚25’W Maximum depth 0.40 m Elevation 120 m asl Specific conductivity 39 μS/cm pH 8.1 TP 12.7 μg/L TN 640 μg/L Chl a 1.60 μg/L DOC 8.2 mg/L DIC 4.8 mg/L NH3 0.011 mg/L NO2 0.001 mg/L Cl 1.93 mg/L SO4 4.8 mg/L SiO2 0.27 mg/L POC 0.559 mg/L PON 0.053 mg/L SRP 0.0057 mg/L TKN 0.648 mg/L Al 0.02 mg/L Ba 0.0023 mg/L Cu 0.001 mg/L Fe 0.353 mg/L Li 0.001 mg/L Mn 0.0052 mg/L Sr 0.0112 mg/L Ca 5.2 mg/L Mg 2.7 mg/L Na 1.5 mg/L K 0.4 mg/L

CHAPTER 6

PROLONGED ICE COVER DAMPENS DIATOM COMMUNITY RESPONSES TO RECENT CLIMATIC

CHANGE IN HIGH ARCTIC LAKES

BRONWYN E. KEATLEY, MARIANNE S.V. DOUGLAS, AND JOHN P. SMOL

This chapter has been accepted for publication and is in press: Keatley, B.E., M.S.V. Douglas, and J.P. Smol. Prolonged ice cover dampens diatom community responses to recent climatic change in high Arctic lakes. Arctic Antarctic and Alpine Research.

ABSTRACT Numerous paleolimnological studies of high Arctic lakes and ponds have shown marked shifts in

both algal and invertebrate taxa within the past ~150 years that are consistent with recent climatic

warming. However, the magnitude and timing of changes are often non-uniform, with large, deep

lakes frequently exhibiting muted assemblage shifts relative to smaller ponds. The hypothesis

that duration and extent of ice cover exerts an overriding influence on habitat availability for

biota has been commonly invoked to explain these differences, and many studies indicate that

changes in ice cover are important drivers of recent biological changes. However, a detailed

paleolimnological comparison of two lakes from the same region that have similar water

chemistry but different ice cover regimes has not yet been attempted. Here we examine the

influence of prolonged ice cover on the rate, magnitude, and direction of fossil diatom species

shifts over time in two remarkably similar and adjacent Ellesmere Island lakes that mainly differ

in their periods of ice cover (by an estimated 7-10 days). These two lakes exhibit strikingly

different paleolimnological diatom profiles, despite their physical proximity, similar depths, and

nearly identical water chemistry. In the lake characterized by prolonged ice cover, we find little

evidence of diatom-inferred environmental change over its recent history, whilst diatom

assemblages have undergone dramatic changes in the lake with the shorter duration of ice cover.

This study supports the general hypothesis that changes in ice cover are a principle determinant of

shifting diatom assemblages in high Arctic lakes.

INTRODUCTION

High latitude regions are especially sensitive to climatic changes (ACIA, 2004) and

therefore represent critical reference areas for studies of long-term environmental change. As

long-term monitoring data are lacking with respect to both temporal and spatial scales, proxy

indicators of environmental change must be relied upon to reconstruct environmental histories in

this climatically sensitive area (Pienitz et al., 2004). The abundance of lakes and ponds in the

Canadian high Arctic allows for regional assessments of environmental change using

paleolimnological techniques. The accumulation of allochthonous and autochthonous biological,

chemical, and physical indicators in lacustrine sediments provides a rich archive of information

about past environments (Smol, 2002). Diatoms are siliceous unicellular algae (class

Bacillariophyceae) that are particularly useful environmental indicators because they are

ubiquitous, respond rapidly to changing conditions, and different species often have distinct

optima with respect to many important environmental variables (Stoermer and Smol, 1999).

Diatoms have been especially effective biomonitors in Arctic regions (Douglas and Smol, 1999;

Douglas et al., 2004; Solovieva et al., 2005), where other sources of proxy data are sometimes

lacking.

In a recent analysis of 55 paleolimnological profiles from circumpolar lakes and ponds,

Smol et al. (2005) summarized the often dramatic changes in biological microfossils, including

diatoms, which have occurred over the last ~150 years. Assemblage changes were found to be

consistent with climatic warming, with the greatest changes observed in the northernmost regions.

Interpretations based on diatoms are also consistent with studies using non-biological proxies

(e.g. Smith et al., 2004). Changes in lake ice cover were first proposed by Smol (1983) as a

major cause for some of the striking species changes, and this idea was developed further by

Smol (1988) and Douglas and Smol (1999). Briefly, during cooler periods, the persistence of ice

cover can reduce primary production by limiting light penetration into the lake. Under warmer

conditions, ice cover is reduced, more light is available for photosynthesis, and as a result more

habitat becomes available for colonization by algae. However, other factors such as lake depth,

also play a role in how changes in climate, manifested through changes in lake ice, could affect

diatom flora. For example, in relatively deep, extensively ice-covered lakes near Alert and

Resolute Bay (Fig. 1), diatom communities showed subtle shifts that began over the last few

decades (Doubleday et al., 1995; Michelutti et al., 2003; Antoniades et al., 2005). In contrast,

small, shallow ponds at Cape Herschel (Ellesmere Island) and Isachsen (Ellef Ringnes Island)

(Fig. 1) tracked relatively dramatic changes in diatoms beginning much earlier, in the late 19th or

early 20th century, respectively (Douglas et al., 1994; Antoniades et al., 2005). Likewise, across

the circumpolar north, many deeper lakes have recorded increases in planktonic species during

recent history (Sorvari et al., 2002; Rühland et al., 2003), whereas shallow ponds have been

characterized by shifts in taxa indicative of greater periphytic (aquatic plant) habitat availability

(Smol et al., 2005). Both the development of plankton (in deeper lakes) and periphyton (in

shallow ponds), as well as the delayed impact of warming in some locations, could be explained

by changes in the persistence of ice cover, and associated limnological changes in the physical,

chemical, and biological characteristics of deeper versus shallower lakes (Smol, 1988; Douglas

and Smol, 1999).

While a large diatom calibration set has been developed to infer ice cover duration in

subarctic lakes in Fennoscandia (Thompson et al., 2005), the ice cover hypothesis has yet to be

explicitly tested using a paleolimnological approach. In this paper, we test the hypothesis that ice

cover is the overriding influence on diatom communities in high Arctic lakes by examining

diatom assemblage shifts in two small, adjacent lakes, Skeleton Lake and Lake EP2, in northern

Ellesmere Island, which are very similar in all limnological respects (e.g. location, morphology,

and water chemistry variables that influence diatoms) other than duration of ice cover.

Although instrumental records of climate change do not exist for the study sites and the

Lake Hazen region of northern Ellesmere Island, previous diatom-based paleolimnological

investigations suggest that environmental changes have occurred at Alert (Doubleday et al., 1995;

Antoniades et al., 2005), the Hazen Plateau (Smith, 2002), and the Fosheim Peninsula (Wolfe,

2000; Perren et al., 2003). Likewise, glacier mass balance records (Braun et al., 2004), melt

records from the nearby Agassiz Ice Cap (Koerner and Fisher, 1990), and instrumental data from

Alert and Eureka weather stations (Meteorological Service of Canada, 2006) indicate a regional

recent warming in recent decades. Thus, while recent environmental change has occurred in this

region, ice cover continues to persist longer on one lake than the other due to local cooling caused

by shading from a nearby hill. If climate modulated changes in the duration and extent of ice

cover is truly the dominant factor driving diatom assemblage shifts in Arctic lakes, we predict the

greatest ecological changes to occur in the lake with the longest open water period.

SITE DESCRIPTION

The two study lakes, Skeleton Lake and EP2 (unofficial name), are small (estimated

maximum depths ~4 m and ~3 m, respectively) high Arctic lakes (81°50’N, 71°28’W), located on

the north side of Lake Hazen, Ellesmere Island, Nunavut, Canada. Skeleton Lake drains into EP2

(less than 20 m to the east) via Skeleton Creek, which in turn flows through a third pond (EP3)

and then into Lake Hazen (Fig. 2). Skeleton Lake has repeatedly been shown to become ice free

later than EP2 (Fig. 3; Oliver and Corbet, 1966; National Air Photo Archives; and field

observations, this study, 2003). This is likely due to the greater shading and protection from

winds on Skeleton Lake by Blister Hill, located to the south (Fig. 2; Oliver and Corbet, 1966).

Skeleton Lake has been noted for its luxuriant cyanobacterial mats (Quesada et al. 1999),

although similar studies have not been conducted on EP2.

For its latitude, the Hazen Basin region experiences anomalously warm summer

conditions due to its continental location and its placement on the leeward side of the Grant Land

Mountains (Thompson, 1994). In July 2003, for example, mean maximum daily temperatures at

Eureka and Alert were 10.9˚C and 4.3˚C, respectively (Meteorological Service of Canada, 2006),

while the mean maximum daily temperature at Lake Hazen camp was 16˚C. The area receives

very little precipitation (~95 mm annually; Thompson, 1994). The frost-free period at Lake

Hazen (Fig. 1) lasts 8-10 weeks and, as such, supports a greater abundance and diversity of

vegetation than the surrounding areas (Soper and Powell, 1985). Skeleton Lake, EP2, and EP3

are surrounded by wet meadows which are predominantly covered with Carex aquatilis var. stans

and Eriophorum spp., as well as Saxifraga, Lychnis, Salix, Stellaria, Arctagrostis, Polygonum,

and Equisetum, and many types of mosses (Soper and Powell, 1985). Bedrock consists of

Jurassic and Cretaceous sandstone and shale; this is overlain by glacial till, sand, gravel, talus and

soils of Pleistocene and more recent origin (Christie, 1964).

METHODS

Water chemistry measurements and water samples were taken from Skeleton Lake, EP2,

and EP3 within approximately one hour of each other on 7 July 2003. Temperature, pH, and

specific conductivity were measured in the field with a handheld thermometer, a handheld Hanna

pHEP pH meter and aYSI model 33 conductivity meter, respectively. Water samples were also

taken for total filtered phosphorus (TPf), total unfiltered phosphorus (TPu), nitrate-nitrite-

nitrogen (NO3-NO2-N), ammonia-nitrogen (NH3-N), total Kjeldahl nitrogen (TKN, filtered), total

dissolved nitrogen (TdN), particulate nitrogen (PON), dissolved organic carbon (DOC), dissolved

inorganic carbon (DIC), particulate carbon (POC), dissolved silica (SiO2), chlorophyll a (Chl a,

uncorrected for phaeophytin), calcium (Ca), magnesium (Mg), sodium (Na), potassium (K),

sulfate (SO4), and chloride (Cl). Laboratory analyses were performed at the National Laboratory

for Environmental Testing (NLET), Environment Canada, Burlington, ON, Canada, using

methods outlined in Environment Canada (1994). Details about sampling methods are given in

Appendix 1.

Sediment cores were obtained using a modified Glew (1989) corer (3.75 cm inner

diameter) from the deepest part of both Skeleton Lake (33 cm long) and EP2 (17 cm long) on 14

July 2003. Cores were sectioned on-site at 0.5 cm intervals using a Glew (1988) extruder. The

uppermost (~2 cm) sediments from the adjacent pond, EP3, were sampled by hand to allow for a

comparison of the recently deposited diatom assemblages between the three sites.

Subsamples from each 0.5 cm thick interval of both cores were analysed for 210Pb using

alpha spectrometry at MyCore Scientific, Deep River, ON, Canada. The 210Pb dates for Skeleton

Lake were calculated using the constant rate of supply (CRS) method (Appendix 7; Binford,

1990). In the sediments from EP2, the 210Pb activity levels were too low to provide a reliable age

model, so trace elements were used to identify anthropogenic pollution horizons as an

approximate marker for the onset of the industrial revolution at ~1850 AD. Trace element

analyses were carried out by Activation Laboratories Inc., in Ancaster, Ontario, Canada. A suite

of metals (including total Pb) were measured using a Perkin Elmer SCIEZ ELAN 6100

Inductively Coupled Plasma Mass Spectrometry (ICP/MS). International certified reference

materials USGS GXR-1, GXR-2, GXR-4 and GXR-6 were analyzed at the beginning and end of

each batch of samples. Internal control standards were analyzed every 10 samples and a duplicate

was run for every 10 samples. Total Hg was measured on a Perkin Elmer FIMS 100 cold vapour

Hg analyzer. Organic matter was estimated by loss on ignition (LOI) at 550°C for 2 hours (Heiri

et al., 2001) at PEARL in Kingston, Ontario, Canada.

Microwave digestion techniques were used to clean diatom frustules from the sediment

with concentrated HNO3 (Parr et al., 2004). Diatom slurries were rinsed with deionized water

until they reached a neutral pH, dried on pre-cleaned glass cover slips, and mounted on slides

with Naphrax® (high refractive index mounting medium). A minimum of 500 diatom valves was

identified and enumerated from each sample using a Leica DMRB microscope. Taxonomy

primarily followed Krammer and Lange-Bertalot (1986-1991), Krammer (2002), and Antoniades

et al. (in press). In order to examine whether rare species were tracking environmental changes in

these lakes, we used all species in a principal components analysis (PCA). We also performed a

detrended correspondence analysis (DCA) to examine species turnover in standard deviation (SD)

units (Lepš and Šmilauer, 2003). This ecologically-relevant interpretation provides an objective

comparison of species compositional change between the two sites.

RESULTS

Modern limnology

The physical and chemical variables measured for Skeleton Lake, EP2, and EP3 are

shown in Table 1. The variables of greatest limnological interest that influence diatom

assemblages, including pH, specific conductivity, and dissolved nutrients, are nearly identical

between the three lakes; notable exceptions to this include temperature and the particulate

fractions of carbon (POC) and nitrogen (PON), and NH3. However, when examined in the

context of 55 lakes and ponds from northern Ellesmere Island, a principal components analysis

(PCA), based on water chemistry variables that are expected to influence diatom assemblages,

indicates that they are more similar to each other than to 52 other sites from northern Ellesmere

Island (Fig. 4).

Diatom assemblages identified from the recently deposited sediments of the three lakes

are listed in Table 2. All three lakes are characterized by small, benthic diatoms and share high

abundances of some common taxa (e.g. small, benthic Fragilaria spp. sensu lato). The diatom

assemblage in the surface sediments from EP3 was the most diverse of the three lakes, and

consisted of small Achnanthes spp. (A. petersenii, A. minutissima), Brachysira zellensis,

Cymbella spp., Denticula spp., and small Nitzschia species (Table 2). Similar to EP2, the most

prevalent diatom in EP3 was Fragilaria brevistriata suggesting that some, as of yet, unidentified

characteristic about these water bodies currently favours the growth of F. brevistriata.

Paleolimnology

Unsupported 210Pb concentrations in the Skeleton Lake sediment core displayed a largely

exponential decline until reaching background levels at 4.75 cm depth, corresponding to an age of

~1877 AD (Fig. 5, Appendix 11). Since the 210Pb concentrations were too low in the EP2

sediment core to obtain reliable dates, we examined total Pb and Hg concentrations as an

approximate time horizon indicating the onset of airborne anthropogenic pollution. Sedimentary

concentrations of Pb and Hg in ponds from nearby Greenland have been shown to increase

starting in the early 19th and 20th centuries (Bindler et al., 2001a; 2001b). Thus, a rise in the Hg

and Pb concentrations at 9 cm depth is interpreted as corresponding to ~150 to 200 years BP (see

Appendix 12 for total Hg and total Pb for Skeleton Lake).

A list of common diatom taxa and the DCA axis 1 gradient lengths (an estimate of

change in species composition) of the sediment cores of Skeleton Lake and EP2 can be found in

Table 3. There is little change in the 33 cm Skeleton Lake core (Fig. 6a); this is reflected in the

short gradient length of DCA axis 1 (1.06 SD units, Table 3). Diatom assemblages are composed

of small, benthic taxa, dominated by Fragiliaria construens var. venter (~80-90%). Throughout

much of this core, F. construens var. venter co-occurs with F. pinnata (up to ~20%). In the top 2

cm (corresponding to approximately the last ~30 years), there is a small increase in F.

brevistriata (up to its maximum relative abundance of ~10%), as well as the near complete

disappearance of F. pinnata.

In nearby Lake EP2, which is unshaded, warmer, and experiences less prolonged ice

cover than Skeleton Lake (Fig. 6b), three periods of diatom change occur within the 17 cm

sediment core (Fig. 6c). These patterns in the dominant species are also reflected when all

species are incorporated in the DCA, which indicates a species turnover of 2.00 SD units (Table

3). Below 10 cm, F. pinnata dominated the record (up to ~80% relative abundance), and was

accompanied by other small benthic diatoms such as F. construens var. venter and F. brevistriata

(Fig. 6c). At ~9.25 cm, estimated to be ~100-150 years BP based on our geochemical proxies,

diatom assemblages diversified to include taxa such as Denticula kuetzingii, small Nitzschia

species (N. frustulum, N. inconspicua, and N. perminuta), and Cymbella spp. In the uppermost

sediments, between 0-2 cm, a small benthic diatom, F. brevistriata, becomes the dominate taxon.

DISCUSSION

Modern limnology

The similarity of the modern limnological variables amongst the three lakes is a function

of their similar bedrock, regional climate, and connectivity via Skeleton Creek. The notable

exceptions to these similarities are temperature and the particulate fractions of carbon (POC) and

nitrogen (PON), which are all higher in EP2 and EP3 than Skeleton Lake, and NH3, which is

highest in EP2 (Table 1). The lower temperature of Skeleton Lake can be attributed to the

persistent ice cover that was present at the time of sampling; this ice cover is in turn related to the

local shading of Skeleton Lake. Both POC and PON are measures of particulate matter in the

water column, and not directly indicative of nutrients that are available to photosynthetic

organisms. Higher particulate concentrations in EP2 and EP3 suggest that they have higher

concentrations of algal and detrital matter, which are indicative of higher production and/or

inflow energy in these systems, as would be expected with decreased ice cover (Smol, 1988).

Moreover, as water flows from Skeleton Lake through EP2 and then EP3, it is reasonable to

expect higher particulate matter downstream. Although NH3 concentrations are twice as high in

EP2 as Skeleton Lake, both values are low compared to the range of NH3 values for 24 ponds and

lakes located within ~10 km radius of our study site (range NH3 = 0.003 to 0.145 mg/L, Keatley

et al., [4]). In any case, as all three lakes are P limited, not N limited (Table 1), slight differences

in these nutrients are not likely to be biologically significant.

Modern diatom assemblages

Although there are some specific taxonomic differences between the diatom assemblages

present in the recently deposited sediments of Skeleton Lake, EP2, and EP3, they also exhibit

some important similarities (Table 2). For example, small benthic Fragilaria (sensu lato) taxa are

the most abundant diatoms in each of the three sites. Small Fragiliara taxa, especially F.

pinnata, are considered to be pioneering diatoms that can commonly exploit harsh conditions and

have been well documented as an indicator of cool conditions with short growing seasons in both

the Arctic (Douglas et al., 1994; Michelutti et al., 2003a), and in alpine regions (Lotter and

Bigler, 2000), and are frequently the dominant taxa found in pre-industrial polar lake sediments

(Smol et al., 2005). Ecological preferences that can distinguish these small Fragilaria taxa are

difficult to define; for example, Karst-Riddoch (2004) suggested that the distribution of small

benthic Fragilaria taxa (F. pinnata, F. construens var. venter, F. brevistriata) in many lakes

across Iceland was not related to measured environmental variables.

The differences amongst the three lakes can best be appreciated with respect to the higher

diversity of taxa representing different growth forms and life strategies. In Lake EP3, which

experiences the least shading and is the shallowest lake, the diatom assemblage consists of many

species known to have secondary growth characteristics (e.g. Cymbella taxa often have

mucilaginous stalks). Such diverse assemblages have been linked to regions with longer growing

seasons, as these more complex life forms would take longer to develop (Douglas and Smol,

1999).

Paleolimnology

Skeleton Lake and EP2 have very similar morphologies, water chemistry, are located

within ~20 m of each other, and are connected by Skeleton Creek, yet the historical composition

of diatom species assemblages is strikingly different between the two lakes. Although both lakes

must have experienced warming to some degree, Skeleton Lake has had more persistent ice cover

than EP2. Thus, it was expected that diatom assemblage shifts consistent with climate warming

would be recorded in EP2, but that only a muted diatom response to the same environmental

drivers would be apparent in Skeleton Lake due to shading that leads to a cooler microclimate and

extended ice cover (Smol, 1988).

In Skeleton Lake, the consistent domination of the diatom assemblages by small, benthic

Fragilaria taxa is likely indicative of cool conditions characterized by extensive ice cover (e.g.

Smol, 1988; Douglas and Smol, 1999; Lotter and Bigler, 2000). In the most recent sediments, the

subtle increase in F. brevistriata is similar to that found in the very recent sediments, representing

the warmest years on record, of Char Lake, a large, mostly ice-covered high Arctic lake on

Cornwallis Island (Michelutti et al., 2003a). Thus, Skeleton Lake sediments indicate a relatively

muted response to recent regional warming, consistent with its more extended ice cover.

In Lake EP2, on the other hand, shifts in diatom assemblages from those indicative of

cool, harsh conditions towards a diversification of taxa indicative of more diverse substrate

availability (e.g. moss epiphytes such as Denticula kuetzingii; Lim et al., 2001; Michelutti et al.,

2003b), and larger, more complex frustules (e.g. Cymbella taxa) are consistent with climate

warming and reduced ice cover. Such an assemblage shift is consistent with climatic warming

and reduced ice cover duration that would have led to a longer growing season for aquatic

macrophytes and algae, as well as enhanced nutrient export from the catchment (Smol, 1988;

Douglas and Smol, 1999). These types of diatom species changes are consistent with many other

diatom changes seen in lakes and ponds throughout the Arctic (e.g. Douglas et al., 1994;

Michelutti et al., 2003a; Antoniades et al., 2005; Smol et al., 2005).

Alternative hypotheses

Other factors, such as acid precipitation, atmospheric deposition of anthropogenic

nitrogen or other nutrients or pollutants, and/or some other unmeasured phenomenon, may

potentially be proposed as alternate drivers explaining the diatom assemblage changes in Skeleton

Lake and EP2. A comparison of the diatom profiles between the two lakes addresses these other

factors.

The remote location of our study sites, far from point sources of pollution of any type,

identifies atmospheric deposition as the major pathway of pollutant transport to these sites. Due

to the proximity of our sites, both lakes are subjected to identical types and quantities of

atmospheric deposition. Nonetheless, if some type of atmospheric deposition was affecting the

lakes differentially, we should see differences in our water chemistry data, especially with respect

to pH, dissolved nitrogen, and sulphate. However, both Skeleton Lake and EP2 are presently

nearly identical with respect to these variables (Table 1); with the exceptions of POC, PON, and

NH3, which have been discussed above. Furthermore, historical water chemistry measurements

indicate that Skeleton Lake and EP2 have remained circumneutral to alkaline and, if anything,

have slightly increased in pH since 1964 (Oliver and Corbet, 1966). These historical data also

confirm that the water chemistry of Skeleton Lake and EP2 were very similar in 1964 (Oliver and

Corbet, 1966), just as it was during our field work in 2003.

Enrichment from atmospheric nitrogen has been linked to diatom community shifts in

lakes in the Rocky Mountains, USA (e.g. Wolfe et al., 2001; Saros et al., 2003). In Arctic lakes

on Baffin Island, nitrogen deposition may have acted in concert with climate changes to cause

diatom shifts after 1950, although diatom community change commenced in the mid-19th century

(Wolfe et al., 2006). Although our extremely remote study sites must have received

comparatively less atmospheric N input than those in the continental USA, persistent ice on

Skeleton Lake could lead to differences in the timing of N delivery to the two lakes, thereby

eliciting unique ecological responses. However, the diatom changes in the very remote Lake EP2

began over the last ~100-150 years, based on our approximate chronological estimate, while

atmospheric N deposition in comparatively impacted lakes in the continental USA only become

an ecologically important source of N after the mid- to late 20th century (e.g. Wolfe et al., 2001;

Saros et al., 2003). Thus, even if differences in timing of delivery of N deposition were affecting

diatom communities, this could only explain much more recent changes, and in any case, would

ultimately still be traced back to differences in ice cover regime.

With respect to other types of pollution, previous paleolimnological studies have shown

that no changes in diatom assemblages have been recorded in lakes that have experienced point-

source pollution from PCBs (Paterson et al., 2003). Thus, differences in acidification, amount of

nitrogen deposition, and pollution cannot explain the types of diatom responses recorded in the

Skeleton Lake and EP2 cores.

The results from the sediment cores from Skeleton Lake and EP2, as well as the surface

sediment diatoms from EP3, all suggest that ice cover plays an important role in dampening

diatom community shifts to environmental change. This is likely due to the role of ice cover in

restricting the growing season for both algae and their potential substrates (e.g. mosses and

grasses). However, due to microclimatic differences, the ice cover continues to persist longer at

Skeleton Lake than at EP2. Thus, the changes recorded in the diatom records may reflect the

crossing of some type of ice cover threshold in EP2 that has not yet been reached in Skeleton

Lake, as the climatic changes have not yet been great enough to reduce ice cover to the point at

which dramatic diatom shifts are occurring (Fig. 7). Furthermore, these results underscore the

importance of careful site selection and contextual interpretation in paleolimnological studies, as

local conditions may greatly influence ecological responses to regional environmental change

(Smol et al., 2005).

CONCLUSIONS

Although Skeleton Lake and EP2 are nearly identical with respect to water chemistry,

morphology, and are located within ~ 20 m of each other, the two lakes have recorded strikingly

different diatom histories. The major difference between the two sites is the longer length of ice

cover on Skeleton Lake relative to EP2. Both the muted diatom assemblage change in Skeleton

Lake and the marked changes in EP2 suggest that ice cover plays an important role in influencing

biological assemblages in polar lakes.

ACKNOWLEDGEMENTS

Funding for this project came from Natural Sciences and Engineering Research Council of

Canada (NSERC) grants to BEK, MSVD, and JPS, and from a Northern Scientific Training

Program (NSTP) grant to BEK. We would also like to thank the Polar Continental Shelf Project

and Parks Canada for logistical support. We are grateful to the National Laboratory for

Environmental Testing (NLET) at Environmental Canada for water chemistry analysis, S. Arnott

and J. Glew for assistance in the field, and J.R. Glew for drafting the detailed inset map. Many

thanks to D. Antoniades, N. Michelutti, and the anonymous journal reviewers for constructive

comments on the manuscript, and to one of the journal reviewers for providing Figure 4. This is

PCSP contribution number 010-07.

REFERENCES

ACIA, 2004: Impacts of a warming Arctic. Arctic Climate Impact Assesment. Cambridge: Cambridge University Press, 140 pp.

Antoniades, D., Douglas, M.S.V., and Smol, J.P., 2005: Quantitative estimates of recent environmental changes in the Canadian High Arctic inferred from diatoms in lake and pond sediments. Journal of Paleolimnology, 33: 349 - 360.

Antoniades, D., Hamilton, P.B., Douglas, M.S.V., and Smol, J.P., in press: Freshwater diatoms from the Canadian High Arctic, submitted to Diatomologica Iconographica.

Bindler, R., Renberg, I., Anderson, N.J., Appleby, P.G., Emteryd, O., and Boyle, J., 2001a: Pb isotope ratios of lake sediments in West Greenland: inferences on pollution sources. Atmospheric Environment, 35: 4675 - 4685.

Bindler, R., Renberg, I., Appleby, P.G., Anderson, N.J., and Rose, N.L., 2001b: Mercury accumulation rates and spatial patterns in lake sediments from west Greenland: A coast to ice margin transect. Environmental Science and Technology, 35: 1736 - 1741.

Binford, M.W., 1990: Calculation and uncertainty analysis of 210Pb dates for PIRLA project lake sediment cores. Journal of Paleolimnology, 3: 253 - 267.

Braun, C., Hardy, D.R., and Bradley, R.S., 2004: Mass balance and area changes of four High Arctic plateau ice caps, 1959-2002. Geografiska Annaler Series A-Physical Geography, 86A: 43 - 52.

Christie, R.L., 1964: Geological reconnaissance of northeastern Ellesmere Island, District of Franklin. Ottawa: Geological Survey of Canada Memoir 331: 79 pp.

Doubleday, N.C., Douglas, M.S.V., and Smol, J.P., 1995: Paleoenvironmental studies of black carbon deposition in the High Arctic - a case-study from northern Ellesmere Island. Science of the Total Environment, 161: 661 - 668.

Douglas, M.S.V., Hamilton, P.B., Pienitz, R., and Smol, J.P., 2004: Algal indicators of environmental change in Arctic and Antarctic lakes and ponds. In Pienitz, R., Douglas, M.S.V., and Smol, J.P. (eds.) Long-term environmental change in Arctic and Antarctic lakes. Dordrecht: Springer, 117 - 158.

Douglas, M.S.V., and Smol, J.P., 1999: Freshwater diatoms as indicators of environmental change in the High Arctic. In Stoermer, E.F. and Smol, J.P. (eds.) The Diatoms: Applications for the Environmental and Earth Sciences. Cambridge: Cambridge University Press, 227 - 244.

Douglas, M.S.V., Smol, J.P., and Blake Jr, W., 1994: Marked post-18th century environmental change in high-arctic ecosystems. Science, 266: 416 - 419.

Environment Canada, 1994: Manual of Analytical Methods National Laboratory for Environmental Testing, Canadian Centre for Inland Waters. Burlington, Canada, 2 vols., 1097 pp.

Glew, J.R., 1988: A portable extruding device for close interval sectioning of unconsolidated core samples. Journal of Paleolimnology, 2: 241 - 243.

Glew, J.R., 1989: A new trigger mechanism for sediment samplers. Journal of Paleolimnology, 2: 241 - 243.

Heiri, O., Lotter, A.F., and Lemcke, G., 2001: Loss on ignition as a method for estimating organic and carbonate content in sediments: reproducibility and comparability of results. Journal of Paleolimnology, 25: 101 - 110.

Hill, M.O., 1973: Diversity and evenness: a unifying notation and its consequences. Ecology, 54: 427 - 432.

Jackson, C.I., 1959: Operation Hazen: The Meteorology of Lake Hazen, N.W.T. Part 1: Analysis of the observations. Ottawa: Defence Research Board, 194 pp.

Joynt, E.H. III, and Wolfe, A.P., 2001: Paleoenvironmental inference models from sediment diatom assemblages in Baffin Island lakes (Nunavut, Canada) and reconstruction of summer water temperatures. Canadian Journal of Fisheries and Aquatic Sciences, 58: 1222 - 1243.

Karst-Riddoch, T.L., 2004: Sedimentary diatoms from sensitive alpine (northwest Canada) and subarctic (Iceland) lakes as indicators of environmental and climatic conditions. PhD thesis, Department of Biology, Queen’s University, Kingston, Canada. 194 pp.

Keatley, B.E., Douglas, M.S.V., and Smol, J.P., [4]. Limnological characteristics of a high Arctic

oasis and comparisons across northern Ellesmere Island. Arctic. in press. Koerner, R.M. and Fisher, D.A., 1990: A record of Holocene summer climate from a Canadian

high-Arctic ice core. Nature, 343: 630 - 631.

Krammer, K., 2002: Diatoms of Europe: Cymbella. Ruggell: A.R.G. Gantner Verlag, 584 pp.

Krammer, K. and Lange-Bertalot, H., 1986-1991: Bacillariophyceae, Stuttgart: Gustav Fisher Verlag., 2485 pp.

Lim, D.S.S., Kwan, C., and Douglas, M.S.V., 2001b: Periphytic diatom assemblages from Bathurst Island, Nunavut, Canadian High Arctic: An examination of community relationships and habitat preferences. Journal of Phycology, 37: 379-392.

Lepš, J., and Šmilauer, P., 2003: Multivariate Analysis of Ecological Data using CANOCO, Cambridge: Cambridge University Press, 269 pp.

Lotter, A.F. and Bigler, C., 2000: Do diatoms in the Swiss Alps reflect the length of ice-cover? Aquatic Sciences, 62: 125 - 141.

Meteorological Service of Canada, 2006: Monthly Climate Data. Environment Canada http://www.climate.weatheroffice.ec.gc.ca/climateData/monthlydata_e.html , Accessed July 2006

Michelutti, N., Douglas, M.S.V., and Smol, J.P., 2003a: Diatom response to recent climatic change in a high arctic lake (Char Lake, Cornwallis Island, Nunavut). Global and Planetary Change, 38: 257 - 271.

Michelutti, N., Douglas, M.S.V., and Smol, J.P., 2003b: Periphytic diatom assemblages from ultra-oligotrophic and UV transparent lakes and ponds on Victoria Island and comparisons with other diatom surveys in the Canadian Arctic. Journal of Phycology, 39: 465-480.

Oliver, D.R. and Corbet, P.S., 1966: Operation Hazen: Aquatic habitats in a high arctic locality: the Hazen Camp study area, Ellesmere Island, N.W.T. Ottawa: Department of National Defence. 126 pp.

Parr, J.F., Taffs, K.H., and Lane, C.M., 2004: A microwave digestion technique for the extraction of fossil diatoms from coastal lake and swamp sediments. Journal of Paleolimnology, 31: 383 - 390.

Paterson, A.M., Betts-Piper, A.A., Smol, J.P., and Zeeb, B.A., 2003: Diatom and chrysophyte algal response to long-term PCB contamination from a point-source in northern Labrador, Canada. Water Air and Soil Pollution, 145: 377 - 393.

Perren, B.B., Bradley, R.S., and Francus, P., 2003: Rapid lacustrine response to recent High Arctic warming: A diatom record from Sawtooth Lake, Ellesmere Island, Nunavut. Arctic Antarctic and Alpine Research, 35: 271 - 278.

Pienitz, R., Douglas, M.S.V., and Smol, J.P., 2004: Long-Term Environmental Change in Arctic and Antarctic Lakes. Dordrect: Springer, 550 pp.

Quesada, A., Vincent, W. F., and Lean, D.R.S., 1999: Community and pigment structure of Arctic cyanobacterial assemblages: the occurrence and distribution of UV-absorbing compounds. FEMS Microbiology Ecology, 28: 315-323.

Rühland, K., Priesnitz, A., and Smol, J.P., 2003: Paleolimnological evidence from diatoms for recent environmental changes in 50 lakes across Canadian Arctic treeline. Arctic Antarctic and Alpine Research, 35: 110 - 123.

Saros, J.E., Interlandi, S.J., Wolfe, A.P., and Engstrom, D.R., 2003: Recent changes in the diatom community structure of lakes in the Beartooth Mountain Range, USA. Arctic Antarctic and Alpine Research, 35: 18 - 23.

Smith, I.R., 2002: Diatom-based holocene paleoenvironmental records from continental sites on northeastern Ellesmere Island, high Arctic, Canada. Journal of Paleolimnology, 27: 9 - 28.

Smith, Jr.S.V., Bradley, R.S., and Abbott, M.B., 2004: A 300 year record of environmental change from Lake Tuborg, Ellesmere Island, Nunavut, Canada. Journal of Paleolimnology, 32: 137 - 148.

Smol, J.P., 1983: Paleophycology of a high arctic lake Cape Herschel, Ellesmere Island. Canadian Journal of Botany, 61: 2195 - 2204.

Smol, J.P., 1988: Paleoclimate proxy data from freshwater Arctic diatoms. Verh. int. Ver. Limnol., 23: 837 - 844.

Smol, J.P., 2002: Pollution of lakes and rivers: a paleoenvironmental perspective. London: Arnold, 280 pp.

Smol, J.P., Wolfe, A.P., Birks, H.J.B., Douglas, M.S.V., Jones, V.J., Korhola, A., Pienitz, R., Ruhland, K., Sorvari, S., Antoniades, D., Brooks, S.J., Fallu, M.A., Hughes, M., Keatley, B.E., Laing, T.E., Michelutti, N., Nazarova, L., Nyman, M., Paterson, A.M., Perren, B., Quinlan, R., Rautio, M., Saulnier-Talbot, E., Siitoneni, S., Solovieva, N., and Weckstrom, J., 2005: Climate-driven regime shifts in the biological communities of arctic lakes. Proceedings of the National Academy of Sciences of the United States of America, 102: 4397 - 4402.

Solovieva, N., Jones, V.J., Nazarova, L., Brooks, S.J., Birks, H.J.B., Grytnes, J.A., Appleby, P.G., Kauppila, T., Kondratenok, B., Renberg, I., and Ponomarev, V., 2005: Palaeolimnological evidence for recent climatic change in lakes from the northern Urals, Arctic Russia. Journal of Paleolimnology, 33: 463 - 482.

Soper, J.H. and Powell, J.M., 1985: Botanical studies in the Lake Hazen Region, northern Ellesmere Island, Northwest Territories, Canada. Ottawa: Publications in Natural Sciences: National Museums of Canada, 67 pp.

Sorvari, S., Korhola, A., and Thompson, R., 2002: Lake diatom response to recent Arctic warming in Finnish Lapland. Global Change Biology, 8: 171 - 181.

Stoermer, E.F. and Smol, J.P. (eds), 1999: The diatoms: applications for the environmental and earth sciences. Cambridge: Cambridge University Press, 469 pp.

Thompson, W., 1994: Climate. In: Resource description and analysis: Ellesmere Island National Park Reserve. National Resource Conservation Section, Prairie and Northern Region, Parks Canada, Department of Canadian Heritage, Winnipeg. 78 pp.

Thompson, R., Price, D., Cameron, N., Jones, V., Bigler, C., Rosen, P., Hall, R.I., Catalan, J.,

Garcia, J., Weckstrom, J., and Korhola, A., 2005: Quantitative calibration of remote mountain-lake sediments as climatic recorders of air temperature and ice-cover duration. Arctic Antarctic and Alpine Research, 37: 626 - 635.

Wolfe, A.P., 2000: A 6500-year record from southwestern Fosheim Peninsula, Ellesmere Island, Nunavut. In Garneau, M. and Alt, B.T. (eds.) Environmental Response to Climate Change in the Canadian High Arctic. Ottawa: Geological Survey of Canada, Bulletin 529, 249 - 256.

Wolfe, A.P., Baron, J.S., and Cornett, R.J., 2001: Anthropogenic nitrogen deposition induces rapid ecological changes in alpine lakes of the Colorado Front Range (USA). Journal of Paleolimnology, 25: 1 - 7.

Wolfe, A.P., Cooke, C.A., and Hobbs, W.O., 2006: Are current rates of atmospheric nitrogen deposition influencing lakes in the Eastern Canadian Arctic? Arctic Antarctic and Alpine Research, 38: 465 - 476.

Figure captions. Figure 1. Map showing the locations of the study site (star) and the other sites (numbered) mentioned in the text: 1. Alert, 2. Hazen Plateau, 3. Agassiz Ice Cap, 4. Fosheim Peninsula, 5. Cape Herschel, 6. Isachsen, Ellef Ringnes Island, 7. Char Lake, Cornwallis Island. Figure 2. Topographical map detailing the local topography near Skeleton Lake, EP2 and EP3. Figure 3. Ice-off dates for Skeleton Lake (solid bars) and Lake EP2 (hatched bars). The data from the 1960s is from Oliver and Corbett (1966), while the 2003 data is from our own field observations. These data are also corroborated by archival air photos that show ice persisting on Skeleton Lake after Lake EP2 has become ice-free. Figure 4. Principal components analysis (PCA) biplot based on measured water chemistry variables that are considered to influence diatom assemblages. The arrows represent the measured environmental variables, while the open circles represent 52 lakes and ponds across northern Ellesmere Island. Skeleton Lake is represented by the solid square, EP2 is represented by the open square, and EP3 is represented by the solid circle. The proximity of these three sites represents their highly similar water chemistry. Figure 5. a) Age-depth model for Skeleton Lake based on CRS model (Binford, 1990), and b) Total 210Pb activities as estimated by alpha spectroscopy. The dashed line indicates estimated supported 210Pb. Figure 6a. Diatom profile of Skeleton Lake showing taxa present in at least at least one interval with a relative abundance of >3%. See Table 2 for synonyms for some of the common taxa. Percent loss-on-ignition (%LOI; an estimate of organic matter) and PCA axis 2 sample scores (PCA2; a summary of change occurring in rare taxa) are presented at the right side of the profile. Figure 6b. Photographs of ice cover on Skeleton Lake and Lake EP2, indicating the physical proximity of the two lakes, Blister Hill, and a nearby pingo for reference (photographs taken 7 July 2003). Figure 6c. Diatom profile of EP2 showing only species that are present in at least 3% relative abundance in at least one interval. See Table 2 for synonyms for some of the common taxa. Percent loss-on-ignition (% LOI), total Pb and total Hg (both expressed per gram organic carbon), and PCA axis two sample scores (PCA2) are presented at the right side of the profile. The rise in total Pb and Hg are interpreted to mark the onset of anthropogenic pollution (mid-19th to early 20th century). Figure 7. Schematic diagrams illustrating possible diatom responses to changing ice cover conditions in the two study lakes. Although the length of the ice-free season has likely increased in both lakes, it has yet to reach a critical threshold in Skeleton Lake, resulting in a muted diatom response. In Lake EP2, reduction in ice cover has crossed this critical threshold and resulted in marked diatom changes.

Figure 1.

Figure 2.

27-Jun

02-Jul

07-Jul

12-Jul

17-Jul

22-Jul

27-Jul

1963 1964 2003

Year

Dat

e

Figure 3.

-0.4 1.0

-1.0

1.0

pH

conductivity

Chla

DICDOC

NH3

TKN

TdNTN

TPu

TPf

SRP

SiO2

Ca

K

Na

Mg

ClSO4

Figure 4.

Year AD

1860

1880

1900

1920

1940

1960

1980

2000

Dep

th (c

m)

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

A

Total 210Pb (pCi/g)

0.51.01.52.0

Dep

th (c

m)

0.5

1.5

2.5

3.5

4.5

5.5

6.5

7.5

32.5B

Figure 5.

0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32Depth (cm)

20F. pinn

ata 2550

100

F. con

strue

nsva

r.

vente

rcomple

x

20F. brev

istria

ta

F. pse

udoc

onstr

uens

2040

Rel

ativ

e ab

unda

nce

(%)

%

LOI

-20

2

PCA2

75

Depth (cm)

Rel

ativ

e ab

unda

nce

(%)

%pp

b/g

Cpp

m/g

C

F. con

strue

nsva

r.

vente

r F. pinn

ataN. s

oehre

nsis

A. minu

tissim

a

C. ang

ustat

a

Nitzsc

hiaco

mplex

C. silic

ulaA. p

eterse

nii

D. kue

tzing

ii F. brev

istria

ta

LOI

Hg

Pb

PCA2

0 2 4 6 8 10 12 14 16

2020

4060

8020

2040

2025

5075

100

2040

1575

135

2040

60-2

02

A C

Skel

eton

Lak

e

ping

o

Blis

ter

Hill

Skel

eton

Cre

ek

EP2

B

Fi

gure

6.

ice-free period

chan

ge in

dia

tom

as

sem

blag

eSkeleton Lake

~1850 2003

small

large

several weeks

several months

ice-free period

chan

ge in

dia

tom

as

sem

blag

e

EP2

~1850

2003

small

large

several weeks

several months

Figure 7.

Table 1. Selected limnological characteristics for the three study lakes. Skeleton Lake EP2 EP3Latitude (N) 81˚49.798’ 81˚49.845’ 81˚49.884’Longitude (W) 71˚28.483’ 71˚28.352’ 71˚28.052’Date sampled 08/07/2003 08/07/2003 08/07/2003Depth (m) ~4 ~3 ~2Temperature (°C) 6 13.5 12pH 8.2 8.2 8.1Specific conductivity (μS/cm) 175 180 187

SiO2 (mg/L) 4.61 5.21 5.7Na (mg/L) 1.42 1.26 1.41Ca (mg/L) 38.8 41.7 44.3K (mg/L) 1.09 0.85 0.98Mg (mg/L) 6.53 5.59 6.04SO4 (mg/L) 41.7 30.3 33.5Cl (mg/L) 0.61 0.58 0.56Particulate organic carbon (POC; mg/L) 0.529 1.24 1.49Particulate nitrogen (PON; mg/L) 0.084 0.259 0.334Dissolved organic carbon (DOC; mg/L) 5.8 4.6 4.1Dissolved inorganic carbon (DIC; mg/L) 20.7 22.8 25.9NO3-NO2 (mg/L) 0.005 0.006 <0.005NH3 (mg/L) 0.014 0.032 0.01Total Kjeldahl nitrogen (TKN; mg/L) 0.314 0.296 0.298Total dissolved nitrogen (TdN; mg/L) 0.354 0.346 0.275Total phosphorus (Tpu; mg/L) 0.0095 0.0075 0.0056Total phosphorus filtered (TPf; mg/L) 0.0052 0.0036 0.0032Chlorophyll a (μg/L) 0.6 0.6 0.9Total N : Total P (molar) 61 112 159

Table 2. Diatom taxa (>3% relative abundance) from the top 2.5 cm intervals from the sediment cores of Skeleton Lake and EP2 and the surface sediment (~2 cm) diatoms from EP3.

Relative abundance

(%) Skeleton Lake Fragilaria brevistriata 11.26 Fragilaria construens var. venter 85.78 EP2 Denticula kuetzingii 7.55 Fragilaria brevistriata 83.88 Nitzschia complex 4.95 EP3 Achnanthes minutissima 7.24 Achnanthes petersenii 5.36 Brachysira zellensis 3.22 Cymbella angustata 5.36 Cymbella microcephala 4.83 Cymbopleura sp. 8.85 Denticula tenuis 4.29 Eunotia praerupta 4.29 Fragilaria brevistriata 19.03 Fragilaria pinnata 5.90 Navicula chiarae 4.29 Nitzschia complex 7.24

Tabl

e 3.

Com

mon

dia

tom

taxa

(>3%

rela

tive

abun

danc

e) fo

und

in th

e se

dim

ent c

ores

from

Ske

leto

n La

ke a

nd E

P2.

Num

bers

in

pare

nthe

ses a

fter t

he la

ke n

ame

repr

esen

t the

leng

ths o

f the

DC

A a

xis 1

gra

dien

ts, i

n st

anda

rd d

evia

tion

units

, N is

the

num

ber o

f oc

curr

ence

s, N

2 is

the

effe

ctiv

e nu

mbe

r of o

ccur

renc

es (H

ill, 1

973)

and

Max

% is

the

max

imum

rela

tive

abun

danc

e. S

ynon

yms f

or

taxa

with

revi

sed

nam

es a

re li

sted

in th

e fa

r rig

ht c

olum

n.

N

N

2 M

ax

%

Syno

nym

Sk

elet

on L

ake

(1.0

6 SD

)

Frag

ilari

a br

evis

tria

ta G

runo

w in

Van

Heu

rck

33

29.6

12

Ps

eudo

stau

rosi

ra b

revi

stri

ata

(Gru

now

in V

an H

eurc

k)

Will

iam

s & R

ound

Fr

agila

ria

cons

true

ns v

ar. v

ente

r (Eh

renb

erg)

Gru

now

in V

an

Heu

rck

33

33.0

97

St

auro

sira

con

stru

ens v

ar. v

ente

r (Eh

renb

erg)

Ham

ilton

Fr

agila

ria

pinn

ata

Ehre

nber

g 33

27

.1

18

Stau

rosi

rella

pin

nata

(Ehr

enbe

rg) W

illia

ms &

Rou

nd

Frag

ilari

a ps

eudo

cons

true

ns

26

18.8

10

Ps

eudo

stau

rosi

ra p

seud

ocon

stru

ens

EP2

(2.0

0 SD

)

Ac

hnan

thes

min

utis

sim

a K

ützi

ng

14

10.5

13

Ac

hnan

thid

ium

min

utis

sim

um (K

ützi

ng) C

zarn

ecki

Achn

anth

es p

eter

seni

i Hus

tedt

14

11

.0

5 Ro

ssith

idiu

m p

eter

seni

i (H

uste

dt) B

ukht

iyar

ova

&

Rou

nd

Cal

onei

s sili

cula

(Ehr

enbe

rg) C

leve

12

10

.8

4 C

alon

eis s

ilicu

la (E

hren

berg

) Cle

ve

Cym

bella

ang

usta

ta

13

10.9

5

Cym

bopl

eura

ang

usta

ta v

ar. s

pitz

berg

ensi

s Kra

mm

er

Den

ticul

a ku

etzi

ngii

Gru

now

16

13

.4

24

Den

ticul

a ku

etzi

ngii

Gru

now

Frag

ilari

a br

evis

tria

ta G

runo

w in

Van

Heu

rck

17

12.3

98

Ps

eudo

stau

rosi

ra b

revi

stri

ata

(Gru

now

in V

an H

eurc

k)

Will

iam

s & R

ound

Fr

agila

ria

cons

true

ns v

ar. v

ente

r (Eh

renb

erg)

Gru

now

in V

an

Heu

rck

10

8.8

10

Stau

rosi

ra c

onst

ruen

s var

. ven

ter (

Ehre

nber

g) H

amilt

on

Frag

ilari

a pi

nnat

a Eh

renb

erg

17

13.1

79

St

auro

sire

lla p

inna

ta (E

hren

berg

) Will

iam

s & R

ound

N

avic

ula

soeh

rens

is K

rass

ke

11

8.2

6 C

ham

aepi

nnul

aria

soeh

rens

is (K

rass

ke) L

ange

-Ber

talo

t N

itzsc

hia

com

plex

16

12

.7

32

Nitz

schi

a co

mpl

ex

CHAPTER 7

IMPACTS OF SEABIRD-DERIVED NUTRIENTS ON WATER QUALITY AND DIATOM SPECIES

ASSEMBLAGES FROM CAPE VERA, DEVON ISLAND, CANADIAN HIGH ARCTIC

BRONWYN E. KEATLEY, MARIANNE S.V. DOUGLAS, JULES M. BLAIS, MARK MALLORY, AND

JOHN P. SMOL

Abstract

Allochthonous nutrient subsidies play an important role in shaping biotic communities,

particularly in naturally oligotrophic ecosystems such as the Canadian Arctic. Seabirds have been

identified as one vector capable of transporting significant quantities of nutrients from marine to

terrestrial environments. However, the influence of seabird-derived nutrients on water quality

and ecosystem functioning of Arctic freshwater systems has been largely unexplored. Here we

sampled 24 ponds on Cape Vera, Devon Island (Arctic Canada), an area that is home to over 10

000 nesting pairs of northern fulmars, in order to explore the influence of seabirds on water

quality and diatom ecology. Our data reveal distinctive water chemistry changes (i.e. high pH,

nutrients) and diatom assemblages (i.e. characterized by extremely low diversity), related to

varying degrees of bird influence. Using δ15N as an ecological tracer, we show that a small but

significant portion of the variance in the diatom species data can be explained by seabird-derived

nutrients. Although the generation of robust quantitative models predicting δ15N from fossil

diatom data was limited by the overwhelming dominance of a few taxa, this study provides

evidence that seabird-derived nutrients play a critical role in influencing the water quality of

Arctic ponds. Interestingly, diatom assemblage composition does not respond in a simple fashion

to seabird-derived nutrients, in contrast to diatom-nutrient relationships recorded in many

temperate regions.

Introduction

Allochthonous nutrient subsidies (e.g. from fertilizer, sewage, fish, or birds) can play

significant roles in food web functioning across diverse environments (Polis et al. 1997; Polis et

al. 2004). Generalizations concerning the ways ecosystems respond to nutrient inputs, however,

are difficult to make, in part due to the differences in temporal and spatial scales, trophic

interactions within food webs, and differences in the types of nutrient fluxes (Polis et al. 1997;

Vanni et al. 2004; Ellis 2005). Previous research has shown, for example, that production of

terrestrial plants and the abundance of beetles and arthropods are greater on areas influenced by

seabird activities than those without (Polis & Hurd 1996; Sánchez-Piñero & Polis 2000).

Conversely, terrestrial production may be reduced in areas of high seabird activity (see Ellis

2005) and plant species richness can be suppressed in regions receiving large quantities of seabird

guano (Anderson & Polis 2004; Wait et al. 2005). Similarly divergent results have been found

for microbial communities (Tscherko et al. 2003) and abiotic aspects (Ellis 2005) of seabird-

influenced soils.

Seabirds that feed in marine environments, but nest on land, represent important links

between marine, terrestrial, and, by extension, freshwater environments. Indeed, on a global

basis, seabirds are estimated to consume a similar volume of fish to that harvested by commercial

fishing operations (Brooke 2004). Yet, while seabirds have been used as monitors of

environmental change in both marine (reviewed in Diamond & Devlin 2003) and terrestrial

(reviewed in Ellis 2005) ecosystems, few studies have assessed the impact of seabird-derived

nutrients on freshwater quality or biota (Izaguirre et al. 1998; Harding et al. 2004). In temperate

and subarctic regions, non-marine birds, including ducks, geese, and gulls, have been shown to

deposit significant amounts of nutrients into freshwater lakes (Manny et al. 1994; Kitchell et al.

1999; Payne & Moore 2006). In Arctic regions, only a few studies have documented the

transport of nutrients and pollutants into freshwater ponds and lakes via seabird guano (Evenset et

al. 2004; Blais et al. 2005; Evenset et al. 2007).

As many Arctic lakes and ponds are naturally oligotrophic (Vincent & Hobbie 2000),

nutrient subsidies from allochtononous sources can significantly alter their water quality and

biological communities, although ecological responses can be complex. For example, Bonilla et

al. (2005) found that nutrient additions stimulated phytoplanktonic, but not benthic, communities

in a High Arctic lake. Diatoms (unicellular algae of the class Bacillariophyceae) have been used

to track nutrient additions in a wide range of freshwater habitats (reviewed by Hall & Smol

1999), and previous work in Arctic regions has shown that diatoms have responded, albeit in a

manner different from responses found in temperate regions, to nutrient enrichment from sewage

(Douglas & Smol 2000; Michelutti et al. 2002a), the decay of whale carcasses (Douglas et al.

2004a), and salmon (Gregory-Eaves et al. 2003). Stable isotopes of N (δ15N) are commonly

employed to track marine-derived eutrophication, as upper trophic level consumers in marine

foodwebs are enriched in δ15N compared to their terrestrial counterparts (Finney et al. 2000;

Robinson 2001; Douglas et al. 2004a). Seabirds, however, represent a far more common nutrient

source to Arctic lakes and ponds than salmon or whales as the Canadian Arctic is home to over 10

million seabirds (Mallory & Fontaine 2004). Recently, Blais et al. (2005) demonstrated that

seabirds are important vectors of marine-derived nutrients and contaminants in a series of 11

ponds located near Cape Vera, Devon Island, Arctic Canada, but the significance of seabird-

derived nutrients on freshwater biota remains unknown.

Here, we build upon the Blais et al. (2005) study by extending the original 11 study sites

to encompass 24 ponds across a seabird-derived nutrient gradient near Cape Vera, Devon Island

in order to: 1) determine the influence of seabirds on water chemistry of freshwater Arctic ponds

by comparing limnological changes in ponds across this gradient of seabird influence (as well as

to other High Arctic ponds unaffected by seabirds); 2) examine the distributions and patterns of

community structure of diatoms across this gradient of seabird influence; 3) determine which

environment variables can best explain diatom distributions in this region. This study also

provides autecological diatom data for a companion paleolimnological study that extends the

results of the modern limnological survey to an historical context (Keatley et al. [8]).

We found that sites located within a range of seabird influence at Cape Vera exhibited

highly distinctive water chemistry, exemplified by elevated nutrients and related variables, as

well as by high trace metal concentrations. The diatom assemblages recorded in surface

sediments of the ponds are unlike those found elsewhere in the High Arctic and are characterized

by high dominance of very few taxa. Finally, we determined that δ15N (a proxy for seabird

influence) can explain a small but significant portion of the variation in diatom species

distributions, although, surprisingly, the most abundant taxon showed no direct relationship to

seabird-derived nutrients.

Site description

Cape Vera (76°15’N, 89°15’W; Figure 1) is located at the northwestern portion of Devon

Island, Nunavut, in the Canadian High Arctic. Cape Vera is characterized by cold winters, cool

summers, and little precipitation. Long-term meteorological data do not exist for Cape Vera, but

measurements from the nearest weather station at Grise Fiord, Ellesmere Island, indicate mean

January and July temperatures of -31.8°C and 3.4°C, respectively, and a mean annual

precipitation of 167 mm (Meteorological Survey of Canada 2007).

The ~250 m high dolostone and limestone cliffs (Mayr et al. 1998) at Cape Vera house a

large colony (~10 000 breeding pairs) of northern fulmars (Fulmarus glacialis), which are

medium-sized petrels. These birds feed near the top of the marine food chain on squid, capelin,

crustaceans (Hatch & Nettleship 1998), and are also known to scavenge polar bear kills in the

High Arctic; as such, their tissues are elevated in 15N, and have δ15N values averaging between

13-17‰ (reviewed in Mallory 2006). The northern fulmars arrive at Cape Vera to breed in early

May, during which time they travel as far as ~250 - 400 km away to feed. They incubate their

nests at various times throughout the summer season and depart in September for their over-

wintering grounds, thought to be in the North Atlantic (Mallory 2006). Beneath the nesting cliffs

at Cape Vera is a small foreland that contains several freshwater ponds, which, depending on

their proximity to the bird colony and local hydrographic features, receive varying amount of

seabird subsidies (including guano, regurgitated stomach oil, feathers, eggshells and carcasses).

Northern fulmars return to their breeding sites with extremely high fidelity, and, unlike many

other Arctic breeding seabirds, lay eggs at the same time each year, regardless of weather

conditions (Mallory 2006). In the Canadian Arctic, the northern fulmar colony at Cape Vera does

not appear to have changed in colony size since the 1970s (Gaston et al. 2006), although these

data are admittedly based on only a few surveys performed at irregular intervals.

We selected 24 ponds located across a “bird-affected” gradient near the colony at Cape

Vera (Figure 1). Thirteen of the twenty-four ponds (unofficial names: CV5-10, 12-15, 20, 31)

were in the immediate vicinity of the most densely populated section of the colony, while four

sites (CV1-4) were situated just to the north of this heavily-used section. Seven of our sites

currently receive no run-off from the bird-colony (Figure 1).

Methods

Sampling

We conducted our sampling during the first two weeks of July over a three year period

(2004, 2005, 2006); a strategy consistent with our other water quality surveys. Some sites were

sampled in multiple years (e.g. CV5, 6, 7, 8, 9, 10, 12, 13, 14, 15), but due to helicopter

availability, some sites were only sampled once during the three year period (e.g. CV1-4, 11, 22-

24). Water samples were collected from all ponds for water chemistry analyses, whereas surface

sediment samples were taken for diatom identification and enumeration. The protocols we used

were identical to those from our previous sampling years in other Arctic regions (Lim et al. 2001;

Michelutti et al. 2002b; Michelutti et al. 2002c; Lim & Douglas 2003; Antoniades et al. 2003a;

Antoniades et al. 2003b; Lim et al. 2005; Keatley et al. [2]; Keatley et al. [4]; Michelutti et al. in

press a), and sampling procedures followed those described in Environment Canada

(Environment Canada 1994). Specific conductivity (YSI model 33 conductivity meter), pH

(handheld Hanna pHEP meter), and temperature (handheld thermometer) were measured each

day at our base camp, while all other water chemistry analyses (major ions (Ca, Mg, Na, K, Cl,

SO4), nutrients (total phosphorus unfiltered (TPu), total phosphorus filtered (TPf), total dissolved

nitrogen (TdN), total Kjeldahl nitrogen (TKN, filtered), ammonia-nitrogen (NH3-N), nitrate-

nitrogen (NO3-N), nitrate-nitrite-nitrogen (NO3-NO2-N), particulate nitrogen (PON), trace metals

(Ag, Al, As, B, Ba, Be, Bi, Cd, Co, Cr, Cu, Fe, Ga, La, Li, Mn, Mo, Ni, Pb, Rb, Sb, Se, Sr, Tl, U,

V, Zn), dissolved organic carbon (DOC), dissolved inorganic carbon (DIC), particulate carbon

(POC), chlorophyll a) were performed at the National Laboratory for Environmental Testing

(Environment Canada 1994). Details on sampling methodology can be found in Appendix 1.

Surface sediments were sampled by hand from the top ~1 cm from bottom of each pond,

placed into Whirlpak® bags, and kept dark and cold (~4°C) until analysed for δ15N and diatoms.

Diatom frustules were prepared for identification following standard techniques with a MARSX

microwave digester (Parr et al. 2004). Digested material was rinsed several times with deionized

water until a neutral pH was reached, and the resultant slurries were plated on microscope slides

using Naphrax®, a high refractive mounting medium. Surface sediments were also analysed for

% C and N, and stable isotopes of δ15N at the G.G. Hatch Stable Isotope Laboratory, Ottawa,

Canada. Elemental analyses were performed for C and N using a Vario EL III (Elementar,

Germany), while stable isotopes of N were measured using a continuous-flow DeltaPlus

Advantage isotope ratio mass spectrometer (ThermoFinnigan, Germany) coupled with a ConFlo

III. Isotope data was normalized using previously calibrated internal standards, and analytical

precision is +/-0.2 ‰.

A minimum of 300, and often > 500, diatom valves were enumerated and identified from

each sediment sample using primarily Antoniades et al. (in press), Krammer (2002) and Krammer

and Lange-Bertalot (1991).

Statistical analyses

A Pearson’s correlation matrix with Bonferroni-adjusted probabilities was generated

using Systat 9.0 (SPSS 1999) to identify pairs of significantly correlated variables because highly

collinear variables can unduly influence multivariate ordinations. All multivariate statistical

analyses were performed using CANOCO version 4.5 (ter Braak & Šmilauer 2002). Measured

environmental variables were analysed using a series of Principal Components Analyses (PCA) to

assess main directions of variation in the dataset, and to validate our initial visual ranking across

the gradient of seabird influence. Since we had multi-year environmental data from some sites,

we used this to semi-quantitatively assess whether inter-year variation within a site was greater

than inter-pond variation within a sampling period (one sample per pond per year). For all

subsequent analyses, however, we used a single sample year. A PCA was used to determine the

main directions of variation in the measured environmental variables amongst the sites. To better

explore the impact of seabird activity on the most ecologically-relevant variables (e.g.,

phosphorus, nitrogen, carbon, pH, specific conductivity, chlorophyll a), we compared the

differences between the Cape Vera sites to other Devon Island sites far from seabird influence

(Lim & Douglas 2003; Lim 2004). In an attempt to identify how unique the Cape Vera ponds

were with respect to other Arctic limnological surveys, we performed a PCA of our measured

environmental variables (pH, specific conductivity, TPu, TPf, TNu, DOC) with those from

previous studies throughout the Canadian Arctic (Lim et al. 2001; Michelutti et al. 2002b;

Michelutti et al. 2002c; Lim & Douglas 2003; Antoniades et al. 2003a; Antoniades et al. 2003b;

Lim et al. 2005; Keatley et al. [2]; Keatley et al. [4]; Michelutti et al. in press a).

Diatom species counts were converted to relative abundances and these data were

analysed using a detrended correspondence analysis (DCA) to assess the gradient length of the

species data, and to determine whether linear or unimodal statistical techniques were more

appropriate. As the DCA indicated a short gradient (<2 standard deviation units), linear

ordination techniques (PCA and RDA) were further used to assess patterns in the diatom species

distributions (Lepš & Šmilauer 2003). In order to compare patterns of diatom diversity between

sites, Hill’s N2 (a measure of species evenness and essentially the number of very abundant

species) was calculated (Hill 1973).

A series of constrained ordinations (RDAs) were used to determine whether measured

environmental variables could explain significant portions of the variance in diatom species

distributions. First, the correlations between environmental variables were assessed using a

Pearson correlation with Bonferroni-adjusted probabilities. In this manner, the environmental

dataset was truncated to reduce the influence of highly collinear environmental variables. Next,

the species data were constrained to the measured environmental variables, and these were

assessed using forward selection techniques to determine which environmental variables could

explain significant portions of the species data using a series of RDAs. We constructed predictive

models to explore the feasibility of quantitatively inferring the most important measured

environmental variable using regression-calibration techniques in C2 version 1.3 (Juggins 2003).

Results and Discussion

Water chemistry

A summary of the measured environmental variables for each Cape Vera site is given in

Table 1. For the sake of brevity, we limit our discussion of water chemistry variables to pH,

specific conductivity, nutrients and related variables, and some metals, due to their ecological

interest. Full details of all water chemistry data can be found in Appendix 5. In the PCA biplot of

measured environmental variables, the sites we a priori identified as being most influenced by

seabirds plotted in the right-hand side of the diagram and were characterized by elevated pH,

conductivity, production (Chla), nutrients and related variables (N, P), and δ15N, our isotope

proxy of bird influence (Figure 2). In contrast, our control sites plotted on the left-hand site of the

ordination plot (Figure 2). The distinctive water chemistry of the most heavily-affected ponds

becomes even more apparent when compared to sites from the nearby Haughton Crater region

(Figure 3), which shares similar climate and geological features, yet is not influenced by seabird

activities (Lim & Douglas 2003).

While pH values across the High Arctic are generally reflective of alkaline conditions

(Hamilton et al. 2001), the seabird-influenced ponds at Cape Vera had the highest pH yet

recorded in the Canadian Arctic (10.75, Table 1). For comparison, the mean pH of 36 lakes and

ponds elsewhere on Devon Island was 8.3, despite sharing similar climatic and geological

features (Lim 2004). The only comparably high pH values recorded in the Canadian Arctic come

from small, hyper-eutrophic sewage ponds near Resolute Bay (Douglas & Smol 2000). The

elevated pH values at Cape Vera likely reflect the enhanced aquatic production stimulated by the

nutrients from seabird wastes. In temperate regions, for example, shallow rock pools influenced

by gull guano are known to exceed pH 10 on a daily basis, linked to high photosynthetic activity

(Loder et al. 1996).

Specific conductivity, although relatively high compared to elsewhere on Devon Island,

was not as high as might have been expected given the proximity of the ocean to many of these

ponds (Cape Vera range: 85 – 312 µS/cm; mean = 160 µS/cm; Table 1). The relatively low

conductivity, coupled with the large concentrations of DIC and Ca as the dominant ions, suggest

that, despite their proximity to the sea, the ponds at Cape Vera are not receiving large quantities

of marine inputs. Coastal ponds elsewhere in the Canadian Arctic often have recorded specific

conductivity values > 1000 µS/cm and are driven by high concentrations of Cl and Na (e.g.

Antoniades et al. 2003a).

Interestingly, trace metal concentrations in the ponds at Cape Vera were almost all above

method detection limits, which is unusual because metal concentrations rarely exceed detection

limits in the High Arctic. For example, of the nine High Arctic water chemistry surveys

completed to date, only sites located on northern Ellesmere Island had metal concentrations that

were as high as, or higher than, those from Cape Vera (Keatley et al. [4]). On northern Ellesmere

Island, the elevated metal concentrations may be related to geological sources. That the ponds

closest to the bird colony at Cape Vera had the highest metal concentrations is consistent with an

ornithogenic source of metals (Table 1). Indeed, previous studies have found that certain metals

are known to bioaccumulate in animals (e.g. Rb, Campbell et al. 2005; Cd, Cu, Hg, Zn, Liu et al.

2006).

The nutrient concentrations of the Cape Vera ponds were highest in the ponds closest to

the bird colony (Table 1, Figure 2). The high production in the bird-influenced ponds (as

estimated by chlorophyll a concentrations, Chlamean: 9.0 µg/L, Table 1) was accompanied by,

although not significantly correlated to, high phosphorus and nitrogen (e.g. TPu: range 4 – 207

µg/L, mean: 36.1 µg/L; TPf: 2.2-72.0 µg/L, mean: 14.3 µg/L; TdN: 0.072 – 1.53 mg/L, mean:

0.375 mg/L). In the context of over 400 lakes and ponds our labs have sampled from the

Canadian Arctic Archipelago, the Cape Vera sites include several of the most nutrient rich Arctic

freshwater ecosystems (Figure 4). A few other Arctic sites have equally high, or higher, TPu, but

these have been previously attributed to re-suspended particulate matter rather than to available

phosphorus, and thus not considered to be truly eutrophic systems (Antoniades et al. 2003b;

Michelutti et al. 2007a; Keatley et al. [2]).

In general, sites in the High Arctic that are characterized by relatively high production,

lush vegetation, and high nutrients also have high DOC (Lim et al. 2001; Antoniades et al. 2003a;

Lim et al. 2005; Keatley et al. [2]; Keatley et al. [4]). Interestingly, DOC concentrations at Cape

Vera were not nearly as high as we might have expected (range: 0.6 – 12.6 mg/L; mean: 3.0

mg/L), given the lush vegetation that surrounds many of these ponds. This can be most easily

seen by the composite PCA, in which the sites from the Arctic oasis near Lake Hazen are clearly

distinguished by their high DOC (Figure 4). This interesting finding might be due to a

decoupling of the dissolved organic matter cycle, whereby the high terrestrial and aquatic

production at Cape Vera exists, in spite of very harsh climatic conditions, as a function of the

external nutrient subsidies deposited by birds. In contrast, localized sites of relatively high

production elsewhere in the Arctic Archipelago are generally found in concert with warm

microclimatic conditions that are likely more conducive to decomposition of terrestrial organic

matter and thus higher DOC. The relatively low DOC in the Cape Vera sites corroborates our

evidence from δ15N that the nutrients in the ponds are generated from the seabirds rather than

terrestrial plant matter.

Diatom assemblages

A total of 95 taxa were identified from the 24 study sites at Cape Vera, although 44% of

these were considered very rare (<1% relative abundance). The assemblages were dominated by

small, benthic diatoms (Figure 5, Table 2), a typical finding in shallow Arctic ponds (Douglas et

al. 2004b). The dominant taxa are all common components of diatom assemblages in ponds in

the Canadian High Arctic (Antoniades et al. in press). However, their distributional patterns are

quite distinctive at Cape Vera. For example, many sites were dominated by small Nitzschia spp.

(often >90% relative abundance) and Cymbella sensu lato spp. (Figure 5). Elsewhere in the High

Arctic, the Cymbella spp. found at Cape Vera (e.g. C. cleve-eulerae, C. botellus, Encyonema.

minutum) are generally not found at >10% relative abundances and are typically associated with

oligotrophic, low electrolyte waters (Antoniades et al. in press). On the other hand, small

Nitzschia spp. (e.g. N. perminuta, N. frustulum (recorded by Antoniades et al. (in press) as N.

alpina)) are perhaps the most ubiquitous and abundant freshwater diatoms (max abundance 68%

and 26%, respectively) found in the Canadian High Arctic (Antoniades et al. in press), although

morphological similarities often make it difficult to distinguish these two taxa and hence

appreciate their true abundances and environmental preferences. For example, Michelutti et al.

(2003) found very high relative abundances (often > 90%) of the Nitzschia complex (N.

frustulum, N. perminuta, N. libertruthii, N. inconspicua) in the periphytic samples of ponds

receiving sewage near Resolute Bay, Nunavut. The extremely high relative abundance of these

small Nitzschia taxa at Cape Vera (often > 90%) suggests that N. frustulum and N. perminuta do

exceedingly well under high nutrient, high pH conditions. Interestingly, none of the most

common species found at Cape Vera were common components of the diatom flora of lakes and

ponds elsewhere on Devon Island (Lim 2004), further highlighting the uniqueness of the Cape

Vera sites.

In the PCA biplot, sites with more species in common plot closer together (e.g. CV9, 9a,

10), while those with very different species assemblages plot further away from each other

(Figure 6). Species associated with the higher δ15N sites included Encyonema minutim, E. fogedii,

and E. silesiacum, while N. frustulum plotted amongst the higher DOC sites (Figure 6). Species

characteristic of lower δ15N sites included small Achnanthes sensu lato spp. (Achnanthidium

minutissimum, Psammothidium marginulatum, Rossithidium petersenii, Achnanthes kriegeri,

Eucocconeis laevis), Denticula kuetzingii, Diatoma tenuis, Cymbella botellus, C. cleve-eulerae,

C. designata, Encyonopsis microcephala, and Cymbopleura angustata (Figure 6). Not

surprisingly, the control sites at the Cape Hawes (CV22-24) sites plotted closely together based

on the species data (Figure 6).

The diatom assemblage diversity is extremely low in most of the bird-influenced sites, as

indicated by the Hill’s N2 number (Table 2; N2 mean: 7.28; N2 range 1.0 – 15.0; Hill 1973).

High Arctic ponds are generally species-poor (mean Hill’s N2 values range from 6.72 (Cape

Herschel, Ellesmere Island; Douglas & Smol 1993) to 13.09 (Alert, Ellesmere Island; Antoniades

et al. 2005), but the overwhelming dominance by only a few taxa at Cape Vera is unusual.

Previous Arctic limnological studies have linked increased diversity of diatom assemblages with

the increased habitat heterogeneity that occurs as a result of increased moss and macrophyte

growth, which in turn has been attributed to recent warming (e.g. Douglas & Smol 1999). Thus,

one might have expected to find an increase in diatom diversity with high nutrient concentrations

and the presence of relatively diverse substrates (including rocks, sediments, mosses, and

filamentous algae) at Cape Vera. However, only a weak negative relationship (R2 = 0.081, p =

0.177) was found between diatom diversity and seabird-derived nutrients, while no relationship

was found between diatom diversity and primary production (as estimated by chlorophyll a) in

these sites. Previous ecological studies have suggested that plant and algal diversity (or richness)

follows a hump-shaped curve along a productivity gradient (Dodson et al. 2000; Mittelbach et al.

2001; Groner & Novoplansky 2003), but we find no evidence of such a relationship with respect

to diatom species assemblages at Cape Vera, perhaps reflecting dominance by other algal groups.

Species-environment relationships

Three measured environmental variables explained significant portions of the variance in

the diatom data based on the RDA: δ15N, 13.5%, specific conductivity 10.5%, and dissolved

organic carbon (DOC) 10.2% (Figure 7). Only the first axis of the RDA was significant,

however, and the low eigenvalue of this axis suggests that other factors not captured by the RDA

are also influencing diatom species distributions (Axis 1 λ1= 0.168, p = 0.008; Figure 7).

Therefore, in order to explore which environmental factor was most important for explaining the

variation in diatom species distributions in the absence of δ15N, we performed an RDA without

δ15N. In this more restricted analysis, filtered phosphorus (TPf) explained a significant amount of

the variation in the diatom species data (12.4%). Together, these two findings suggest that

nutrients, which are associated with seabirds, play a role in shaping diatom communities at Cape

Vera, but the relationship is not simple and not directly related to nutrients.

To explore which species are most closely associated with the δ15N gradient, we

performed an additional RDA constrained only to δ15N (Table 4). The RDA axis 1 species scores

(a measure of how well the species can be explained by the sole environmental variable) indicate

that N. frustulum, although by far the most dominant taxon in this dataset, has little relationship to

δ15N, the measured environmental variable that explained the most variance in the diatom species

data (Table 4). The high relative abundance of this taxon in the control sites at Cape Hawes and

CV11, which all exhibit low δ15N, underscores the lack of a relationship between δ15N and N.

frustulum. Instead, the relationship between diatom species distributions and δ15N must be

influenced by species associated with low δ15N sites, such as Cymbella botellus, C. designata,

Cymbopleura angustata, and Psammothidium marginulatum (RDA axis 1 species scores > +0.5),

rather than those characteristic of the high δ15N sites (RDA axis 1 score < - 0.5: Encyonema

minutum). Furthermore, although δ15N explains the largest percentage of diatom species

distributions in our dataset, the high dominance of relatively few species at Cape Vera precluded

the generation of a robust transfer function.

That the relationships between the most dominant diatom taxa at Cape Vera are not easily

explainable by patterns of nutrients across such an extreme nutrient gradient further supports

studies suggesting that Arctic diatoms do not respond to nutrient enrichment in a simple fashion,

unlike diatoms from temperate regions (Hall & Smol 1999). While paleolimnological studies of

Arctic lakes receiving large quantities of human sewage have shown distinct diatom responses,

these changes have been very different from those recorded in temperate regions and generally

are indicative of species associated with periphytic habitats rather than high nutrient

concentrations (Douglas & Smol 2000; Michelutti et al. 2002a; Michelutti et al. in press b).

These subtle, and sometimes delayed, diatom responses to eutrophication have been attributed to

the overriding influence of climate (Douglas & Smol 2000; Michelutti et al. in press b). In the

ponds closest to the northern fulmar colony, harsh climatic conditions have not prevented the

development of extremely high standing stocks of mosses, large masses of filamentous algae

(qualitatively identified as Cosmarium sp., Zygnema spp., Ulothrix sp., and Tetraspora sp.) and

cyanobacteria (Nostoc commune and Lyngbya sp.). It is possible that these algae have

outcompeted diatoms in the nutrient-rich ponds at Cape Vera, and therefore precluded any

marked diatom species changes associated with nutrient enrichment. Evidence for differential

nutrient requirements between the phytoplankton and benthos in Arctic lakes is provided by

Bonilla et al. (2005), who found that phytoplankton (including Chrysophyceae and cyanobacteria)

respond much more strongly to nutrient enrichment than do benthic communities, which are

generally not nutrient limited.

Conclusions

Ponds affected by seabird-derived nutrients at Cape Vera are characterized by distinctive

water chemistry (i.e. high pH and nutrients) as compared to ponds less affected by seabird-

nutrients, and compared to lakes and ponds elsewhere in the Canadian Arctic Archipelago.

Surface sediment diatoms were overwhelmingly dominated by very few taxa in almost all ponds,

regardless of seabird influence; nevertheless, a significant portion of the variance in the diatom

species data can be explained by δ15N. Although the generation of robust quantitative models

predicting δ15N from fossil diatom data was stymied by the dominance of a taxon indifferent to

seabird derived nutrients, this study provides evidence that seabird-derived nutrients play a

critical role in shaping the water quality and also, in a more complex manner, the biological

communities of these ponds.

Acknowledgements

We would like to thank NSERC, PCSP, NSTP, CWS for funding and logistical support, and Paul

Hamilton for aid in identifying the non-diatom algae. This fieldwork was completed with the

generous assistance of I. Gregory-Eaves, A. Fontaine, N. North, M. Wayland, N. Michelutti, L.

Kimpe, K. Foster, H. Liu, J. Akearuk, and M. Falconer . This manuscript was greatly improved

by comments from N. Michelutti. This is PCSP/ÉPCP contribution # (TBD).

References

Anderson, W. B., & G. A. Polis, 2004. Allochthonous Nutrient and Food Inputs: Consequences for Temporal Stability. In G. A. Polis, M. E. Power & G. R. Huxel (eds), Food Webs at the Landscape Level. University of Chicago Press, Chicago: 82-95.

Antoniades, D., M. S. V. Douglas, & J. P. Smol, 2003a. Comparative physical and chemical limnology of two Canadian High Arctic regions: Alert (Ellesmere Island, NU) and Mould Bay (Prince Patrick Island, NWT). Archiv für Hydrobiologie 158: 485-516.

Antoniades, D., M. S. V. Douglas, & J. P. Smol, 2003b. The physical and chemical limnology of 24 ponds and one lake from Isachsen, Ellef Ringnes Island, Canadian High Arctic. International Review of Hydrobiology 88: 519-538.

Antoniades, D., M. S. V. Douglas, & J. P. Smol, 2005. Benthic diatom autecology and inference model development from the Canadian High Arctic Archipelago. Journal of Phycology 41: 30-45.

Antoniades, D., P. B. Hamilton, M. S. V. Douglas & J. P. Smol, in press. Freshwater diatoms from the Canadian High Arctic. Iconographica Diatomologica.

Blais, J. M., L. E. Kimpe, D. McMahon, B. E. Keatley, M. L. Mallory, M. S. V. Douglas, & J. P. Smol, 2005. Arctic seabirds transport marine-derived contaminants. Science 309: 445.

Bonilla, S., V. Villeneuve, & W. F. Vincent, 2005. Benthic and planktonic algal communities in a High Arctic Lake: Pigment structure and contrasting responses to nutrient enrichment. Journal of Phycology 41: 1120-1130.

Brooke, M. D., 2004. The food consumption of the world's seabirds. Proceedings of the Royal Society of London Series B-Biological Sciences 271: S246-S248.

Campbell, L. M., A. T. Fisk, X. W. Wang, G. Kock, & D. C. G. Muir, 2005. Evidence for biomagnification of rubidium in freshwater and marine food webs. Canadian Journal of Fisheries and Aquatic Sciences 62: 1161-1167.

Diamond, A. W., & C. M. Devlin, 2003. Seabirds as indicators of changes in marine ecosystems: Ecological monitoring on Machias Seal Island. Environmental Monitoring and Assessment 88: 153-175.

Dodson, S. I., S. E. Arnott, & K. L. Cottingham, 2000. The relationship in lake communities between primary productivity and species richness. Ecology 81: 2662-2679.

Douglas, M. S. V., & J. P. Smol, 1993. Fresh-Water Diatoms from High Arctic Ponds (Cape Herschel, Ellesmere-Island, NWT). Nova Hedwigia 57: 511-552.

Douglas, M. S. V.& J. P. Smol, 1999. Freshwater diatoms as indicators of environmental change in the High Arctic. In E. F. Stoermer & J. P. Smol (eds), The Diatoms: Applications for the Environmental and Earth Sciences. Cambridge University Press, Cambridge: 227-244.

Douglas, M. S. V., & J. P. Smol, 2000. Eutrophication and recovery in the High Arctic: Meretta lake (Cornwallis Island, Nunavut, Canada) revisited. Hydrobiologia 431: 193-204.

Douglas, M. S. V., J. P. Smol, J. M. Savelle, & J. M. Blais, 2004a. Prehistoric Inuit whalers affected arctic freshwater ecosystems. Proceedings of the National Academy of Sciences of the United States of America 101: 1613-1617.

Douglas, M. S. V., P. B. Hamilton, R. Pienitz, & J. P. Smol, 2004b. Algal indicators of environmental change in arctic and antarctic lakes and ponds. In R. Pienitz, M. S. V. Douglas & J. P. Smol (eds), Long-term environmental change in Arctic and Antarctic lakes. Springer, Dordrecht: 117-158.

Ellis, J. C., 2005. Marine birds on land: a review of plant biomass, species richness, and community composition in seabird colonies. Plant Ecology 181: 227-241.

Environment Canada, 1994. Manual of Analytical Methods – National Laboratory for Environmental Testing, Canadian Centre for Inland Waters. Burlington, Canada.

Evenset, A., J. Carroll, G. N. Christensen, R. Kallenborn, D. Gregor, & G. W. Gabrielsen, 2007. Seabird guano is an efficient conveyer of persistent organic pollutants (POPs) to Arctic lake ecosystems. Environmental Science & Technology 41: 1173-1179.

Evenset, A., G. N. Christensen, T. Skotvold, E. Fjeld, M. Schlabach, E. Wartena, & D. Gregor, 2004. A comparison of organic contaminants in two high Arctic lake ecosystems, Bjørnøya (Bear Island), Norway. The Science of the Total Environment 318: 125-141.

Finney, B. P., I. Gregory-Eaves, J. Sweetman, M. S. V. Douglas, & J. P. Smol, 2000. Impacts of climatic change and fishing on Pacific salmon abundance over the past 300 years. Science 290: 795-799.

Gaston, A. J., M. L. Mallory, H. G. Gilchrist, & K. O'Donovan, 2006. Status, trends and attendance patterns of the northern fulmar Fulmarus glacialis in Nunavut, Canada. Arctic 59: 165-178.

Gregory-Eaves, I., J. P. Smol, M. S. V. Douglas, & B. P. Finney, 2003. Diatoms and sockeye salmon (Oncorhynchus nerka) population dynamics: Reconstructions of salmon-derived nutrients over the past 2,200 years in two lakes from Kodiak Island, Alaska. Journal of Paleolimnology 30: 35-53.

Groner, E., & A. Novoplansky, 2003. Reconsidering diversity-productivity relationships: directness of productivity estimates matters. Ecology Letters 6: 695-699.

Hall, R. I., & J. P. Smol, 1999. Diatoms as indicators of lake eutrophication. In E. F. Stoermer & J. P. Smol (eds), The Diatoms: Applications for the Environmental and Earth Sciences. Cambridge University Press, New York: 128-168.

Hamilton, P. B., K. Gajewski, D. E. Atkinson, & D. R. S. Lean, 2001. Physical and chemical limnology of 204 lakes from the Canadian Arctic Archipelago. Hydrobiologia 457: 133-148.

Harding, J. S., D. J. Hawke, R. N. Holdaway, & M. J. Winterbourn, 2004. Incorporation of marine-derived nutrients from petrel breeding colonies into stream food webs. Freshwater Biology 49: 576-586.

Hatch, S. A.& D. N. Nettleship, 1998. Northern fulmar (Fulmarus glacialis). In A. Poole & F. Gill (eds), The Birds of North America, No. 361. The Birds of North America, Inc., Philadelphia, PA: 1-32.

Hill, M. O., 1973. Diversity and evenness: a unifying notation and its consequences. Ecology 54: 427-432.

Izaguirre, I., A. Vinocur, G. Mataloni, & M. Pose, 1998. Phytoplankton communities in relation to trophic status in lakes from Hope Bay (Antarctic Peninsula). Hydrobiologia 370: 73-87.

Juggins, S., 2003. C2: software for ecological and palaeocological data analysis and visualisation, version 1.3. Department of Geography, University of Newcastle.

Keatley, B. E., M. S. V. Douglas, & J. P. Smol, [2]. Physical and chemical limnological characteristics of 46 lakes and ponds across environmental gradients on Melville Island, High Arctic Canada. Fundamental and Applied Limnology 168: 355-376.

Keatley, B. E., M. S. V. Douglas, & J. P. Smol, [4]. Limnological characteristics of a high Arctic oasis and comparisons across northern Ellesmere Island. Arctic.

Kitchell, J. F., D. E. Schindler, B. R. Herwig, D. M. Post, M. H. Olson, & M. Oldham, 1999. Nutrient cycling at the landscape scale: The role of diel foraging migrations by geese at the Bosque del Apache National Wildlife Refuge, New Mexico. Limnology and Oceanography 44: 828-836.

Krammer, K., 2002. Diatoms of Europe: Cymbella. A.R.G. Gantner Verlag, Ruggell.

Krammer, K.& H. Lange-Bertalot, 1991. Bacillariophyceae. In H. Ettl, J. Gerloff, H. Heynig & D. Mollenhauer (eds). Gustav Fischer Verlag, Stuttgart.

Lepš, J.& P. Šmilauer, 2003. Multivariate analysis of ecological data using CANOCO. Cambridge University Press, London.

Lim, D.S.S., 2004. Limnology and diatom paleoecology of diatoms on Banks Island, N.W.T., and Devon Island, Nunavut, Canadian Arctic. University of Toronto. PhD Thesis.

Lim, D. S. S., M. S. V. Douglas, J. P. Smol, & D. R. S. Lean, 2001. Physical and chemical limnological characteristics of 38 lakes and ponds on Bathurst Island, Nunavut, Canadian High Arctic. International Review of Hydrobiology 86: 1-22.

Lim, D. S. S., & M. S. V. Douglas, 2003. Limnological characteristics of 22 lakes and ponds in the Haughton Crater region of Devon Island, Nunavut, Canadian High Arctic. Arctic Antarctic and Alpine Research 35: 509-519.

Lim, D. S. S., M. S. V. Douglas, & J. P. Smol, 2005. Limnology of 46 lakes and ponds on Banks Island, NWT, Canadian Arctic Archipelago. Hydrobiologia 545: 11-32.

Liu, X. D., S. P. Zhao, L. G. Sun, X. B. Yin, Z. Q. Xie, L. Honghao, & Y. H. Wang, 2006. P and trace metal contents in biomaterials, soils, sediments and plants in colony of red-footed booby (Sula sula) in the Dongdao Island of South China Sea. Chemosphere 65: 707-715.

Loder, I. T. C., B. Ganning, & J. A. Love, 1996. Ammonia nitrogen dynamics in coastal rockpools affected by gull guano. Journal of Experimental Marine Biology and Ecology 196: 113-129.

Mallory, M. L., 2006. The Northern Fulmar (Fulmarus glacialis) in Arctic Canada: ecology, threats, and what it tells us about marine environmental conditions. Environmental Reviews 14: 187-216.

Mallory, M. L., & A. J. Fontaine, 2004. Key marine habitat sites for migratory birds in Nunavut and the Northwest Territories. Canadian Wildlife Service, Ottawa. Occasional Paper 109.

Manny, B. A., W. C. Johnson, & R. G. Wetzel, 1994. Nutrient Additions by Waterfowl to Lakes and Reservoirs - Predicting Their Effects on Productivity and Water-Quality. Hydrobiologia 280: 121-132.

Mayr, U., T. de Freitas, G.H. Eisbacher, & J.Wm. Kerr, 1998. Geology, Prince Alfred Bay-Cardigan Strait and parts of adjacent sheets, Northwest Territories. Geological Survey of Canada, Ottawa. Map 1924a; Scale 1:250 000.

Meteorological Service of Canada, 2007: Monthly Climate Data. Environment Canada http://www.climate.weatheroffice.ec.gc.ca/climateData/monthlydata_e.html, Accessed February 2007

Michelutti, N., M. S. V. Douglas, & J. P. Smol, 2002a. Tracking recent recovery from eutrophication in a high arctic lake (Meretta Lake, Cornwallis Island, Nunavut, Canada) using fossil diatom assemblages. Journal of Paleolimnology 28: 377-381.

Michelutti, N., M. S. V. Douglas, D. C. G. Muir, X. Wang, & J. P. Smol, 2002b. Limnological characteristics of 38 lakes and ponds on Axel Heiberg Island, High Arctic Canada. International Review of Hydrobiology 87: 385-399.

Michelutti, N., M. S. V. Douglas, D. R. S. Lean, & J. P. Smol, 2002c. Physical and chemical limnology of 34 ultra-oligotrophic lakes and ponds near Wynniatt Bay, Victoria Island, Arctic Canada. Hydrobiologia 482: 1-13.

Michelutti, N., M. S. V. Douglas, & J. P. Smol, 2003. Tracking recovery in a eutrophied high Arctic lake (Meretta Lake, Cornwallis Island, Canadian Arctic) using periphytic diatoms. Verh. int. Ver. Limnol. 28: 1533-1537.

Michelutti, N., M. S. V. Douglas, & J. P. Smol, in press a. Evaluating diatom community composition in the absence of marked limnological gradients in the high Arctic: A surface sediment calibration set from Cornwallis Island (Nunavut, Canada). Polar Biology.

Michelutti, N., M. H. Hermanson, J. P. Smol, P. J. Dillon, & M. S. V. Douglas, in press b. Delayed response of diatom assemblage changes to sewage inputs in an Arctic lake. Aquatic Sciences.

Mittelbach, G. G., C. F. Steiner, S. M. Scheiner, K. L. Gross, H. L. Reynolds, R. B. Waide, M. R. Willig, S. I. Dodson, & L. Gough, 2001. What is the observed relationship between species richness and productivity? Ecology 82: 2381-2396.

Parr, J. F., K. H. Taffs, & C. M. Lane, 2004. A microwave digestion technique for the extraction of fossil diatoms from coastal lake and swamp sediments. Journal of Paleolimnology 31: 383-390.

Payne, L. X., & J. W. Moore, 2006. Mobile scavengers create hotspots of freshwater productivity. Oikos 115: 69-80.

Polis, G. A., W. B. Anderson, & R. D. Holt, 1997. Toward an integration of landscape and food web ecology: the dynamics of spatially subsidized food webs. Annual Reviews of Ecological Systems 28: 289-316.

Polis, G. A., & S. D. Hurd, 1996. Linking marine and terrestrial food webs: allochthonous input from the ocean supports high secondary productivity on small islands and coastal land communities. The American Naturalist 147: 396-423.

Polis, G. A., M. E. Power & G. R. Huxel, 2004. Food Webs at the Landscape Level. University of Chicago Press, Chicago.

Robinson, D., 2001. δ15N as an integrator of the nitrogen cycle. Trends in Ecology and Evolution 16: 153-162.

Sánchez-Piñero, F., & G. A. Polis, 2000. Bottom-up dynamics of allochthonous input: direct and indirect effects of seabirds on islands. Ecology 81: 3117-3132.

SPSS, 1999. SYSTAT 9. SPSS Inc., Chicago, IL, USA.

ter Braak, C. J. F.& P. Šmilauer, 2002. CANOCO reference manual and CANOdraw for Windows user guide: Software for Canonical Community Ordination (Version 4.5).

Tscherko, D., M. Bolter, L. Beyer, J. Chen, J. Elster, E. Kandeler, D. Kuhn, & H. P. Blume, 2003. Biomass and enzyme activity of two soil transects at King George Island, maritime Antarctica. Arctic Antarctic and Alpine Research 35: 34-47.

Vanni, M. J., D. L. DeAngelis, D. E. Schindler & G. R. Huxel, 2004. Overview: Cross-Habitat Flux of Nutrients and Detritus. In G. A. Polis, M. E. Power & G. R. Huxel (eds), Food Webs at the Landscape Level. University of Chicago Press, Chicago: 3-11.

Vincent, W. F.& J. E. Hobbie, 2000. Ecology of Arctic lakes and rivers. In M. Nuttall & T. V. Callaghan (eds), The Arctic: Environment, People, Policy. Harwood Academic Publishers, Amsterdam: 197-232.

Wait, D. A., D. P. Aubrey, & W. B. Anderson, 2005. Seabird guano influences on desert islands: soil chemistry and herbaceous species richness and productivity. Journal of Arid Environments 60: 681-695.

Figure captions.

Figure 1. Map of study location at Cape Vera, Devon Island. Insets showing location of: a) Devon Island within Canada; b) Cape Vera on Devon Island; and c) sites located throughout Cape Vera. The isoclines are not drawn to scale. The gradient indicates the approximate concentration of northern fulmars along the cliffs at Cape Vera, with the most birds occurring within the southern third of the colony. Figure 2. Principal Components Analysis (PCA) biplot of measured environmental variables and Cape Vera sites. The “control sites” (CV11, 16, 17, 18, 22, 23, 24) are denoted with open circles while the rest of the ponds are denoted with solid circles. Figure 3. Box plots comparing selected measured environmental variables (TPu, TPf, TN, DOC, pH, chla) between Cape Vera ponds (n = 24, this study) and ponds located in the nearby Haughton Crater, Devon Island (n = 22, Lim & Douglas 2003). Figure 4. Composite PCA biplot of selected measured environmental variables common to our labs’ previous limnological surveys in the Canadian Arctic Archipelago. References for each study are as follows: Melville Island (Keatley et al. [2]), Mould Bay, Prince Patrick Island (Antoniades et al. 2003a), Banks Island (Lim et al. 2005), Victoria Island (Michelutti et al. 2002a), Isachsen, Ellef Ringnes Island (Antoniades et al. 2003b), Bathurst Island (Lim et al. 2001), Cornwallis Island (Michelutti et al. in press a), Devon Island (Lim & Douglas 2003), Axel Heiberg Island (Michelutti et al. 2002b), northern Ellesmere Island (Keatley et al. [4]), Alert, Ellesmere Island (Antoniades et al. 2003). Figure 5. Histogram of dominant diatom species (>1% relative abundance in at least 5 sites) found in ponds from Cape Vera. Both the sites and the species are ordered according to their DCA axis 1 scores. Figure 6. Principal components analysis (PCA) biplot of species and sites from the surface sediments of Cape Vera, Devon Island. Figure 7. Redundancy analysis (RDA) biplot constrained to the three measured environmental variables that explained significant portions of the diatom species variance (δ15N, specific conductivity, and dissolved organic carbon (DOC)). Biplot 7a) presents the diatom species, and 7b) the sites in relation to the environmental variables.

b c

a

Figure 1.

Figure 2.

0

10

20

30

40

50

60

70

0.00

0.05

0.10

0.15

0.20

0.25

Cape Vera (n = 24)

Devon Crater (n = 22)

DO

C (m

g/L)

0

2

4

6

8

10

12

14

Cape Vera (n = 24)

Devon Crater (n = 22)

TNu

(mg/

L)

0.0

0.5

1.0

1.5

2.0

2.5

3.0

TPu

(mg/

L)

Chl

a(µ

g/L)

0

10

20

30

40

50

60

70

0.00

0.05

0.10

0.15

0.20

0.25

Cape Vera (n = 24)

Devon Crater (n = 22)

DO

C (m

g/L)

0

2

4

6

8

10

12

14

Cape Vera (n = 24)

Devon Crater (n = 22)

TNu

(mg/

L)

0.0

0.5

1.0

1.5

2.0

2.5

3.0

TPu

(mg/

L)

Chl

a(µ

g/L)

Figure 3.

-0.4 1.0

-0.6

1.0

cond

pH

TPu

TPf

TNu

DOC

Axis 1 λ=0.379

Axi

s 2 λ=

0.18

0

Mould Bay

N. EllesmereOasis Ellesmere

Cape VeraMelville

BanksVictoriaBathurstIsachsenDevonAxelAlert

Cornwallis

-0.4 1.0

-0.6

1.0

cond

pH

TPu

TPf

TNu

DOC

Axis 1 λ=0.379

Axi

s 2 λ=

0.18

0

-0.4 1.0

-0.6

1.0

cond

pH

TPu

TPf

TNu

DOC

Axis 1 λ=0.379

Axi

s 2 λ=

0.18

0

Mould Bay

N. EllesmereOasis Ellesmere

Cape VeraMelville

BanksVictoriaBathurstIsachsenDevonAxelAlert

Cornwallis

Mould Bay

N. EllesmereOasis Ellesmere

Cape VeraMelville

BanksVictoriaBathurstIsachsenDevonAxelAlert

Cornwallis

Cape VeraMelville

BanksVictoriaBathurstIsachsenDevonAxelAlert

Cornwallis

Figure 4.

5-05

7-04

9-04

4-04

31-0

617

-05

3-04

13-0

68-

0410

-04

18-0

59a

-56-

0420

-05

14-0

516

-05

2-04

1-04

12-0

515

-05

24-0

522

-05

23-0

511

-04

020

Microc

ostat

uskra

sske

i

0Nitzsc

hiapu

ra

020

Encyo

nemaminu

tum

0Navicu

lach

iarae

020

Encyo

nemafog

edii

0Encyo

nemasil

esiac

um

0Fragila

riaca

pucin

a

020

4060

8010

0

Nitzsc

hiape

rminu

ta

020

4060

80

Nitzsc

hiafru

stulum

020

Cymbe

llacle

ve-eu

lerae

0Navicu

lavu

lpina

020

40

Diatom

atenuis

0Psammoth

idium

marginu

latum

0Achna

nthidi

umminu

tissim

um

020

Cymbe

llabo

tellus

020

Cymbo

pleura

angu

stata

Pond name

Rel

ativ

e ab

unda

nce

(%)

5-05

7-04

9-04

4-04

31-0

617

-05

3-04

13-0

68-

0410

-04

18-0

59a

-56-

0420

-05

14-0

516

-05

2-04

1-04

12-0

515

-05

24-0

522

-05

23-0

511

-04

020

Microc

ostat

uskra

sske

i

0Nitzsc

hiapu

ra

020

Encyo

nemaminu

tum

0Navicu

lach

iarae

020

Encyo

nemafog

edii

0Encyo

nemasil

esiac

um

0Fragila

riaca

pucin

a

020

4060

8010

0

Nitzsc

hiape

rminu

ta

020

4060

80

Nitzsc

hiafru

stulum

020

Cymbe

llacle

ve-eu

lerae

0Navicu

lavu

lpina

020

40

Diatom

atenuis

0Psammoth

idium

marginu

latum

0Achna

nthidi

umminu

tissim

um

020

Cymbe

llabo

tellus

020

Cymbo

pleura

angu

stata

Pond name

Rel

ativ

e ab

unda

nce

(%)

5-05

7-04

9-04

4-04

31-0

617

-05

3-04

13-0

68-

0410

-04

18-0

59a

-56-

0420

-05

14-0

516

-05

2-04

1-04

12-0

515

-05

24-0

522

-05

23-0

511

-04

020

Microc

ostat

uskra

sske

i

0Nitzsc

hiapu

ra

020

Encyo

nemaminu

tum

0Navicu

lach

iarae

020

Encyo

nemafog

edii

0Encyo

nemasil

esiac

um

0Fragila

riaca

pucin

a

020

4060

8010

0

Nitzsc

hiape

rminu

ta

020

4060

80

Nitzsc

hiafru

stulum

020

Cymbe

llacle

ve-eu

lerae

0Navicu

lavu

lpina

020

40

Diatom

atenuis

0Psammoth

idium

marginu

latum

0Achna

nthidi

umminu

tissim

um

020

Cymbe

llabo

tellus

020

Cymbo

pleura

angu

stata

Pond name

Rel

ativ

e ab

unda

nce

(%)

Figu

re 5

.

-1.5 1.0

-1.0

1.0

P.margA.minu

C.anguC.desi

E.foge

E.late

E.micr

E.minu

C.sile

D.kuet

D.cont

D.tenu

D.marg

F.capu

M.kras

N.chia

N.phyl

N.vulp

N.diss

N.frust

N.perm

N.pura

1

2

34

5

6

78

99a

10

11

12

14

15

16

17

18

20

2223

24

1331

Axis 1 λ = 0.278

Axi

s 2 λ

= 0.

172

R.pete

A.krie

C.clevE.laev

N.pale

-1.5 1.0

-1.0

1.0

P.margA.minu

C.anguC.desi

E.foge

E.late

E.micr

E.minu

C.sile

D.kuet

D.cont

D.tenu

D.marg

F.capu

M.kras

N.chia

N.phyl

N.vulp

N.diss

N.frust

N.perm

N.pura

1

2

34

5

6

78

99a

10

11

12

14

15

16

17

18

20

2223

24

1331

Axis 1 λ = 0.278

Axi

s 2 λ

= 0.

172

R.pete

A.krie

C.clevE.laev

N.pale

-1.5 1.0

-1.0

1.0

P.margA.minu

C.anguC.desi

E.foge

E.late

E.micr

E.minu

C.sile

D.kuet

D.cont

D.tenu

D.marg

F.capu

M.kras

N.chia

N.phyl

N.vulp

N.diss

N.frust

N.perm

N.pura

1

2

34

5

6

78

99a

10

11

12

14

15

16

17

18

20

2223

24

1331

Axis 1 λ = 0.278

Axi

s 2 λ

= 0.

172

R.pete

A.krie

C.clevE.laev

N.pale

Figure 6.

b)

-1.0 1.0

-1.0

1.0

δ15N

cond

DOC

1

2

3

4

5

6

7

8

9

9a

10

11

12

14

15

16

17

18

2022

2324

13

31

Axis 1 λ = 0.132

Axis

2 λ

= 0.

051

a)

-1.0 1.0

-1.0

1.0

A.krie

E.laev

A.marg

A.minu

E.late

D.kuet

N.perm

R.pete

C.angu

C.bote

C.clev

E.foge

E.micr

E.minu

E.sile

D.cont

D.tenu

D.margF.capu

M.kras

N.chia

N.cryp

N.phyl

N.vulp

N.dissN.frus

N.pale

N.pura

C.desi

Axis 1 λ = 0.132

Axis

2 λ

= 0.

051

b)

-1.0 1.0

-1.0

1.0

δ15N

cond

DOC

1

2

3

4

5

6

7

8

9

9a

10

11

12

14

15

16

17

18

2022

2324

13

31

b)

-1.0 1.0

-1.0

1.0

δ15N

cond

DOC

1

2

3

4

5

6

7

8

9

9a

10

11

12

14

15

16

17

18

2022

2324

13

31

Axis 1 λ = 0.132

Axis

2 λ

= 0.

051

a)

-1.0 1.0

-1.0

1.0

A.krie

E.laev

A.marg

A.minu

E.late

D.kuet

N.perm

R.pete

C.angu

C.bote

C.clev

E.foge

E.micr

E.minu

E.sile

D.cont

D.tenu

D.margF.capu

M.kras

N.chia

N.cryp

N.phyl

N.vulp

N.dissN.frus

N.pale

N.pura

C.desi

a)

-1.0 1.0

-1.0

1.0

A.krie

E.laev

A.marg

A.minu

E.late

D.kuet

N.perm

R.pete

C.angu

C.bote

C.clev

E.foge

E.micr

E.minu

E.sile

D.cont

D.tenu

D.margF.capu

M.kras

N.chia

N.cryp

N.phyl

N.vulp

N.dissN.frus

N.pale

N.pura

C.desi

Axis 1 λ = 0.132

Axis

2 λ

= 0.

051

Figure 7.

Tabl

e 1.

Sel

ecte

d lim

nolo

gica

l dat

a fo

r pon

ds lo

cate

d ne

ar C

ape

Ver

a, D

evon

Isla

nd.

Si

teSa

mpl

eSa

mpl

eδ1

5N%

CpH

Con

dN

O3N

O2

N

H3

CH

La

Cl

SO4

DO

CD

ICC

aM

g K

Na

Dat

eco

de(‰

)%

(μS/

cm)

mg/

Lm

g/L

µg/L

mg/

Lm

g/L

mg/

Lm

g/L

mg/

Lm

g/L

mg/

Lm

g/L

CV

-126

/07/

2004

1-04

6.62

17.1

68.

4215

20.

056

0.01

25.

56.

381.

941

19.5

209.

050.

333.

28C

V-2

26/0

7/20

042-

044.

363.

778.

2814

00.

034

0.00

99.

74.

60.

740.

917

.818

.27.

80.

272.

43C

V-3

26/0

7/20

043-

0418

.41

1.45

9.63

240

1.12

0.01

645

.632

.211

.43.

918

.820

.611

.92.

4221

CV

-426

/07/

2004

4-04

14.3

80.

549.

9521

50.

280.

026

10.9

27.4

6.45

7.1

1516

.511

2.25

16.7

CV

-502

/07/

2005

5-05

9.36

10.6

46.

9511

60.

021

0.00

60.

92.

840.

60.

914

.614

.75.

050.

181.

78C

V-6

03/0

7/20

056-

059.

1315

.44

7.5

146.

50.

025

0.01

0.5

3.73

0.64

1.6

19.5

196.

260.

292.

37C

V-7

09/0

7/20

057-

0510

.58

0.50

7.95

134.

90.

020.

009

10.8

1.69

0.37

1.4

17.1

20.7

4.66

0.26

1.24

CV

-826

/07/

2004

8-04

16.3

426

.50

10.7

521

00.

011

0.01

358

.624

.32.

5712

.617

.127

.110

.31.

3413

.2C

V-9

09

/07/

2005

9-05

15.9

520

.57

8.95

121

0.07

90.

011

13.

090.

621.

613

.215

.75.

220.

31.

99C

V-9

a09

/07/

2005

9a-0

513

.02

42.1

49.

7512

6.5

0.06

30.

025

1.9

5.89

1.07

3.9

11.5

146.

20.

443.

44C

V-1

026

/07/

2004

10-0

413

.26

19.3

89.

4915

30.

048

<0.0

0535

.711

.71.

071.

215

.121

.77.

540.

46.

43C

V-1

126

/07/

2004

11-0

40.

748.

438.

7014

20.

017

<0.0

0515

.46.

990.

690.

917

.819

7.03

0.2

2.99

CV

-12

02/0

7/20

0512

-05

3.30

16.5

37.

6531

2.5

0.01

20.

035

246

.98.

734.

121

.617

.611

1.87

29.4

CV

-14

03/0

7/20

0514

-05

8.15

32.4

37.

116

9.2

0.00

70.

006

1.5

9.69

1.45

3.6

19.6

17.3

7.48

0.64

5.32

CV

-15

03/0

7/20

0515

-05

16.3

421

.84

7.55

167.

40.

016

0.01

60.

94.

951.

12.

421

.320

.77.

290.

373.

01C

V-1

604

/07/

2005

16-0

55.

9921

.93

7.65

125.

80.

045

0.00

80.

94.

41.

53.

016

.014

.87.

460.

252.

36C

V-1

704

/07/

2005

17-0

56.

2327

.15

7.35

84.7

0.02

10.

016

0.6

3.96

0.62

1.4

10.3

8.78

4.45

0.16

2.08

CV

-18

04/0

7/20

0518

-05

6.46

32.3

77.

8514

80.

019

0.00

50.

510

.91.

123.

016

.014

.67.

660.

254.

69C

V-2

009

/07/

2005

20-0

57.

6035

.73

10.0

517

30.

020.

018

0.9

14.8

1.69

6.2

12.5

17.8

8.09

0.84

7.94

CV

-22

10/0

7/20

0522

-05

1.34

16.4

17.

5512

5.4

0.01

40.

026

0.9

15.2

1.07

1.8

10.9

10.7

5.19

0.35

7.25

CV

-23

10/0

7/20

0523

-05

1.61

34.5

67.

213

6.5

0.01

70.

013

0.5

20.5

1.19

1.5

10.1

9.99

5.3

0.41

9.73

CV

-24

10/0

7/20

0524

-05

1.62

35.7

58

127.

30.

022

0.00

90.

65.

171.

141.

515

.714

.37

0.35

2.93

CV

-13

10/0

7/20

0613

-06

19.3

825

.68

10.0

023

0.6

0.14

70.

050.

513

.84.

495.

718

.320

.710

.10.

987.

28C

V-3

118

/07/

2006

31-0

611

.95

5.78

8.60

180

0.01

10.

029

0.8

5.15

0.98

1.6

18.1

21.6

5.9

0.28

2.9

Mea

n9.

2619

.70

8.45

161.

60.

089

0.01

78.

6311

.93

2.22

3.0

16.1

17.3

7.46

0.64

6.74

Med

ian

8.64

19.9

88.

1414

7.3

0.02

10.

013

0.95

6.69

1.11

1.7

16.6

17.7

7.38

0.35

3.36

Max

19.3

842

.14

10.7

531

2.5

1.12

00.

050

58.6

046

.90

11.4

012

.621

.627

.111

.90

2.42

29.4

0M

in0.

740.

506.

9584

.70.

007

<0.0

050.

501.

690.

370.

910

.18.

84.

450.

161.

24

Tabl

e 1.

con

tinue

d.

Si

teN

O2

POC

PON

SiO

2SR

P T

KN

T

dN

TPf

TPu

H

ill's

N:P

N o

r P

mg/

Lm

g/L

mg/

Lm

g/L

mg/

Lm

g/L

mg/

Lm

g/L

mg/

LN

2m

olar

limite

d?C

V-1

0.00

40.

197

0.02

0.3

0.00

110.

129

0.20

30.

0052

0.00

564.

478

.4p

C

V-2

0.00

20.

159

0.02

10.

170.

001

0.08

10.

132

0.00

230.

004

11.5

72.9

pC

V-3

0.02

84.

860.

863

0.91

0.00

150.

589

1.59

0.01

390.

19.

3555

.1p

CV

-40.

012

0.89

20.

136

2.64

0.00

170.

856

0.99

50.

0227

0.04

718.

2757

.9p

CV

-50.

003

0.34

20.

013

0.25

0.00

10.

093

0.11

40.

0057

0.00

948.

4729

.0ei

ther

CV

-60.

004

0.28

30.

017

0.31

0.00

070.

131

0.17

60.

0041

0.00

568.

2866

.2p

CV

-70.

003

0.74

20.

051

0.43

0.00

080.

184

0.21

10.

0052

0.02

698.

1220

.3ei

ther

CV

-80.

002

9.27

0.99

1.09

0.00

290.

937

0.82

30.

035

0.20

72.

6520

.1ei

ther

CV

-9

0.00

50.

323

0.02

40.

420.

0226

0.19

20.

293

0.02

860.

0353

6.37

17.9

n

CV

-9a

0.00

85.

620.

666

0.36

0.05

550.

404

0.44

80.

072

0.12

42.

7119

.6n

CV

-10

0.00

54.

80.

694

0.51

0.00

090.

243

0.30

10.

0117

0.01

833.

0811

5.3

pC

V-1

10.

001

1.33

0.23

40.

260.

0008

0.06

10.

072

0.00

220.

0169

13.3

739

.6ei

ther

CV

-12

0.00

40.

888

0.1

2.49

0.00

130.

448

0.43

10.

0134

0.02

615.

9146

.0ei

ther

CV

-14

0.00

20.

777

0.08

70.

090.

013

0.27

30.

286

0.01

640.

0318

6.37

24.7

eith

erC

V-1

50.

003

0.40

70.

033

0.51

0.00

070.

220.

240.

0042

0.01

047.

8455

.4p

CV

-16

0.00

30.

16<0

.005

0.37

0.00

280.

097

0.11

90.

0049

0.01

126.

6427

.4ei

ther

CV

-17

0.00

40.

50.

060.

140.

0011

0.15

40.

153

0.00

580.

0143

3.16

35.2

eith

erC

V-1

80.

004

0.23

0.00

50.

190.

001

0.13

90.

141

0.00

850.

0104

133

.6ei

ther

CV

-20

0.00

50.

479

0.04

90.

810.

009

0.65

70.

628

0.02

890.

0371

3.73

41.9

eith

erC

V-2

20.

004

1.21

0.11

70.

310.

0005

0.26

50.

261

0.00

290.

0359

11.7

723

.6ei

ther

CV

-23

0.00

30.

333

0.02

50.

210.

0003

0.21

30.

210.

0028

0.00

8313

.18

65.8

pC

V-2

40.

003

0.27

0.01

30.

250.

0003

0.14

70.

162

0.01

070.

0096

7.83

40.6

eith

erC

V-1

30.

005

0.32

0.04

52.

610.

0113

0.71

30.

909

0.03

530.

0488

5.92

39.7

eith

erC

V-3

10.

002

0.26

40.

027

0.33

0.00

830.

203

0.20

80.

0031

0.00

8414

.97

61.5

pM

ean

0.00

51.

444

0.18

70.

670.

0058

0.31

00.

379

0.01

440.

0355

7.29

45.3

2ei

ther

Med

ian

0.00

40.

443

0.04

90.

350.

0011

0.20

80.

226

0.00

720.

0176

7.24

40.1

8ei

ther

Max

0.02

89.

270

0.99

02.

640.

0555

0.93

71.

590

0.07

200.

2070

14.9

711

5.34

pM

in0.

001

0.15

9<0

.005

0.09

0.00

030.

061

0.07

20.

0022

0.00

401.

0017

.91

n

Tabl

e 2.

Pea

rson

’s c

orre

latio

n m

atrix

with

Bon

ferr

oni-a

djus

ted

prob

abili

ties.

Sig

nific

ant v

alue

s are

den

oted

in b

old

(p<0

.01)

or i

n un

derli

ned

italic

s (p<

0.05

).

δ15

N%

CpH

CO

ND

NO

3NO

2N

H3

CH

La

DO

CN

O2

PON

SiO

2SR

PT

KN

Cd

FePb

TdN

TPf

Tpu

δ15N

1%

C-0

.169

1pH

0.53

7-0

.076

1C

ON

D0.

328

-0.2

320.

446

1N

O3N

O2

0.49

2-0

.384

0.50

30.

251

1N

H3

0.26

3-0

.003

0.24

0.40

80.

215

1C

HL

a0.

243

-0.5

240.

538

0.34

0.33

9-0

.279

1D

OC

0.48

10.

137

0.63

70.

658

0.20

60.

567

0.19

31

NO

20.

485

-0.1

760.

384

0.28

10.

824

0.39

50.

151

0.38

11

PON

0.28

4-0

.133

0.58

20.

358

0.22

40.

096

0.67

20.

491

0.3

1Si

O2

0.45

-0.2

440.

628

0.76

10.

461

0.55

50.

282

0.68

30.

468

0.33

81

SRP

0.52

90.

241

0.47

20.

145

0.16

50.

322

-0.0

830.

453

0.17

0.18

20.

173

1T

KN

0.53

10.

061

0.67

0.66

20.

332

0.62

20.

230.

937

0.55

30.

572

0.73

80.

396

1C

d0.

579

-0.0

360.

502

0.38

10.

308

0.40

80.

180.

635

0.35

90.

418

0.49

60.

695

0.61

41

Fe0.

351

-0.2

240.

211

0.29

0.23

0.53

70.

016

0.47

40.

450.

376

0.45

30.

312

0.59

70.

731

1Pb

0.33

5-0

.402

0.41

10.

352

0.24

80.

205

0.42

50.

141

0.09

20.

489

0.42

70.

151

0.27

40.

477

0.49

81

TdN

0.61

8-0

.088

0.71

0.68

70.

563

0.58

50.

340.

864

0.70

20.

586

0.74

50.

392

0.95

30.

639

0.57

80.

334

1T

Pf0.

575

0.33

10.

669

0.35

10.

356

0.32

50.

121

0.75

0.49

40.

418

0.51

0.69

20.

742

0.74

10.

453

0.18

70.

722

1T

pu0.

461

0.05

20.

675

0.39

0.32

20.

334

0.40

60.

793

0.43

70.

715

0.52

70.

516

0.79

10.

747

0.58

40.

332

0.77

0.76

81

Table 3. Summary of dominant diatom species (at least 1% relative abundance in at least 3 sites) found in surface sediments of Cape Vera ponds, where “Code” is the shortened species name, “N” is the number of occurrences, “N2” is the effective number of occurrences (Hill 1973), and “Max %” is the maximum relative abundance of a given taxon.

Code N N2 Max %Achnanthidium kriegeri (Krasske) Hamilton, Antoniades & Siver A.krie 3 4.57 2.21Eucocooneis laevis (Østrup) Lange-Bertalot E.laev 3 2.88 3.42Psammothidium marginulatum (Grunow) Bukhtiyarova & Round P.marg 6 7.69 4.32Achnanthidium minutissimum (Kützing) Czarnecki A.minu 9 9.38 8.97Rossithidium petersenii (Hustedt) Round & Bukhtiyarova R.pete 3 6.22 4.32Cymbopleura angustata var spitzbergensis Krammer C.angu 5 4.37 17.32Cymbella botellus (Lagerstedt) Schmidt C.bote 8 7.95 16.30Cymbella cleve-eulerae Krammer C.clev 16 13.94 19.28Cymbella designata Krammer C.desi 4 5.43 2.18Encyonema fogedii Krammer E.foge 19 15.89 24.87Encyonema latens (Krasske) D. Mann E.late 4 7.59 2.56Encyonopsis microcephala (Grunow) Krammer E.micr 4 3.97 15.72Encyonema minutum (Hilse) D. Mann E.minu 14 11.73 26.74Encyonema silesiacum (Bleisch) D. Mann E.sile 5 12.35 3.30Denticula kuetzingii Grunow D.kuet 4 3.43 23.99Diadesmis contenta (Grunow ex. Van Huerck) D. Mann D.cont 3 2.48 14.50Diatoma tenuis Agardh D.tenu 8 7.17 43.11Diploneis marginestriata Hustedt D.marg 2 1.42 26.06Fragilaria capucina Desmazières F.capu 6 7.71 8.92Microcostatus krasskei (Hustedt) Johansen & Sray M.kras 7 6.74 19.44Navicula chiarae Lange-Bertalot & Genkal N.chia 4 4.71 3.55Navicula cryptocephala Kützing N.cryp 3 4.48 2.03Navicula phyllepta Kützing N.phyl 4 7.5 4.5Navicula vulpina Kützing N.vulp 7 9.81 4.30Nitzschia dissipata var media (Hantzsch) Grunow N.diss 3 5.97 3.05Nitzschia frustulum (Kützing) Grunow N.frus 23 19.99 87.45Nitzschia perminuta (Grunow) M. Peragallo N.perm 24 20.58 99.82Nitzschia paleacea Grunow in Van Heurck N.pale 3 3.7 3.1Nitzschia pura Hustedt N.pura 4 5 7.

9

825

Table 4. List of dominant diatom species arranged in order of their axis 1 species scores based on a redundancy analysis (RDA) constrained solely to δ15N.

RDA δ15N onlyaxis 1 scores

Cymbella botellus 0.6333Cymbella designata 0.628Cymbopleura angustata var spitzbergensis 0.6078Psammothidium marginulatam 0.5572Encyonopsis microcephala 0.4765Eucocooneis laevis 0.467Achnanthidium minutissimum 0.4564Nitzschia paleacea 0.3705Achnanthidium kriegeri 0.3174Rossithidium petersenii 0.2998Denticula kuetzigii 0.2864Diatoma tenuis 0.2562Cymbella cleve-eulerae 0.2311Diploneis marginestriata 0.1465Navicula vulpina 0.1334Nitzschia dissipata var media 0.1054Navicula chiarae 0.0788Navicula cryptocephala 0.0698Diadesmis contenta 0.0182Fragilaria capucina -0.0447Encyonema latens -0.1294Nitzschia perminuta -0.1305Navicula phyllepta -0.1342Encyonema silesiacum -0.2058Encyonema fogedii -0.2169Nitzschia frustulum -0.2169Nitzschia pura -0.2671Microcostatus krasskei -0.2803Encyonema minutum -0.6591

CHAPTER 8

TRACKING SEABIRD POPULATION DYNAMICS USING PALEOLIMNOLOGY: A CASE STUDY FROM

DEVON ISLAND, ARCTIC CANADA

BRONWYN KEATLEY, MARIANNE S.V. DOUGLAS, NEAL MICHELUTTI, JULES BLAIS, MARK

MALLORY, AND JOHN P. SMOL

ABSTRACT

Rapid environmental changes occurring in polar regions may pose a significant threat to the ~10

million seabirds currently inhabiting the Canadian Arctic. However, a paucity of long-term data

on seabird population dynamics makes proper management difficult. As top predators in the

marine environment, seabirds provide marine-derived nutrients (enriched in δ15N) to freshwater

habitats via their guano, causing trophic cascades that can be tracked using paleolimnological

techniques. Here, we analyze multiple proxies (δ15N, inferred-chlorophyll a, and diatom

assemblage composition) archived in the sediments of freshwater ponds to reconstruct population

dynamics of a colony of northern fulmars (Fulmarus glacialis) over the past few centuries at

Cape Vera, Devon Island, Nunavut, High Arctic Canada. Sedimentary δ15N and inferred-

chlorophyll a appear to track seabird activity, and indicate that seabirds have been present since

the formation of these ponds over the past ~200 years, and that their numbers have increased in

recent years. Interestingly, diatoms do not show any changes in species composition that are

consistent with shifts in δ15N and inferred-chlorophyll a. The lack of a species level response in

diatoms may be related to the dominance of chlorophytes and cyanobacteria in these systems.

INTRODUCTION

On a worldwide scale, seabirds exert comparable pressure on fish stocks to that of commercial

fisheries (Brooke 2004) and, because they feed in the marine environment and nest on land,

seabirds represent important linkages between marine and terrestrial food webs (Polis et al.

2004). For example, densely populated seabird colonies have been shown to transport significant

quantities of marine-derived nutrients and contaminants from the ocean to terrestrial sites via

their guano (Evenset et al. 2004; Blais et al. 2005; Keatley et al. [7]).

The Canadian Arctic is a critical migrating, nesting, and breeding habitat for ~10 million

marine birds (Mallory & Fontaine 2004), however, Arctic regions are likely to change

dramatically over the next few decades in response recent warming and other environmental

stressors (ACIA 2004). Such environmental changes are likely to influence seabird populations,

but understanding the magnitude or even the direction of seabird responses is unclear (Gaston &

Hipfner 1998; Grosbois & Thompson 2005; Gaston et al. 2005a; Gaston et al. 2005b). This

confusion directly relates to the lack of data on long-term, natural variability of seabird

populations. For example, seabird monitoring programs conducted under the auspices of the

Canadian Wildlife Service extend back only to the 1970s, and consist of either cursory aerial

surveys conducted at 5 – 10 year intervals, with intensive colony monitoring occurring at only a

few sites (Gilchrist et al. 2005). Other sources of long-term data, such as traditional ecological

knowledge (TEK), are often not available for species of seabirds that either roost in largely

inaccessible locations and/or are not frequent components of traditional diets (Gilchrist et al.

2005). Thus, alternative proxy data with which seabird population dynamics could be

reconstructed are highly desirable.

Paleolimnology offers the potential to track past seabird population dynamics as,

compared to un-impacted sites, seabird-influenced lakes and ponds have markedly different

limnological properties (i.e., high concentrations of nutrients), unique algal and invertebrate

species compositions, and distinct sedimentary geochemical signatures (Evenset et al. 2004; Blais

et al. 2005; Keatley et al. [7]; Evenset et al. 2007; Michelutti unpublished data). The freshwater

ponds located at Cape Vera on Devon Island, Arctic Canada represent ideal study sites to track

past seabird activity using paleolimnology. This area is home to ~10,000 nesting pairs of

northern fulmars (Fulmarus glacialis), which form dense nesting sites on cliffs that extend above

several small ponds. Recently, Keatley et al. ([7]) documented the influence of seabirds on the

ponds at Cape Vera, and Blais et al. (2005) showed a significant correlation between seabird

influence and concentration of contaminants in surface sediments.

Here, we utilize a multi-proxy approach to reconstruct the influence of seabird activity

over the past several centuries on a series of ponds at Cape Vera. In addition, we analyze two

ponds located near an abandoned colony, at Cape Hawes ~10 km from Cape Vera, but with no

present seabird activity. The proxies we have utilized to track seabird influence include (1) δ15N,

which serves as an ecological tracer for marine-derived nutrients where animals that feed high in

the trophic system, such as seabirds, have elevated δ15N ratios (Mallory 2006); (2) sedimentary

chlorophyll a, which tracks overall aquatic production; and (3) fossil diatom assemblages, which

are known to respond sensitively to changes in nutrient levels (Hall & Smol 1999). Indeed,

similar paleolimnological approaches have been successfully applied in Arctic regions to track

past abundances of sockeye salmon (Gregory-Eaves et al. 2003), as well as to assess the degree of

freshwater eutrophication from slaughtered whale carcasses (Douglas et al. 2004) and human

sewage (Douglas & Smol 2000; Michelutti et al. in press).

The main questions we address in this study are: (1) Do diatom species assemblages and

sedimentary-chla respond synchronously to seabird-derived nutrients (as measured by δ15N)

through time? (2) Have there been historical changes in the current northern fulmar colony at

Cape Vera? If so, in what direction are these changes? (3) Is there any paleolimnological

evidence to support the anecdotal evidence of an abandoned seabird colony at nearby Cape

Hawes?

We show that in most of the eight study ponds at Cape Vera, δ15N and inferred-

chlorophyll a were indicative of seabird-derived nutrient enrichment throughout the entire history

encapsulated in the cores, and had highest concentrations in the most recent sediments. As

expected, the control sites at Cape Hawes recorded low δ15N and chlorophyll a compared to the

Cape Vera sites, although one core recorded evidence of a once active seabird colony. Diatom-

based paleolimnological indices, however, do not show consistent patterns to either δ15N or

sedimentary inferred-chla, confirming previous research that diatom assemblages do not respond

in a simple manner to prolonged nutrient enrichment in High Arctic ponds.

SITE DESCRIPTION

Located in the Canadian High Arctic, Cape Vera, Devon Island, Nunavut (Fig. 1) is home

to a large colony (~10 000 breeding pairs) of northern fulmars (Fulmarus glacialis) (Gaston et al.

2006). Mallory (2006) provides a review of northern fulmar ecology in the Canadian High

Arctic. Current estimates suggest that the Cape Vera colony has not changed appreciably in size

since the 1970s, although this information is largely based on aerial surveys with which accurate

counts of the cryptically-coloured northern fulmar are difficult to obtain (Gaston et al. 2006).

Northern fulmars feed near the top of the marine food web and thus are enriched in δ15N with

average δ15N of ~13-17‰ (Mallory 2006).

The northern fulmars at Cape Vera nest in dolostone / limestone cliffs rising ~250 m

above sea level (Mayr et al. 1998). Stretching below these cliffs are a suite of freshwater ponds

that receive varying amounts of seabird-derived subsidies (e.g. guano, regurgitated stomach oil,

eggshells, feathers, carcasses) that significantly impact the water quality and biota (i.e. diatom

assemblages) within the ponds (Fig. 2; Keatley et al. [7]). Details on the modern limnology and

diatom distributions from these ponds are summarized in Keatley et al. [7]). Briefly, the ponds

most affected by seabirds are characterized by high pH, primary production, and nutrient-related

variables (Table 1). The diatom assemblages recorded at Cape Vera rank amongst the lowest in

the Canadian Arctic in terms of species diversity, and are dominated by small Nitzschia and

Cymbella taxa, some of which show significant relationships to seabird-derived nutrients (δ15N).

METHODS

We identified several ponds as coring candidates based on the criteria that they spanned

an apparent gradient of seabird influence and contained adequate sediment for coring. Sediment

cores were taken from 10 ponds during our field seasons in July 2005 and July 2006 using a

plexi-glass core tube from a Glew gravity corer (inner diameter = 3.75 cm; Glew et al. 2001). For

all but three ponds (CV12, CV13, CV24), replicate cores were obtained in order to perform multi-

proxy analyses. All cores were sectioned in the field at 0.5 cm intervals using a Glew (1988)

extruder, except CV5 which was sectioned into 1.0 cm intervals.

Diatom frustules were prepared for analysis using HNO3 in a MARSX microwave

digestion procedure (Parr et al. 2004) to remove organic material from each sample. The

resultant slurries were repeatedly rinsed with deionized water until a neutral pH was reached,

after which subsamples were strewn onto pre-cleaned cover slips and mounted on slides using

Naphrax®, a mounting medium with a high refractive index. Identifications primarily followed

Antoniades et al. (in press), Krammer and Lange-Bertalot (1991), and Krammer (2002).

Sediment samples from the cores were also analysed for stable isotopes of δ15N, as well

as for elemental analysis (C, N) at G.G. Hatch Laboratory, University of Ottawa, Ottawa, Canada.

The ratio of C/N can be used to assess whether sedimentary organic matter is largely of aquatic

(C/N <10) or terrestrial (C/N >20) origin (Meyers & Ishiwatari 1993). The elemental analyses

were conducted on a Vario EL III (Elementar, Germany), while stable isotopes were measured

using a continuous-flow DeltaPlus Advantage isotope ratio mass spectrometer (ThermoFinnigan,

Germany) coupled with a ConFlo III. Isotope data was normalized using previously calibrated

internal standards, and analytical precision is +/-0.2 ‰.

Sedimentary chlorophyll a (chla) was inferred using a new application of visible

reflectance spectroscopy (Das et al. 2005; Wolfe et al. 2006). Briefly, sediment spectra were

obtained using a FOSS NIR Systems Rapid Content Analyzer, and these data were converted to

chla estimations using the algorithm developed by Michelutti et al. (2005). This technique

effectively measures chla and its derivatives and is sensitive to chla changes on the order of 0 –

0.1 mg/g dry weight. Due to a shortage of sample material in some of the cores, inferred-chla

was analysed for replicate cores when necessary (CV9a, CV22). These replicate cores were

cross-validated using independent 210Pb dating where available.

The sediment cores were radiometrically dated using 210Pb and 137Cs from gamma

spectrometry (Appleby 2001) at either the University of Ottawa or at PEARL, Queen’s

University, Kingston. Both Constant Rate of Supply (CRS) and Constant Initial Concentration

(CIC) age models were explored to determine the most appropriate fit for each core, based on the

137Cs peak (Michelutti et al., submitted).

Statistical analyses

Diatom assemblage data are reported as relative abundances and dominant taxa (>3%

relative abundance) from each core were plotted in a stratigraphy using the program C2 (Juggins

2003). Species data were analysed using a variety of multivariate ordination approaches using

the program CANOCO version 4.5 (ter Braak & Šmilauer 2002). First, a detrended

correspondence analysis (DCA) was used to determine the length of the species gradient (or

amount of species turnover) along the main axis of variation in each core. Since the DCA axis 1

gradient lengths were short (i.e. < 2 standard deviation units), we used the linear ordination

technique of Principal Components Analysis (PCA) to summarize the main directions of variation

in the diatom data (Lepš & Šmilauer 2003).

RESULTS

Core summaries

We obtained sediment cores from eight ponds at Cape Vera (CV5, CV6, CV7, CV9,

CV9a, CV12, CV13, CV20) and two ponds at Cape Hawes (CV22, CV24). The sediment cores

ranged in length from 6 cm to 23 cm long. CIC models dating models were invalidated by non-

monotonic features of the 210Pb profiles in all cores, necessitating the use of CRS models

(Michelutti et al. submitted). However, agreement between the CRS dates and independent

markers (137Cs, PCB data) was only obtained for CV6, CV9a, and CV22. In CV7, CV9 and

CV20, dates are approximated based on 137Cs peaks (Michelutti et al. submitted). The other cores

were challenging to date. Details on all 210Pb profiles and model comparisons can be found in

Michelutti et al. (submitted). Given these challenges, we compare the paleoenvironmental data

with respect to depth for the purposes of this paper. Nevertheless, we suggest that we likely have

the full sedimentary record in these sites as, in most ponds, the core tube hit hard material (likely

bedrock or ice) that prevented the retrieval of longer sediment records.

Due to the wide variety of paleolimnological proxies and divergent responses with the

diatom data, ecosystem-level generalizations are difficult. However, in the interest of space

limitation, we summarize the main findings below, organized by type of proxy indicator. This

section is followed by a more detailed description of each core.

Stable Isotopes: δ15N

The lowest average δ15N (~2‰) was found in the two control sites at Cape Hawes

(CV24, CV22) and in CV12, the Cape Vera pond located furthest from the colony (Figs. 3-11).

The highest average δ15N (~12‰) was found in CV13, CV9, CV9a (Figs. 3-11). With the

exception of CV12, CV22, and CV24, all cores showed general patterns of δ15N enrichment

towards the tops of the cores with the greatest increase in the Cape Vera cores; this pattern was

much smoother in some cores than in others (Figs. 3-11).

Carbon/Nitrogen

Ratios of C/N in the Cape Vera cores were generally >20 in the bottom of most cores,

shifting towards values <10 near the top of the cores (Figs. 3-11). At Cape Hawes, the C/N ratios

are generally below 10 for most of the sedimentary record. Trends in C/N were generally not

consistent with δ15N, chla, or diatom proxies.

Inferred chlorophyll a

Due to a lack of sediment, chla was measured from only seven cores (CV5, CV6, CV9,

CV9a, CV20, CV22, CV24). Chla trends were generally correlated with those found for δ15N

(Figs. 3-11). The two control cores CV22 and CV24, as well as the snow pack-influenced CV5

had the lowest average chla (Figs. 3, 10, 11). At the other end of the spectrum, CV20 (~0.4

mg/g) had the highest chla, closely followed by CV9a, CV9, and CV6 (Figs. 4, 5, 6, 9). All cores

showed increases in chla in the uppermost sediments, although the more heavily affected ponds

exhibited more dramatic increases and these generally began earlier in the sedimentary records

(Fig.1, Figs. 3-11).

Diatoms

Diatom frustules were present throughout the entire length of all cores except CV7,

suggesting that diatoms were relatively well-preserved throughout the cores. While some cores

showed monotonic changes that are consistent with trends in δ15N and/or chla (e.g. CV9, CV13,

CV22), others indicated large fluctuations in diatom assemblage composition that bear little

relationship to either δ15N or chla (e.g. CV5, CV6, CV9a; Figs. 3-11).

Trends in diatom assemblage diversity, as estimated by Hill’s N2, indicated little relation

to patterns of main direction of species variation (PCA1), seabird-derived nutrients (δ15N) or chla

(Figs. 3-11). Diatom species diversity showed no relationship with time, as some cores displayed

greatest diversity in most recent sediments (e.g. CV5, CV9; Figs. 3, 5), while others had the

greatest diversity in the oldest sediments (e.g. CV6, CV9a; Figs. 4, 6). In general, the absolute

value of the Hill’s N2 number from site to site followed a trend of the most affected ponds

exhibiting lower diversity (i.e. CV9, CV20, CV13 (Hill’s N2 = ~5)) than the least affected ponds

(i.e. CV5, CV22 (Hill’s N2 = ~14)).

Detailed core descriptions – Cape Vera cores (CV5 – CV20)

Pond CV5 (Figs. 3, 12)

The 10 cm long core from CV5 represents at least ~100 years of sediment accumulation,

as estimated by the 210Pb date of 1929 at 8.25 cm (Michelutti et al. submitted). Nitzschia

frustulum and N. perminuta are overwhelmingly dominant, accounting for ~20-80% of the

assemblage at any one interval (Fig. 12). Other prominent features of this core included species

associated with higher δ15N sites based on our calibration set (e.g. Encyonema silesiacum, E.

fogedii, E. minutum; Fig. 12). Above ~3.5 cm, Nitzschia spp. decreases in abundance (~10 -

20%), while Navicula spp. (N. chiarae, N. phyllepta, N. salinarum, N. sp. aff veneta),

Microcostatus krasskei and Cymbella spp. sensu lato (~20%) become more important

components of the assemblage (Fig. 12). The species changes are summarized by the PCA axis 1

sample scores, which show non-monotonic changes throughout the core (Fig. 3). The δ15N is high

throughout this core (~7‰) and increases at the top (to ~10‰; Fig. 3). Elemental data shows little

change in %C (~12% throughout), while %N increases slightly at the top of the core (to a

maximum of ~1%); C/N remains above 20 for most of the record, although a recent decline (~15)

is apparent in the upper most sediments (Fig. 3). Inferred-chla is low throughout this core (0.05

mg/g), although a very small increase is apparent at the top of the core. Species diversity (Hill’s

N2) is high and increases towards the top of the core (~11 – 18).

Pond CV6 (Figs. 4, 13)

Pond CV6 drains pond CV5, yet the two sediment cores share little in common. The 12

cm long core from CV6 represents a much longer time period than the CV5 core, as 210Pb

background at 5 cm corresponds to an age of 1920 AD (Michelutti et al. submitted). Although

the entire record is dominated by Nitzschia frustulum (~30-70%), the bottom half of the core is

the most diverse (Hill’s N2 = ~12) and features species characteristic of low δ15N (e.g. Cymbella

botellus, C. cleve-eulerae, Achnanthidium minutissimum, Navicula vulpina, Cyclotella

pseudostelligera, and Fragilaria construens var. venter; Fig. 13). In the upper ~5.5 cm, diversity

decreases (Hill’s N2 = ~6) as Nitzschia frustulum reaches its maximum abundance and the low

δ15N species disappear (Figs. 4, 13). The δ15N remains steady at ~5‰ for most of the core, but

increase to ~10‰ in the top 2.5 cm. The relative stability of both %N is consistent at (~1%) and

%C (~16%) result in a relatively stable C/N ~20 (Fig. 4). The chla is characterized by a doubling

at the top from ~0.1 – 0.2 mg/g (Fig. 4).

Pond CV7 (Fig. 14)

Due to an absence of diatoms throughout much of the CV7 core, we only briefly describe

it here. At 20 cm long, the core from CV7 was amongst the longest retrieved and may represents

~ 200 years of sediment accumulation (137Cs peak at 3 cm = 1963 AD, Michelutti et al.

submitted). The oldest sediments in the assemblage are characterized by relatively few valves

representing Caloneis spp., Cocconeis spp., Cymbella botellus, C. cleve-eulerae, Encyonema

fogedii, Navicula chiarae, N. vulpina, Nitzschia perminuta, and an unidentified fragment,

possibly of marine origin (Fig. 14). No diatoms are present between 19 and 5.25 cm, although

diatoms are abundant in the uppermost ~5 cm (Fig. 14). Sediments above 3 cm depth were

characterized by greater relative abundances of small Nitzschia spp. and Psammothidium

marginulatum, and decreased abundances of Cymbella botellus (Fig. 14).

Pond CV9 (Figs. 5, 15)

Pond CV9 is situated in closest proximity to the bird colony; the 22 cm long core likely

represents ~200 years of accumulation (137Cs peak at 3 cm = 1963 AD, Michelutti et al.

submitted). Below ~14 cm depth, the core is heavily dominated by Nitzschia frustulum (up to

~70% relative abundance); this species is accompanied by small percentages of Cymbella cleve-

eulerae, Navicula vulpina, and Nitzschia perminuta (Fig. 15). Beginning at ~13.25 cm, Nitzschia

frustulum decreases in abundance (although never dropping below 30%) while Encyonema

fogedii and E. minutum (species associated with higher δ15N) become more abundant (Fig. 15).

The PCA1 exhibits a largely monotonic increase toward the top of the core (Fig. 5), and this is

mirrored by increases in δ15N (from ~9 - 17‰), %N (from ~0.5 - ~2%), %C (from ~0 – 16%),

C/N steadily decreasing (above 19 cm), chla (from 0.05 – 0.3 mg/g), and chironomid head

capsule abundance (Michelutti, unpublished data).

Pond CV9a (Figs. 6, 16)

Situated beside CV9, the 12.5 cm core from pond CV9a represents a slower

sedimentation rate, as 7 cm depth represents ~1900 AD (Michelutti et al. submitted). The greatest

diversity (Hill’s N2 = 14) in the sediment core occurs in the bottom sediments and is represented

by Cyclotella pseudostelligera, Encyonema minutum, Fragilaria construens var. venter, F.

pinnata, and Nitzschia frustulum (Fig. 16). The rest of the core is overwhelmingly dominated by

N. frustulum; these diatom assemblage changes are clearly captured by the fluctuations in the

PCA1 scores below 8 cm and by the static PCA1 values from 8 cm (~1910) to the top of the core

(Fig. 6). The δ15N was very high throughout the core (between 9-11‰), while C/N showed little

change (Fig. 6). CV9a had the highest %N (2-5%), %C (16-34%), and chla (0.1-0.5 mg/g) of all

the cores in our study (Fig. 6).

Pond CV12 (Figs. 7, 17)

Of the ponds located at Cape Vera, CV12 was the furthest from the colony. Fossil

diatoms in this 6.5 cm core were difficult to identify due to a large percentage of broken valves,

particularly in the bottom half of the core. Between 3 - 6.5 cm, the diatom assemblage was

characterized by what appeared to be central areas of large Neidium spp. and Navicula spp., and

to a lesser degree smaller Cocconeis spp. (Fig. 17). These may reflect marine influence, as several

Cocconeis spp. have been associated with high conductivity values in Arctic regions (Ng & King

1999). Beginning at ~2.25 cm, these species largely disappear and the core is dominated by

Nitzschia frustulum, N. perminuta and Denticula kuetzingii (Fig. 17). PCA1 scores exhibit a

steady increase towards the top of the core, while δ15N decreases monotonically from ~7‰ to

~1.5‰ (Fig. 7). Percent N and %C exhibit the opposite trend, increasing from ~0 - 2% and 12-

24%, respectively, at top of the core (Fig. 7). Aside from a peak at 4 - 4.5 cm, the C/N declines

towards the top of the core (Fig. 7). Diatom assemblage diversity decreases at the top of the core

(from Hill’s N2 of ~11 to ~5; Fig. 7).

Pond CV13 (Figs. 8, 18)

The presence of stone circles around CV13 suggests that early Arctic peoples were active

in the watershed of this pond. This 13.25 cm core is characterized by relatively low diversity

(Hill’s N2: ~5 – 7), although the upper samples are slightly more diverse (Fig. 8). In the bottom

half of the 13.25 cm core, Nitzschia frustulum (~60%) is the dominant diatom (Fig. 18). Above 4

cm, Cymbella cleve-eulerae, Encyonema silesiacum, E. minutum, and Navicula vulpina become

more common, while Diatoma tenuis largely disappears and Nitzschia frustulum drops to ~30%

(Fig. 18). Seabird-derived nutrients (δ15N) are extremely high throughout this record, and

experience increases in the upper sediments (from ~8 – 18‰; Fig. 8). Likewise, %N and %C

increase from ~0.5% to ~3%, and from ~14-24%, respectively, in concert with the PCA1 scores

and δ15N. C/N shows some fluctuations, but generally decreases from ~24 to ~10 at the top of the

core (Fig. 8).

Pond CV20 (Figs. 9, 19)

The CV20 core had one of the fastest sediment accumulation rates encountered at Cape

Vera, with ~40 years of accumulation contained in the uppermost 6.25 cm (137Cs peak at 6.25 cm

= 1963 AD, Michelutti et al. submitted). There are few changes in the dominant diatom

assemblages through time as Nitzschia frustulum remains relatively stable at ~80% relative

abundance and both Cymbella cleve-eulerae and Encyonema minutum are present at relatively

constant, yet low (<10%) abundances (Fig. 19). The muted changes that occur in the top of the

core include increased relative abundances of Cyclotella pseudostelligera and Nitzschia

perminuta, and decreased abundances of Cymbella silesiaca (Fig. 19). The δ15N remains

relatively constant throughout the core at ~7‰, while %N gradually increases from 3 to 4% at the

top (Fig. 9). The relative stability of %C (~40%) throughout the core results in an equally stable

C/N (~12; Fig. 9). Chla, while very high throughout the record, reaches its maximum

concentration of ~0.5 mg/g in the upper sediments (Fig. 9).

Cape Hawes cores (CV22 and CV24)

Pond CV22 (Figs. 10, 20)

In the CV22 core, 210Pb indicates an age of ~1950 AD at 5 cm (Michelutti et al.

submitted). This record is characterized by relatively high diversity (Hill’s N2 consistently >10;

amongst the highest in our study) and by large changes in diatom composition, as reflected in the

PCA1 scores (Figs. 10, 20). From the bottom to ~4 cm depth, the dominant taxa vary between

either combinations of Cocconeis spp. and Navicula vulpina or Nitzschia frustulum and Nitzschia

perminuta (Fig. 20). Above ~3 cm, N. frustulum and N. perminuta remain dominant, both

Achnanthidium minutissimum and Denticula tenuis reach their greatest relative abundances (up

to ~8%), but meanwhile Navicula vulpina decreases and Cocconeis spp. disappear (Fig. 20).

Seabird-derived nutrients, reflected by δ15N, decrease from ~9‰ in the bottom half of the core to

~1‰ in the top half of the core, a trend opposite to most of the Cape Vera cores (Fig. 10).

Percent N, %C, C/N, and chla, on the other hand, are very low and stable throughout most of the

core (~0%, 0%, ~8, and ~0 mg/g, respectively; Fig. 10). Inferred-chla exhibits a modest increase

in the uppermost sediments (Fig. 10).

Pond CV24 (Figs. 11, 21)

The 6.5 cm core from pond CV24 had an erratic 210Pb profile precluding the construction

of a reliable geochronology. The dominant diatom species in this core do not change through

time, and are continuously characterized by Nitzschia frustulum (60-70% relative abundance; Fig.

21). Seabird-influence was minimal to non-existent throughout this core, as evidenced by the

very low δ15N (~1‰; Fig. 11). At ~3.5% and ~36%, N and C were relatively high, leading to

C/N generally <10 (Fig. 11). Chla was low (~0.1 mg/g) and indicated a very small increase at the

top of the core (Fig. 11).

DISCUSSION

Geochemical evidence

The paleolimnological records of seabird influence in the ponds at Cape Vera and Cape

Hawes are complex. Concentrations of δ15N and chla support the hypothesis that seabirds have

influenced most of the ponds at Cape Vera (except CV12, the pond most distant from the cliffs,

see below), leading to high nutrient conditions and increased primary production, for the entire

history captured by the sediment cores (maximum of ~200 years). Seabirds generally choose to

nest and breed in cliffs that are inaccessible to predators (Mallory & Fontaine 2004), and thus it is

consistent with seabird behavioural traits that they would have colonized the cliffs at Cape Vera

even before the land beneath the cliffs isostatically rebounded from the ocean (i.e. not more than

~5000 years BP; Lowdon & Blake Jr. 1973; Blake Jr. 1975).

Ratios of C/N in the CV cores generally reflect the terrestrial origin of sedimentary

organic matter in the bottom of most cores (C/N>20), shifting towards values indicative of mostly

aquatic origin at the top of the cores (C/N<10; Meyers & Ishiwatari 1993). At Cape Hawes,

however, the C/N ratios are generally below 10 for most of the sedimentary record, suggesting

that they have never received significant amounts of terrestrial organic matter.

Inferred-chla and δ15N from the Cape Hawes sites confirm that few, if any, seabird-

derived subsidies currently reach these ponds, and that primary production is low. Indeed, these

two ponds located ~7 km from the nearest active seabird colony have limnological characteristics

typical of undisturbed High Arctic ponds. However, historically higher δ15N, sedimentary chla,

and chironomid abundance (Michelutti, unpublished data) prior to the 20th century in CV22

suggests that the cliffs at Cape Hawes may have been the location of a seabird colony in the past;

this would be consistent with CWS observations of jewel lichen, a commonly used indicator of

bird guano (Pielou 1994), still visible on the nearby cliff face. As CV24 is slightly further away

from the cliffs at Cape Hawes, the lack of an elevated δ15N earlier in its sedimentary record

suggests that it was not within range of the nutrient deposition from the seabird colony believed

to have existed at this site.

There is some evidence that hydrological factors may be responsible for the varying

responses in some of the cores within the Cape Vera sites. For example, CV9 is one of the

closest ponds to the bird cliffs and displays amongst the highest δ15N and chla concentrations,

which attain their greatest values in the most recent sediments (Fig. 5). In CV9, the main pattern

of diatom species variation and diatom diversity (Hill’s N2) follow the trends recorded by the

δ15N and chla. In contrast, CV12, which is the pond located furthest from the bird colony, has the

lowest δ15N (Fig. 7). The decrease of bird-derived nutrients in the most recent sediment of CV12,

concomitant with the increases recorded in CV9, imply that changes in drainage patterns have

certainly occurred over time in this region. However, the majority of ponds sampled show

similar trends over time with respect to bird-derived nutrients.

Diatom species composition

Unlike the δ15N and inferred-chla data, diatom-based paleolimnological data are less

easily interpretable with respect to impacts from marine-derived nutrients. Although the taxa

found in the sediment cores were generally encountered from the surface sediment calibration of

24 ponds near Cape Vera (Keatley et al. [7]), suggesting that the historical record covers similar

limnological conditions to those captured by our modern pond survey, shifts in assemblages, as

summarized by PCA axis 1 sample scores, show a wide variety of patterns that do not appear to

be related to proximity to the northern fulmar colony or to the trends recorded in δ15N and chla

(Figs. 3-11). Species assemblage changes do not show consistent patterns through time, whether

they are examined as shifts in species composition, the main directions of variation in all taxa

(PCA axis 1 scores), or by a diversity index (i.e. Hill’s N2). Likewise, the most common species,

small Nitzschia taxa, are alternately most dominant in the recent sediments of some ponds (e.g.

CV6), the oldest sediments of others (e.g. CV9), or may show little change throughout the

sediment cores (e.g. CV24). Moreover, Nitzschia frustulum attains extremely high relative

abundance (>50%) even in sites with very different modern bird influences. While the patterns of

diatom species and diversity changes are correlated to both δ15N and chla concentrations in some

ponds (i.e. CV9, CV2), when the diatom changes in all ten ponds are compared, the complex

responses suggest that species assemblage changes alone cannot provide robust historical

estimations of seabird-derived nutrients at Cape Vera.

Given that the Cape Vera ponds are clearly productive, are known to contain seabird

derived nutrients, and that our previous work calibrating diatom surface sediments to measured

ecological variables indicated a significant relationship between diatom species assemblage

variance and δ15N (Keatley et al. [7]), the lack of a diatom assemblage signal that agrees with the

other proxies of seabird influence was initially surprising. However, it is likely that a number of

factors, working in concert, may explain our results.

Previous studies of diatom-inferred eutrophication in Arctic lakes and ponds (Douglas &

Smol 2000; Michelutti et al. 2002; Douglas et al. 2004; Michelutti et al. in press) indicate that

diatoms respond quite differently to nutrient enrichment than those in temperate (Hall & Smol

1999), or even in subarctic regions (e.g. Gregory-Eaves et al. 2003). For example, diatom

assemblage changes in response to human sewage dumped directly into Arctic lakes were

relatively minor, compared to responses recorded in similarly-impacted temperate lakes. This

relatively muted diatom response in previous studies of nutrient-enriched Arctic lakes has been

attributed to the over-riding influence of prolonged ice cover and short growing seasons (Douglas

& Smol 2000; Michelutti et al. 2007). However, at Cape Vera, the cool conditions that

characterize even the summer months have not limited chlorophytes and some cyanobacteria

from achieving very high standing stocks in response to the elevated nutrient inputs from

seabirds. Therefore, it is possible that at Cape Vera, the diatoms are being outcompeted by other

taxa.

In addition, the diatom assemblages in the Cape Vera ponds are almost all dominated by

small Nitzschia frustulum and Nitzschia perminuta taxa. In the Canadian High Arctic, these taxa

are amongst the most common and abundant species found, and are thought to be environmental

generalists (Antoniades et al. in press). For example, taxa such as N. perminuta and N. frustulum

have been found to be the dominant taxa in both highly eutrophic sites (Michelutti et al. 2003)

and in ultraoligotrophic sites (Antoniades et al. in press). In our surface sediment calibration, we

found that δ15N could significantly explain diatom species distributions, yet the relationship

between δ15N and N. frustulum, one of the most dominant taxa, was weak (Keatley et al. [7]).

Regardless of the specific cause, the inability of diatom species assemblages to

consistently track changes in δ15N or chla at Cape Vera suggest that diatom species assemblages

should not be used in isolation to track changes in trophic status over time at Cape Vera. Given

that diatom identification and enumeration are often the most time consuming steps in

paleolimnological studies, their usefulness as paleo-proxies of seabird influence is limited.

Provided that a high degree of chronological control can be achieved, diatom concentration data,

biogenic silica, or some measure of diatom production would likely be a better indicator of

trophic status than diatom species changes. Our findings contrast sharply with those of temperate

regions, in which relatively small changes in nutrient input often result in dramatic shifts in

diatom assemblage composition (Hall & Smol 1999).

Reconstructing seabird populations

Sediment cores from Cape Vera and Cape Hawes reveal highly variable patterns of

diatom species assemblage change, despite relatively coherent trends in δ15N and chla

concentrations. At Cape Vera, the ponds most affected by seabird-derived nutrients (inferred by

δ15N) generally exhibit records of elevated δ15N throughout their entire histories, suggesting that

the northern fulmar colony has been a continuous source of marine-derived nutrients into these

ponds. In comparison, ponds at Cape Hawes have much lower δ15N and chla concentrations in

the recent sediments, consistent with the observed lack of seabird activity. However, one of the

Cape Hawes ponds records historically higher δ15N that corroborates observational evidence that

seabirds abandoned a colony at this site prior to the 20th century.

Recent increases in both δ15N and chla concentrations, as well as decreases in C/N

suggest that aquatic primary production has increased in recent years at Cape Vera, and that this

is likely due to an increase in seabird populations. However, increased primary production may

also be partially attributable to climate warming, as small increases in chla and decreases in C/N

are also apparent in Cape Hawes, despite the lack of a corresponding increase in δ15N. Indeed,

using the same techniques, Michelutti et al. (2005) have attributed increased chla in the recent

sediments of lakes from Baffin Island to climate-related increases in primary production.

Changes in climate, and especially sea ice conditions, are known to have significant repercussions

on some seabird population dynamics (e.g. Gaston et al. 2005), although the influence of

changing ice conditions on northern fulmars in the Canadian High Arctic is unclear. Nevertheless,

it is probable that climate plays a role in determining the success of northern fulmars at Cape

Vera, and that both shifts in seabird population dynamics and climate, acting both alone and in

concert, have resulted in increased production at Cape Vera.

Conversely, diatom-based paleolimnological records were complex and displayed

varying patterns through time in different ponds, several of which were inconsistent with both

δ15N and inferred-chla. At Cape Vera, the lack of diatom species assemblage response to nutrient

enrichment is related to the overwhelming dominance of a few taxa that do not show strong

relationships to δ15N. The diatom-based results contrast those from studies of eutrophication in

temperate regions, and suggest that diatom species in High Arctic freshwaters respond in a highly

complex manner to nutrient enrichment. This relationship may be related to the overwhelming

importance of the phytobenthos relative to the phytoplankton as primary producers in Arctic

regions.

ACKNOWLEDGEMENTS

We would like to thank NSERC, PCSP, NSTP, and CWS for funding and logistical support, and

Irene Gregory-Eaves, Alain Fontaine, Norm North, Mark Wayland, Linda Kimpe, Karen Foster,

Huijun Liu, Jason Akearuk, and Miles Falconer for assistance in the field.

REFERENCES

ACIA, 2004. Impacts of a warming Arctic. Cambridge University Press, Cambridge, UK.

Antoniades, D., P. B. Hamilton, M. S. V. Douglas & J. P. Smol, in press. Freshwater diatoms from the Canadian High Arctic. Iconographica Diatomologica.

Appleby, P. G., 2001. Chronostratigraphic techniques in recent sediments. In W. M. Last & J. P. Smol (eds), Tracking Environmental Change Using Lake Sediments Volume 1: Basin Analysis, Coring, and Chronological Techniques. Kluwer Academic Publishers, Dordrecht: 171-203.

Blais, J. M., L. E. Kimpe, D. McMahon, B. E. Keatley, M. L. Mallory, M. S. V. Douglas, & J. P. Smol, 2005. Arctic seabirds transport marine-derived contaminants. Science 309: 445.

Blake Jr, W., 1975. Radiocarbon age determinations and postglacial emergence at Cape Storm, southern Ellesmere Island, Arctic Canada. Geografiska Annaler Series A-Physical Geography 57: 1-71.

Brooke, M. D., 2004. The food consumption of the world's seabirds. Proceedings of the Royal Society of London Series B-Biological Sciences 271: S246-S248.

Das, B., R. D. Vinebrooke, A. Sanchez-Azofeifa, B. Rivard, & A. P. Wolfe, 2005. Inferring sedimentary chlorophyll concentrations with reflectance spectroscopy: a novel approach to reconstructing historical changes in the trophic status of mountain lakes. Canadian Journal of Fisheries and Aquatic Sciences 62: 1067-1078.

Douglas, M. S. V., & J. P. Smol, 2000. Eutrophication and recovery in the High Arctic: Meretta lake (Cornwallis Island, Nunavut, Canada) revisited. Hydrobiologia 431: 193-204.

Douglas, M. S. V., J. P. Smol, J. M. Savelle, & J. M. Blais, 2004. Prehistoric Inuit whalers affected arctic freshwater ecosystems. Proceedings of the National Academy of Sciences of the United States of America 101: 1613-1617.

Evenset, A., J. Carroll, G. N. Christensen, R. Kallenborn, D. Gregor, & G. W. Gabrielsen, 2007. Seabird guano is an efficient conveyer of persistent organic pollutants (POPs) to Arctic lake ecosystems. Environmental Science & Technology 41: 1173-1179.

Evenset, A., G. N. Christensen, T. Skotvold, E. Fjeld, M. Schlabach, E. Wartena, & D. Gregor, 2004. A comparison of organic contaminants in two high Arctic lake ecosystems, Bjørnøya (Bear Island), Norway. The Science of the Total Environment 318: 125-141.

Gaston, A. J., & M. Hipfner, 1998. The effect of ice conditions in northern Hudson Bay on breeding by Thick-billed Murres (Uria lomvia). Canadian Journal of Zoology 76: 480-492.

Gaston, A. J., H. G. Gilchrist, & J. M. Hipfner, 2005a. Climate change, ice conditions and reproduction in an Arctic nesting marine bird: Brunnich's guillemot (Uria lomvia L.). Journal of Animal Ecology 74: 832-841.

Gaston, A. J., H. G. Gilchrist, & M. L. Mallory, 2005b. Variation in ice conditions has strong effects on the breeding of marine birds at Prince Leopold Island, Nunavut. Ecography 28: 331-344.

Gaston, A. J., M. L. Mallory, H. G. Gilchrist, & K. O'Donovan, 2006. Status, trends and attendance patterns of the northern fulmar Fulmarus glacialis in Nunavut, Canada. Arctic 59: 165-178.

Gilchrist, G., M. Mallory, & F. Merkel, 2005. Can local ecological knowledge contribute to wildlife management? Case studies of migratory birds. Ecology and Society 10.

Glew, J. R., 1988. A portable extruding device for close interval sectioning of unconsolidated core samples. Journal of Paleolimnology 2: 241-243.

Glew, J. R., J. P. Smol, & W. M. Last, 2001. Sediment core collection and extrusion. In W. M. Last & J. P. Smol (eds), Basin Analysis, Coring, and Chronological Techniques. Kluwer Academic Publisher, Norwell: 73-106.

Gregory-Eaves, I., J. P. Smol, M. S. V. Douglas, & B. P. Finney, 2003. Diatoms and sockeye salmon (Oncorhynchus nerka) population dynamics: Reconstructions of salmon-derived nutrients over the past 2,200 years in two lakes from Kodiak Island, Alaska. Journal of Paleolimnology 30: 35-53.

Grosbois, V., & P. M. Thompson, 2005. North Atlantic climate variation influences survival in adult fulmars. Oikos 109: 273-290.

Hall, R. I., & J. P. Smol, 1999. Diatoms as indicators of lake eutrophication. In E. F. Stoermer & J. P. Smol (eds), The Diatoms: Applications for the Environmental and Earth Sciences. Cambridge University Press, New York: 128-168.

Juggins, S., 2003. C2: software for ecological and palaeocological data analysis and visualisation, version 1.3. Department of Geography, University of Newcastle.

Keatley, B. E., M. S. V. Douglas, J. M. Blais, M. Mallory, & J. P. Smol. [7]. Impacts of seabird-derived nutrients on water quality and diatom species assemblages from Cape Vera, Devon Island, Canadian High Arctic. PhD thesis chapter; in preparation.

Krammer, K., 2002. Diatoms of Europe: Cymbella. A.R.G. Gantner Verlag, Ruggell.

Krammer, K., & H. Lange-Bertalot, 1991. Bacillariophyceae. In H. Ettl, J. Gerloff, H. Heynig & D. Mollenhauer (eds). Gustav Fischer Verlag, Stuttgart.

Lepš, J., & P. Šmilauer, 2003. Multivariate analysis of ecological data using CANOCO. Cambridge University Press, London.

Lowdon, J. A.& W. Blake Jr, 1973. Geological Survey of Canada Radiocarbon Dates XIII. Ottawa, Geological Survey of Canada, Ottawa.

Mallory, M. L., 2006. The Northern Fulmar (Fulmarus glacialis) in Arctic Canada: ecology, threats, and what it tells us about marine environmental conditions. Environmental Reviews 14: 187-216.

Mallory, M. L., & A. J. Fontaine, 2004. Key marine habitat sites for migratory birds in Nunavut and the Northwest Territories. Canadian Wildlife Service, Ottawa.

Mayr, U., T. de Freitas, G.H. Eisbacher, & J.Wm. Kerr, 1998. Geology, Prince Alfred Bay-Cardigan Strait and parts of adjacent sheets, Northwest Territories; Geological Survey of Canada, Map 1924A, scale 1:250 000.

Meyers, P. A., & R. Ishiwatari, 1993. Lacustrine Organic Geochemistry - An Overview of Indicators of Organic-Matter Sources and Diagenesis in Lake-Sediments. Organic Geochemistry 20: 867-900.

Michelutti, N., M. S. V. Douglas, & J. P. Smol, 2002. Tracking recovery in a eutrophied high Arctic lake (Meretta Lake, Cornwallis Island, Canadian Arctic) using periphytic diatoms. Verh. int. Ver. Limnol. 28.

Michelutti, N., M. S. V. Douglas, & J. P. Smol, 2003. Tracking recovery in a eutrophied high Arctic lake (Meretta Lake, Cornwallis Island, Canadian Arctic) using periphytic diatoms. Verh. int. Ver. Limnol. 28: 1533-1537.

Michelutti, N., M. H. Hermanson, J. P. Smol, P. J. Dillon, & M. S. V. Douglas, in press. Delayed response of diatom assemblage changes to sewage inputs in an Arctic lake. Aquatic Sciences.

Michelutti, N., A. P. Wolfe, R. D. Vinebrooke, B. Rivard, & J. P. Briner, 2005. Recent primary production increases in arctic lakes. Geophysical Research Letters 32: L19715-doi: 10.1029/2005GL023693.

Ng, S. L., & R. H. King, 1999. Development of a diatom-based specific conductivity model for the glacio-isostatic lakes of Truelove Lowland: implications for paleoconductivity and paleoenvironmental reconstructions in Devon Island lakes, NWT, Canada. Journal of Paleolimnology 22: 367-382.

Parr, J. F., K. H. Taffs, & C. M. Lane, 2004. A microwave digestion technique for the extraction of fossil diatoms from coastal lake and swamp sediments. Journal of Paleolimnology 31: 383-390.

Pielou, E. C., 1994. A Naturalist’s Guide to the Arctic. University of Chicago Press, Chicago.

Polis, G. A., F. Sánchez-Piñero, P. T. Stapp, W. B. Anderson & M. D. Rose, 2004. Trophic Flows from Water to Land: Marine Input Affects Food Webs of Islands and Coastal Ecosystems Worldwide. In G. A. Polis, M. E. Power & G. R. Huxel (eds), Food Webs at the Landscape Level. University of Chicago Press, Chicago: 200-216.

ter Braak, C. J. F., & P. Šmilauer, 2002. CANOCO reference manual and CANOdraw for Windows user guide: Software for Canonical Community Ordination (Version 4.5).

Wolfe, A. P., R. D. Vinebrooke, N. Michelutti, B. Rivard, & B. Das, 2006. Experimental calibration of lake-sediment spectral reflectance to chlorophyll a concentrations: methodology and paleolimnological validation. Journal of Paleolimnology 36: 91-100.

Figure captions.

Figure 1. Location of study site. A) Regional map of the Canadian Arctic with inset detailing location of the Canadian Arctic within Canada; B) Enlargement of Devon Island region with star identifying the location of Cape Vera; and C) Location of study ponds at Cape Vera, Devon Island. Figure 2. Photograph of a suite of ponds below the densely occupied cliffs at Cape Vera, Devon Island. Figures 3 – 11. Plots for each core summarizing overall changes in diatom assemblage composition based on Principal Components Analysis Axis 1 (PCA1) and PCA axis 2 (PCA2) sample scores, changes in diatom assemblage diversity based on Hill’s N2 (N2; Hill 1973), seabird-derived nutrients based on (δ15N (‰), sedimentary-inferred chlorophyll a concentrations (Chla mg/g dry weight), %N, %C, and C/N ratios. Figure 3. CV5 summary plot. Figure 4. CV6 summary plot. Figure 5. CV9 summary plot. Figure 6. CV9a summary plot. Figure 7. CV12 summary plot. Figure 8. CV13 summary plot. Figure 9. CV20 summary plot. Figure 10. CV22 summary plot. Figure 11. CV24 summary plot. Figures 12 – 21. Diatom stratigraphic plots for each core from the Cape Vera ponds. These plots include all species present in at least 3% relative abundance in at least one sample. Figure 12. CV5 diatom stratigraphy. Figure 13. CV6 diatom stratigraphy. Figure 14. CV7 diatom stratigraphy. Figure 15. CV9 diatom stratigraphy. Figure 16. CV9a diatom stratigraphy. Figure 17. CV12 diatom stratigraphy. Figure 18. CV13 diatom stratigraphy. Figure 19. CV20 diatom stratigraphy. Figure 20. CV22 diatom stratigraphy. Figure 21. CV24 diatom stratigraphy.

A

BC

A

BC

Fi

gure

1.

Figure 2.

0.25

1.25

2.25

3.25

4.25

5.25

Depth (cm)

-2.0

-1.0

0.0

1.0

2.0

PCA1

-2.0

-1.0

0.0

1.0

2.0

PCA2

04

812

1620

N2

04

812

1620

δ15N (‰

)

0.0

0.2

0.4

0.6

Chla(m

g/g dw

)

0

%N

020

40

%C

0.0

2.4

4.8

7.2

9.6

12.0

C/N

0.25

1.25

2.25

3.25

4.25

5.25

Depth (cm)

-2.0

-1.0

0.0

1.0

2.0

PCA1

-2.0

-1.0

0.0

1.0

2.0

PCA2

04

812

1620

N2

04

812

1620

δ15N (‰

)

0.0

0.2

0.4

0.6

Chla(m

g/g dw

)

0

%N

020

40

%C

0.0

2.4

4.8

7.2

9.6

12.0

C/N

0.25

1.25

2.25

3.25

4.25

5.25

Depth (cm)

-2.0

-1.0

0.0

1.0

2.0

PCA1

-2.0

-1.0

0.0

1.0

2.0

PCA2

04

812

1620

N2

04

812

1620

δ15N (‰

)

0.0

0.2

0.4

0.6

Chla(m

g/g dw

)

0

%N

020

40

%C

0.0

2.4

4.8

7.2

9.6

12.0

C/N

Fi

gure

3.

0.25

1.25

2.25

3.25

4.25

5.25

6.25

7.25

8.25

9.25

10.2

5

11.2

5

Depth (cm)

-2.0

-1.0

0.0

1.0

2.0

PCA1

-2.0

-1.0

0.0

1.0

2.0

3.0

PCA2

04

812

1620

N20

48

1216

20

δ15N (‰

)

0.0

0.2

0.4

0.6

Chla (mg/g

dw)

0%N

020

%C

010

2030

C/N

0.25

1.25

2.25

3.25

4.25

5.25

6.25

7.25

8.25

9.25

10.2

5

11.2

5

Depth (cm)

-2.0

-1.0

0.0

1.0

2.0

PCA1

-2.0

-1.0

0.0

1.0

2.0

3.0

PCA2

04

812

1620

N20

48

1216

20

δ15N (‰

)

0.0

0.2

0.4

0.6

Chla (mg/g

dw)

0%N

020

%C

010

2030

C/N

0.25

1.25

2.25

3.25

4.25

5.25

6.25

7.25

8.25

9.25

10.2

5

11.2

5

Depth (cm)

-2.0

-1.0

0.0

1.0

2.0

PCA1

-2.0

-1.0

0.0

1.0

2.0

3.0

PCA2

04

812

1620

N20

48

1216

20

δ15N (‰

)

0.0

0.2

0.4

0.6

Chla (mg/g

dw)

0%N

020

%C

010

2030

C/N

010

2030

C/N Fi

gure

4.

0.25

2.25

4.25

6.25

8.25

10.2

5

12.2

5

14.2

5

16.2

5

18.2

5

20.2

5

21.7

5

Depth (cm)

-2.0

-1.0

0.0

1.0

2.0

PCA1

-2.0

-1.0

0.0

1.0

2.0

PCA2

04

812

1620

N2

04

812

1620

δ15N (‰

)

0.0

0.2

0.4

0.6

Chla (mg/g

dw)

0%N

020%C 0

816

2432

40

C/N0.

25

2.25

4.25

6.25

8.25

10.2

5

12.2

5

14.2

5

16.2

5

18.2

5

20.2

5

21.7

5

Depth (cm)

-2.0

-1.0

0.0

1.0

2.0

PCA1

-2.0

-1.0

0.0

1.0

2.0

PCA2

04

812

1620

N2

04

812

1620

δ15N (‰

)

0.0

0.2

0.4

0.6

Chla (mg/g

dw)

0%N

020%C 0

816

2432

40

C/N0.

25

2.25

4.25

6.25

8.25

10.2

5

12.2

5

14.2

5

16.2

5

18.2

5

20.2

5

21.7

5

Depth (cm)

-2.0

-1.0

0.0

1.0

2.0

PCA1

-2.0

-1.0

0.0

1.0

2.0

PCA2

04

812

1620

N2

04

812

1620

δ15N (‰

)

0.0

0.2

0.4

0.6

Chla (mg/g

dw)

04

812

1620

N2

04

812

1620

δ15N (‰

)

0.0

0.2

0.4

0.6

Chla (mg/g

dw)

0%N

020%C 0

816

2432

40

C/N

0%N

020%C 0

816

2432

40

C/N

Fi

gure

5.

0.25

1.25

2.25

3.25

4.25

5.25

6.25

7.25

8.25

9.25

10.2

5

11.2

5

12.2

5

Depth (cm)

-1.0

0.0

1.0

2.0

3.0

4.0

PCA1

-2.0

-1.0

0.0

1.0

2.0

3.0

PCA2

05

1015

20

N2

04

812

1620

δ15N (‰

)

0.0

0.2

0.4

0.6

Chla (mg/g

dw)

0

%N 020

40

%C

03

69

1215

C/N

0.25

1.25

2.25

3.25

4.25

5.25

6.25

7.25

8.25

9.25

10.2

5

11.2

5

12.2

5

Depth (cm)

-1.0

0.0

1.0

2.0

3.0

4.0

PCA1

-2.0

-1.0

0.0

1.0

2.0

3.0

PCA2

05

1015

20

N2

04

812

1620

δ15N (‰

)

0.0

0.2

0.4

0.6

Chla (mg/g

dw)

0

%N 020

40

%C

03

69

1215

C/N

0.25

1.25

2.25

3.25

4.25

5.25

6.25

7.25

8.25

9.25

10.2

5

11.2

5

12.2

5

Depth (cm)

-1.0

0.0

1.0

2.0

3.0

4.0

PCA1

-2.0

-1.0

0.0

1.0

2.0

3.0

PCA2

05

1015

20

N2

04

812

1620

δ15N (‰

)

0.0

0.2

0.4

0.6

Chla (mg/g

dw)

0

%N 020

40

%C

04

812

1620

δ15N (‰

)

0.0

0.2

0.4

0.6

Chla (mg/g

dw)

0

%N 020

40

%C

03

69

1215

C/N

Fi

gure

6.

0.25

1.25

2.25

3.25

4.25

5.25

6.25

Depth (cm)

-2.0

-1.0

0.0

1.0

2.0

PCA1

-2.0

-1.0

0.0

1.0

2.0

PCA2

04

812

1620

N20

48

1216

20

δ15N (‰

)

0

%N

020%C

080

160

240

320

C/N0.

25

1.25

2.25

3.25

4.25

5.25

6.25

Depth (cm)

-2.0

-1.0

0.0

1.0

2.0

PCA1

-2.0

-1.0

0.0

1.0

2.0

PCA2

04

812

1620

N20

48

1216

20

δ15N (‰

)

0

%N

020%C

080

160

240

320

C/N

Fi

gure

7.

0.25

1.25

2.25

3.25

4.25

5.25

6.25

7.25

8.25

9.25

10.2

5

11.2

5

12.2

5

13.2

5

Depth (cm)

-2.0

-1.0

0.0

1.0

2.0

PCA1

-3.0

-1.0

1.0

3.0

PCA2

04

812

1620

N20

48

1216

20

δ15N (‰

)

0

%N

020%C

05

1015

2025

C/N0.

25

1.25

2.25

3.25

4.25

5.25

6.25

7.25

8.25

9.25

10.2

5

11.2

5

12.2

5

13.2

5

Depth (cm)

-2.0

-1.0

0.0

1.0

2.0

PCA1

-3.0

-1.0

1.0

3.0

PCA2

04

812

1620

N20

48

1216

20

δ15N (‰

)

0

%N

020%C

05

1015

2025

C/N

Fi

gure

8.

0.25

1.25

2.25

3.25

4.25

5.25

6.25

7.25

Depth (cm)

-2.0

-1.0

0.0

1.0

2.0

PCA1

-2.0

-1.0

0.0

1.0

2.0

PCA2

04

812

1620

N2

04

812

1620

δ15N (‰

)

0.0

0.2

0.4

0.6

ChlaC2

(mg/g

dw)

0

%N

020

40

%C

0.0

2.4

4.8

7.2

9.6

12.0

C/N

0.25

1.25

2.25

3.25

4.25

5.25

6.25

7.25

Depth (cm)

-2.0

-1.0

0.0

1.0

2.0

PCA1

-2.0

-1.0

0.0

1.0

2.0

PCA2

04

812

1620

N2

04

812

1620

δ15N (‰

)

0.0

0.2

0.4

0.6

ChlaC2

(mg/g

dw)

0

%N

020

40

%C

0.0

2.4

4.8

7.2

9.6

12.0

C/N

Figu

re 9

.

0.25

1.25

2.25

3.25

4.25

5.25

6.25

7.25

Depth (cm)

-2.0

-1.0

0.0

1.0

2.0

PCA1

-2.0

-1.0

0.0

1.0

2.0

PCA2

04

812

1620

N2

04

812

1620

δ15N (‰

)

0.0

0.2

0.4

0.6

ChlaC2

(mg/g

dw)

0

%N

0

%C

0.0

1.0

2.0

3.0

4.0

5.0

C/N

0.25

1.25

2.25

3.25

4.25

5.25

6.25

7.25

Depth (cm)

-2.0

-1.0

0.0

1.0

2.0

PCA1

-2.0

-1.0

0.0

1.0

2.0

PCA2

04

812

1620

N2

04

812

1620

δ15N (‰

)

0.0

0.2

0.4

0.6

ChlaC2

(mg/g

dw)

0

%N

0

%C

0.0

1.0

2.0

3.0

4.0

5.0

C/N

Figu

re 1

0.

0.25

1.25

2.25

3.25

4.25

5.25

Depth (cm)

-2.0

-1.0

0.0

1.0

2.0

PCA1

-2.0

-1.0

0.0

1.0

2.0

PCA2

04

812

1620

N2

04

812

1620

δ15N (‰

)

0.0

0.2

0.4

0.6

Chla (mg/g

dw)

0

%N

020

40

%C

0.0

2.4

4.8

7.2

9.6

12.0

C/N

0.25

1.25

2.25

3.25

4.25

5.25

Depth (cm)

-2.0

-1.0

0.0

1.0

2.0

PCA1

-2.0

-1.0

0.0

1.0

2.0

PCA2

04

812

1620

N2

04

812

1620

δ15N (‰

)

0.0

0.2

0.4

0.6

Chla (mg/g

dw)

0

%N

020

40

%C

0.0

2.4

4.8

7.2

9.6

12.0

C/N

Figu

re 1

1.

0.5

1.5

2.5

3.5

4.5

5.5

6.5

7.5

8.5

9.5

Depth (cm)

0Cymbe

llabo

tellus

0Psammoth

idium

chlid

anos

020

Encyo

nemaminu

tum

0Navicu

lach

iarae

0Encyo

nemasil

esiac

um0

20

Rel

ativ

e ab

unda

nce

(%)

Encyo

nemafog

edii

020

40Nitz

schia

perm

inuta

020

40

Nitzsc

hiafru

stulum

0Navicu

laph

yllepta

0Microc

ostat

uskra

sske

i

0Navicu

lasa

linaru

m

0Navicu

lasp

affv

eneta

0.5

1.5

2.5

3.5

4.5

5.5

6.5

7.5

8.5

9.5

Depth (cm)

0Cymbe

llabo

tellus

0Psammoth

idium

chlid

anos

020

Encyo

nemaminu

tum

0Navicu

lach

iarae

0Encyo

nemasil

esiac

um0

20

Rel

ativ

e ab

unda

nce

(%)

Encyo

nemafog

edii

020

40Nitz

schia

perm

inuta

020

40

Nitzsc

hiafru

stulum

0Navicu

laph

yllepta

0Microc

ostat

uskra

sske

i0.

5

1.5

2.5

3.5

4.5

5.5

6.5

7.5

8.5

9.5

Depth (cm)

0Cymbe

llabo

tellus

0Psammoth

idium

chlid

anos

020

Encyo

nemaminu

tum

0Navicu

lach

iarae

0Encyo

nemasil

esiac

um0

20

Rel

ativ

e ab

unda

nce

(%)

Encyo

nemafog

edii

020

40Nitz

schia

perm

inuta

020

40

Nitzsc

hiafru

stulum

0Navicu

laph

yllepta

0Microc

ostat

uskra

sske

i

0Navicu

lasa

linaru

m

0Navicu

lasp

affv

eneta

Fi

gure

12.

0.25

1.25

2.25

3.25

4.25

5.25

6.25

7.25

8.25

9.25

10.2

5

11.2

5

Depth (cm)

0Stauros

iraco

nstru

ensva

rvente

r

0Cyclot

ellaps

eudo

stellig

era

020

Cymbe

llabo

tellus 0Ach

nanth

idium

minutis

simum

0Cymbe

llacle

ve-eu

lerae

0

Rel

ativ

e ab

unda

nce

(%)

Encyo

nops

ismicr

ocep

hala

020

Navicu

lavu

lpina

020

Nitzsc

hiape

rminu

ta

020

4060

80

Nitzsc

hiafru

stulum

0Navicu

lach

iarae

0Encyo

nemafog

edii

0.25

1.25

2.25

3.25

4.25

5.25

6.25

7.25

8.25

9.25

10.2

5

11.2

5

Depth (cm)

0Stauros

iraco

nstru

ensva

rvente

r

0Cyclot

ellaps

eudo

stellig

era

020

Cymbe

llabo

tellus 0Ach

nanth

idium

minutis

simum

0Cymbe

llacle

ve-eu

lerae

0

Rel

ativ

e ab

unda

nce

(%)

Encyo

nops

ismicr

ocep

hala

020

Navicu

lavu

lpina

020

Nitzsc

hiape

rminu

ta

020

4060

80

Nitzsc

hiafru

stulum

0.25

1.25

2.25

3.25

4.25

5.25

6.25

7.25

8.25

9.25

10.2

5

11.2

5

Depth (cm)

0Stauros

iraco

nstru

ensva

rvente

r

0Cyclot

ellaps

eudo

stellig

era

020

Cymbe

llabo

tellus 0Ach

nanth

idium

minutis

simum

0Cymbe

llacle

ve-eu

lerae

0

Rel

ativ

e ab

unda

nce

(%)

Encyo

nops

ismicr

ocep

hala

020

Navicu

lavu

lpina

020

Nitzsc

hiape

rminu

ta

020

4060

80

Nitzsc

hiafru

stulum

0Navicu

lach

iarae

0Encyo

nemafog

edii

Fi

gure

13.

0.25

1.25

2.25

3.25

4.25

5.25

6.25

7.25

8.25

9.25

10.2

5

11.2

5

12.2

5

13.2

5

14.2

5

15.2

5

16.2

5

17.2

5

18.2

5

19.2

5

Depth (cm)

020

Cocco

neis

sp 0Uniden

tified

fragm

ent

0Calone

issili

cula

0Navicu

lavu

lpina

020

Cymbe

llabo

tellus

020

40

Nitzsc

hiape

rminu

ta 020

Rel

ativ

e ab

unda

nce

(%)

Navicu

lach

iarae

0Navicu

laph

yllepta

020

Nitzsc

hiafru

stulum 0Cym

bella

cleve

-euler

ae

0Encyo

nemafog

edii

0Navicu

lasp

affv

eneta

0Navicu

laps

eudo

tenell

oides

0Encyo

nemaminu

tum

0Fragila

riaca

pucin

a

0Achna

nthidi

ummarg

inulat

um

0Microc

ostat

uskra

sske

i

0Navicu

lasp

.

(cf.N

. cryp

tocep

hala)

0.25

1.25

2.25

3.25

4.25

5.25

6.25

7.25

8.25

9.25

10.2

5

11.2

5

12.2

5

13.2

5

14.2

5

15.2

5

16.2

5

17.2

5

18.2

5

19.2

5

Depth (cm)

020

Cocco

neis

sp 0Uniden

tified

fragm

ent

0Calone

issili

cula

0Navicu

lavu

lpina

020

Cymbe

llabo

tellus

020

40

Nitzsc

hiape

rminu

ta 020

Rel

ativ

e ab

unda

nce

(%)

Navicu

lach

iarae

0Navicu

laph

yllepta

020

Nitzsc

hiafru

stulum

0.25

1.25

2.25

3.25

4.25

5.25

6.25

7.25

8.25

9.25

10.2

5

11.2

5

12.2

5

13.2

5

14.2

5

15.2

5

16.2

5

17.2

5

18.2

5

19.2

5

Depth (cm)

020

Cocco

neis

sp 0Uniden

tified

fragm

ent

0Calone

issili

cula

0Navicu

lavu

lpina

020

Cymbe

llabo

tellus

020

40

Nitzsc

hiape

rminu

ta 020

Rel

ativ

e ab

unda

nce

(%)

Navicu

lach

iarae

0Navicu

laph

yllepta

020

Nitzsc

hiafru

stulum 0Cym

bella

cleve

-euler

ae

0Encyo

nemafog

edii

0Navicu

lasp

affv

eneta

0Navicu

laps

eudo

tenell

oides

0Encyo

nemaminu

tum

0Fragila

riaca

pucin

a

0Achna

nthidi

ummarg

inulat

um

0Microc

ostat

uskra

sske

i

0Navicu

lasp

.

(cf.N

. cryp

tocep

hala)

Fi

gure

14.

0.25

1.25

2.25

3.25

4.25

5.25

6.25

7.25

8.25

9.25

10.2

511

.25

12.2

513

.25

14.2

515

.25

16.2

517

.25

18.2

519

.25

20.2

521

.25

21.7

5

Depth (cm)

0Cymbe

llacle

ve-eu

lerae

020

4060

80

Nitzsc

hiafru

stulum

020

Navicu

lavu

lpina 0

Rel

ativ

e ab

unda

nce

(%)

Encyo

nops

ismicr

ocep

hala

020

Nitzsc

hiape

rminu

ta

0Encyo

nemasil

esiac

um

020

Encyo

nemafog

edii

020

Encyo

nemaminu

tum

020

Navicu

laph

yllepta

0Navicu

lacry

ptoce

phala

0.25

1.25

2.25

3.25

4.25

5.25

6.25

7.25

8.25

9.25

10.2

511

.25

12.2

513

.25

14.2

515

.25

16.2

517

.25

18.2

519

.25

20.2

521

.25

21.7

5

Depth (cm)

0Cymbe

llacle

ve-eu

lerae

020

4060

80

Nitzsc

hiafru

stulum

020

Navicu

lavu

lpina 0

Rel

ativ

e ab

unda

nce

(%)

Encyo

nops

ismicr

ocep

hala

0.25

1.25

2.25

3.25

4.25

5.25

6.25

7.25

8.25

9.25

10.2

511

.25

12.2

513

.25

14.2

515

.25

16.2

517

.25

18.2

519

.25

20.2

521

.25

21.7

5

Depth (cm)

0Cymbe

llacle

ve-eu

lerae

020

4060

80

Nitzsc

hiafru

stulum

020

Navicu

lavu

lpina 0

Rel

ativ

e ab

unda

nce

(%)

Encyo

nops

ismicr

ocep

hala

020

Nitzsc

hiape

rminu

ta

0Encyo

nemasil

esiac

um

020

Encyo

nemafog

edii

020

Encyo

nemaminu

tum

020

Navicu

laph

yllepta

0Navicu

lacry

ptoce

phala

Fi

gure

15.

0.25

1.25

2.25

3.25

4.25

5.25

6.25

7.25

8.25

9.25

10.2

5

11.2

5

12.2

5

Depth (cm)

0Fragila

riaca

pucin

avargra

cilis

020

Stauros

irella

pinna

ta

020

Cyclot

ellaps

eudo

stellig

era0Cym

bella

cleve

-euler

ae0

20

Rel

ativ

e ab

unda

nce

(%)

Stauros

iraco

nstru

ensva

rvente

r

020

Encyo

nemaminu

tum

020

4060

80

Nitzsc

hiafru

stulum

020

4060

Nitzsc

hiape

rminu

ta0.

25

1.25

2.25

3.25

4.25

5.25

6.25

7.25

8.25

9.25

10.2

5

11.2

5

12.2

5

Depth (cm)

0Fragila

riaca

pucin

avargra

cilis

020

Stauros

irella

pinna

ta

020

Cyclot

ellaps

eudo

stellig

era0Cym

bella

cleve

-euler

ae0

20

Rel

ativ

e ab

unda

nce

(%)

Stauros

iraco

nstru

ensva

rvente

r

020

Encyo

nemaminu

tum

020

4060

80

Nitzsc

hiafru

stulum

020

4060

Nitzsc

hiape

rminu

ta0.

25

1.25

2.25

3.25

4.25

5.25

6.25

7.25

8.25

9.25

10.2

5

11.2

5

12.2

5

Depth (cm)

0Fragila

riaca

pucin

avargra

cilis

020

Stauros

irella

pinna

ta

020

Cyclot

ellaps

eudo

stellig

era0Cym

bella

cleve

-euler

ae0

20

Rel

ativ

e ab

unda

nce

(%)

Stauros

iraco

nstru

ensva

rvente

r

020

Encyo

nemaminu

tum

020

4060

80

Nitzsc

hiafru

stulum

020

4060

Nitzsc

hiape

rminu

ta

Fi

gure

16.

0.25

1.25

2.25

3.25

4.25

5.25

6.25

Depth (cm)

020Nav

icula

sp. (l

arge c

entre

sonly)

0Cocco

neis

sp

020

4060

Neidium

sp. (l

arge c

entre

sonly)

0Navicu

lasp

.

(cfN. rh

ynch

ocep

hala

f. eleg

ans)

0

Rel

ativ

e ab

unda

nce

(%)

Diatom

atenuis

020Nav

icula

phylle

pta

020Den

ticula

kuetz

ingii

020

Nitzsc

hiape

rminu

ta

0Nitzsc

hiaco

mmuntat

a

020

4060

Nitzsc

hiafru

stulum

0Cymbe

llacle

ve-eu

lerae

0.25

1.25

2.25

3.25

4.25

5.25

6.25

Depth (cm)

020Nav

icula

sp. (l

arge c

entre

sonly)

0Cocco

neis

sp

020

4060

Neidium

sp. (l

arge c

entre

sonly)

0Navicu

lasp

.

(cfN. rh

ynch

ocep

hala

f. eleg

ans)

0

Rel

ativ

e ab

unda

nce

(%)

Diatom

atenuis

020Nav

icula

phylle

pta

020Den

ticula

kuetz

ingii

020

Nitzsc

hiape

rminu

ta

0Nitzsc

hiaco

mmuntat

a

020

4060

Nitzsc

hiafru

stulum

0Cymbe

llacle

ve-eu

lerae

0.25

1.25

2.25

3.25

4.25

5.25

6.25

Depth (cm)

020Nav

icula

sp. (l

arge c

entre

sonly)

0Cocco

neis

sp

020

4060

Neidium

sp. (l

arge c

entre

sonly)

0Navicu

lasp

.

(cfN. rh

ynch

ocep

hala

f. eleg

ans)

0

Rel

ativ

e ab

unda

nce

(%)

Diatom

atenuis

020Nav

icula

phylle

pta

020Den

ticula

kuetz

ingii

020

Nitzsc

hiape

rminu

ta

0Nitzsc

hiaco

mmuntat

a

020

4060

Nitzsc

hiafru

stulum

0Cymbe

llacle

ve-eu

lerae

Fi

gure

17.

0.25

1.25

2.25

3.25

4.25

5.25

6.25

7.25

8.25

9.25

10.2

5

11.2

5

12.2

5

13.2

5

Depth (cm)

020

Diatom

atenuis 0

2040

60

Nitzsc

hiafru

stulum

020

Rel

ativ

e ab

unda

nce

(%)

Encyo

nemafog

edii

020

Nitzsc

hiape

rminu

ta

0Encyo

nemasil

esiac

um

020

Encyo

nemaminu

tum

020

Cymbe

llacle

ve-eu

lerae

0Navicu

lavu

lpina

0.25

1.25

2.25

3.25

4.25

5.25

6.25

7.25

8.25

9.25

10.2

5

11.2

5

12.2

5

13.2

5

Depth (cm)

020

Diatom

atenuis 0

2040

60

Nitzsc

hiafru

stulum

020

Rel

ativ

e ab

unda

nce

(%)

Encyo

nemafog

edii

020

Nitzsc

hiape

rminu

ta

0Encyo

nemasil

esiac

um

020

Encyo

nemaminu

tum

020

Cymbe

llacle

ve-eu

lerae

0.25

1.25

2.25

3.25

4.25

5.25

6.25

7.25

8.25

9.25

10.2

5

11.2

5

12.2

5

13.2

5

Depth (cm)

020

Diatom

atenuis 0

2040

60

Nitzsc

hiafru

stulum

020

Rel

ativ

e ab

unda

nce

(%)

Encyo

nemafog

edii

020

Nitzsc

hiape

rminu

ta

0Encyo

nemasil

esiac

um

020

Encyo

nemaminu

tum

020

Cymbe

llacle

ve-eu

lerae

0Navicu

lavu

lpina

Fi

gure

18.

0.25

0.75

1.25

1.75

2.25

2.75

3.25

3.75

4.25

4.75

5.25

5.75

6.25

6.75

7.25

7.75

Depth (cm)

0Encyo

nemasile

siacu

m0

2040

6080

Rel

ativ

e ab

unda

nce

(%)

Nitzsc

hiafru

stulum

0Cymbe

llacle

ve-eu

lerae

020

Encyo

nemaminu

tum

020

Nitzsc

hiape

rminu

ta

0Encyo

nemafog

edii

0Cyclot

ellaps

eudo

stellig

era

0.25

0.75

1.25

1.75

2.25

2.75

3.25

3.75

4.25

4.75

5.25

5.75

6.25

6.75

7.25

7.75

Depth (cm)

0Encyo

nemasile

siacu

m0

2040

6080

Rel

ativ

e ab

unda

nce

(%)

Nitzsc

hiafru

stulum

0Cymbe

llacle

ve-eu

lerae

020

Encyo

nemaminu

tum

020

Nitzsc

hiape

rminu

ta

0Encyo

nemafog

edii

0Cyclot

ellaps

eudo

stellig

era

Fi

gure

19.

0.25

0.75

1.25

1.75

2.25

2.75

3.25

3.75

4.25

4.75

5.25

5.75

6.25

6.75

7.25

7.75

Depth (cm)

0Cocco

neis

sp. s

mall

020

Cocco

neis

sp. la

rge

020

40

Navicu

lavu

lpina

0Navicu

ladig

itorad

iata

0Nitzsc

hiadis

sipata

varmed

ia

0

Rel

ativ

e ab

unda

nce

(%)

Cymbe

llacle

ve-eu

lerae

0Cymbe

llabo

tellus

0Achna

nthidi

umminu

tissim

um

0Dentic

ulaten

uis

020

40

Nitzsc

hiape

rminu

ta

020

40

Nitzsc

hiafru

stulum

0Rossit

hidium

peter

senii

0.25

0.75

1.25

1.75

2.25

2.75

3.25

3.75

4.25

4.75

5.25

5.75

6.25

6.75

7.25

7.75

Depth (cm)

0Cocco

neis

sp. s

mall

020

Cocco

neis

sp. la

rge

020

40

Navicu

lavu

lpina

0Navicu

ladig

itorad

iata

0Nitzsc

hiadis

sipata

varmed

ia

0

Rel

ativ

e ab

unda

nce

(%)

Cymbe

llacle

ve-eu

lerae

0Cymbe

llabo

tellus

0Achna

nthidi

umminu

tissim

um

0Dentic

ulaten

uis

020

40

Nitzsc

hiape

rminu

ta

020

40

Nitzsc

hiafru

stulum

0.25

0.75

1.25

1.75

2.25

2.75

3.25

3.75

4.25

4.75

5.25

5.75

6.25

6.75

7.25

7.75

Depth (cm)

0Cocco

neis

sp. s

mall

020

Cocco

neis

sp. la

rge

020

40

Navicu

lavu

lpina

0Navicu

ladig

itorad

iata

0Nitzsc

hiadis

sipata

varmed

ia

0

Rel

ativ

e ab

unda

nce

(%)

Cymbe

llacle

ve-eu

lerae

0Cymbe

llabo

tellus

0Achna

nthidi

umminu

tissim

um

0Dentic

ulaten

uis

020

40

Nitzsc

hiape

rminu

ta

020

40

Nitzsc

hiafru

stulum

0Rossit

hidium

peter

senii

Fi

gure

20.

0.25

1.25

2.25

3.25

4.25

5.25

Depth (cm)

0Navicu

lavu

lpina

0Achna

nthidi

umminu

tissim

um0Enc

yono

psis

microc

ephe

la0Cym

bella

botel

lus0

2040

6080

Rel

ativ

e ab

unda

nce

(%)

Nitzsc

hiafru

stulum

0Cymbe

llacle

ve-eu

lerae

020

Nitzsc

hiape

rminu

ta

0Dentic

ulaku

etzing

ii

0Cymbo

pleura

angu

stata

varsp

itzbe

rgens

is0.

25

1.25

2.25

3.25

4.25

5.25

Depth (cm)

0Navicu

lavu

lpina

0Achna

nthidi

umminu

tissim

um0Enc

yono

psis

microc

ephe

la0Cym

bella

botel

lus0

2040

6080

Rel

ativ

e ab

unda

nce

(%)

Nitzsc

hiafru

stulum

0Cymbe

llacle

ve-eu

lerae

020

Nitzsc

hiape

rminu

ta

0Dentic

ulaku

etzing

ii

0Cymbo

pleura

angu

stata

varsp

itzbe

rgens

is

Fi

gure

21.

Tabl

e 1.

Sum

mar

y of

lim

nolo

gica

l dat

a fr

om e

ach

pond

cor

ed a

t Cap

e V

era,

Dev

on Is

land

.

Site

Sa

mpl

e co

de

δ15N

%

C

pH

Con

d C

HL

a Si

O2

DO

C

DIC

PO

C

PON

T

dN

TPu

T

Pf

N:P

N

or

P

Dat

e

(‰)

%

(μ

S/cm

) µg

/L

mg/

L

mg/

L

mg/

L

mg/

L

mg/

L

mg/

L

mg/

L

mg/

L

mol

ar

limite

d

CV

-5

02/0

7/20

05

5-05

9.

36

10.6

4 6.

95

116

0.9

0.25

0.

9 14

.6

0.34

2 0.

013

0.11

4 0.

0094

0.

0057

29

.0

eith

er

CV

-6

03/0

7/20

05

6-05

9.

13

15.4

4 7.

5 14

6.5

0.5

0.31

1.

6 19

.5

0.28

3 0.

017

0.17

6 0.

0056

0.

0041

66

.2

p C

V-7

09

/07/

2005

7-

05

10.5

8 0.

50

7.95

13

4.9

10.8

0.

43

1.4

17.1

0.

742

0.05

1 0.

211

0.02

69

0.00

52

20.3

ei

ther

C

V-9

09

/07/

2005

9-

05

15.9

5 20

.57

8.95

12

1 1

0.42

1.

6 13

.2

0.32

3 0.

024

0.29

3 0.

0353

0.

0286

17

.9

n

CV

-9a

09/0

7/20

05

9a-0

5 13

.02

42.1

4 9.

75

126.

5 1.

9 0.

36

3.9

11.5

5.

62

0.66

6 0.

448

0.12

4 0.

072

19.6

n

CV

-12

02/0

7/20

05

12-0

5 3.

30

16.5

3 7.

65

312.

5 2

2.49

4.

1 21

.6

0.88

8 0.

1 0.

431

0.02

61

0.01

34

46.0

ei

ther

C

V-1

3 10

/07/

2006

13

-06

19.3

8 25

.68

10.0

0 23

0.6

0.5

2.61

5.

7 18

.3

0.32

0.

045

0.90

9 0.

0488

0.

0353

39

.7

eith

er

CV

-20

09/0

7/20

05

20-0

5 7.

60

35.7

3 10

.05

173

0.9

0.81

6.

2 12

.5

0.47

9 0.

049

0.62

8 0.

0371

0.

0289

41

.9

eith

er

CV

-22

10/0

7/20

05

22-0

5 1.

34

16.4

1 7.

55

125.

4 0.

9 0.

31

1.8

10.9

1.

21

0.11

7 0.

261

0.03

59

0.00

29

23.6

ei

ther

C

V-2

4 10

/07/

2005

24

-05

1.62

35

.75

8 12

7.3

0.6

0.25

1.

5 15

.7

0.27

0.

013

0.16

2 0.

0096

0.

0107

40

.6

eith

er

Mea

n

9.

13

21.9

4 8.

44

161.

37

2.00

0.

82

2.87

15

.49

1.05

0.

11

0.36

0.

04

0.02

34

.49

eith

er

Med

ian

9.25

18

.55

7.98

13

1.10

0.

90

0.39

1.

70

15.1

5 0.

41

0.05

0.

28

0.03

0.

01

34.3

5 ei

ther

M

ax

19.3

8 42

.14

10.0

5 31

2.50

10

.80

2.61

6.

20

21.6

0 5.

62

0.67

0.

91

0.12

0.

07

66.2

0 p

Min

1.

34

0.50

6.

95

116.

00

0.50

0.

25

0.90

10

.90

0.27

0.

01

0.11

0.

01

0.00

17

.91

n

CHAPTER 9

GENERAL DISCUSSION AND CONCLUSIONS

The heightened sensitivity of Arctic ecosystems to environmental change has been well

documented (ACIA 2004; Schindler & Smol 2006). Climate-related changes have already

resulted in major ecosystem consequences in Canadian High Arctic freshwaters, including

cascading trophic effects (Smol et al. 2005), as well as the complete desiccation of ponds (Smol

& Douglas 2007). High Arctic ecosystems, therefore, are key reference areas for studies of

global environmental change. The appropriate use of reference areas, however, requires a

thorough understanding of natural environmental variability and baseline conditions. Due to the

abundance of lakes and ponds in the Canadian High Arctic, paleolimnological techniques, based

on sound limnological data, have the potential to reconstruct historical patterns of environmental

change.

Modern limnology

Physical and chemical data from over 400 lakes and ponds have highlighted the

limnological variability that currently exists in the Canadian High Arctic and underscored the

need for more data from under-represented ecosystem types. This thesis has contributed new

limnological information from some of these under-studied areas, including the western High

Arctic (Chapter 2), High Arctic oases (Chapter 4), and from freshwater systems influenced by

seabird-derived nutrients (Chapter 7).

In general, the studies in this thesis showed that lakes and ponds located in more lushly

vegetated regions were often characterized by higher dissolved organic carbon (DOC), nutrients,

pH and specific conductivity (Chapters 2 and 4). The relationship between these environmental

variables and lush vegetation had previously been noted, albeit on a smaller scale (Antoniades et

al. 2003; Lim et al. 2005). The differences in environmental variables are likely due, at least in

part, to the underlying reasons for the lush vegetation, whereby microclimatic and topographic

features have resulted in conditions more amenable for plant growth (i.e. availability of water,

richer soils; Edlund & Alt 1989), which in turn leads to increased DOC and nutrient transport

from the catchment. Nutrient enrichment will increase photosynthesis where nutrients limit

primary production (Bonilla et al. 2005), and greater rates of photosynthesis ultimately result in

higher pH. DOC may also increase production by stimulating the microbial loop (i.e. Vincent &

Hobbie 2000), and by providing UV protection for phytoplankton and benthic microbial mats

(Rae & Vincent 1998). Collectively, the freshwater ecosystems found in the more lushly

vegetated regions had notably different limnological characteristics relative to other High Arctic

sites.

On the other hand, the Cape Vera sites, although lushly vegetated and replete with

nutrients, did not have particularly high DOC (Chapter 7). We suggest this may be due to the role

of allochthonous (seabird-derived) nutrients permitting the development of greater terrestrial

production in the absence of a favourable climate. As DOC is derived from the decomposition of

terrestrial organic matter, a slower rate of decomposition related to particularly harsh climatic

conditions would result in less DOC export to freshwater systems than would be otherwise

expected. Furthermore, the type of vegetation may also play a role in DOC cycling; for example,

Cape Vera was characterized by many mosses, yet few grasses and sedges. Mosses in Arctic

lakes have been shown to withstand cold temperatures and to decompose very slowly (Sand-

Jensen et al. 1999).

Previous Arctic limnological surveys (e.g. Lim et al. 2001; Michelutti et al. 2002;

Antoniades et al. 2003) have attributed the lack of a relationship between nutrients and

cholorphyll a measured in the water columns of High Arctic lakes and ponds to the dominance of

benthic production (Vezina & Vincent 1997; Villeneuve et al. 2001; Bonilla et al. 2005). Similar

to these previous studies, the water chemistry data from Melville Island (Chapter 2), northern

Ellesmere Island (Chapter 4), and Cape Vera (Chapter 7) also showed no clear relationship

between nutrients and chlorophyll a concentrations and that is likely attributable to similar factors

(i.e. the dominance of benthic production that is not captured by cholophyll a concentrations in

the water column). However, at Cape Vera, it is clear that several ponds are highly productive,

with luxuriant strands of filamentous algae and cyanobacteria, as well as relatively dense

populations of invertebrates (e.g. chironomids and cladocerans), so this lack of relationship

between measured nutrients and chlorophyll a is slightly more surprising. Due to a lack of upper

trophic level predators (e.g. fish), the Cape Vera ponds might present an ideal site to examine the

influence of grazing pressure on primary production in a High Arctic context, as has been found

in studies of Alaskan lakes (e.g. Hobbie et al. 1999).

Diatom ecology

Chapters 3 and 7 provided the first data on diatom species distributions across a range of

ecozones on Melville Island and from a suite of ponds influenced by seabird-derived nutrients at

Cape Vera, Devon Island, respectively. While diatom assemblages from all samples in this thesis

contained species that have previously been identified from High Arctic ponds and lakes

(Antoniades et al. in press), patterns of relative abundance were quite distinct. On Melville

Island, for example, similar taxa were found in ponds located in all four bioclimatic zones, yet the

relative proportions of the dominant taxa were different, leading to significant differences

between the most lushly vegetated bioclimatic zone and all other zones (Chapter 3). An

interesting conclusion drawn from this study is that spatial variables appear relatively

unimportant in structuring diatom species distributions, at least at the regional scale of Melville

Island.

Chapter 7 presents an examination of diatom species data from ponds across a gradient of

seabird influence. Significant relationships between diatom species distributions and seabird-

derived nutrients (inferred by δ15N) were identified. However, the high abundance of very few,

overwhelmingly dominant, diatom taxa in nearly all sites suggested that seabird-derived nutrients

are not the most important factor directly determining the most abundant diatom assemblages in

these ponds. The data from Chapter 7 add to the existing literature of other diatom-based studies

of nutrient enrichment (from sewage inputs and whale carcasses) in Arctic lakes and ponds

(Douglas & Smol 2000; Douglas et al. 2004; Michelutti et al. in press); together these data

indicate highly complex relationships between diatoms and nutrient enrichment. Interestingly,

our data contrast those of other studies that have suggested anthropogenic nitrogen deposition

may influence diatom assemblage shifts in Arctic lakes and ponds (Wolfe et al. 2006).

Paleolimnology

Chapters 5, 6, and 8 each present a different application of diatom-based paleolimnology.

The timing of environmental change recorded by diatoms from paleolimnological studies has

been attributed, in part, to the hypothesis that ice-cover is an important factor influencing diatom

assemblage shifts in High Arctic lakes and ponds (Smol 1983; Smol 1988; Douglas & Smol

1999). In Chapter 6, the difference in ice duration between two small, adjacent, and

limnologically similar lakes from northern Ellesmere was used as a test of the ice cover

hypothesis. The marked diatom changes in the lake with less ice cover, combined with the lack

of diatom change in the lake with more ice cover, strongly support the idea that the duration and

extent of ice cover impact the magnitude of the diatom response to environmental change

(Chapter 6).

The diatom record from a small pond on Melville Island provides the only

paleoenvironmental data available from this large western High Arctic island, and indicates

marked environmental change, consistent with climate warming, beginning in the early 20th

century (Chapter 5). The results from Chapter 5 are consistent with those from two small mid-

Arctic ponds on the more southerly Banks Island (Lim 2004). As small ponds are expected to be

more sensitive to environmental change than larger lakes (Smol et al. 2005), together these two

western Arctic diatom-based paleolimnological studies suggest a later onset of environmental

change as compared to small ponds from the eastern Canadian Arctic (Douglas et al. 1994;

Antoniades et al. 2005).

The diatom-based paleolimnological studies of ten ponds across a gradient of seabird

influence at Cape Vera, Devon Island, represent the first paleolimnological assessments of

seabird-influence in Arctic regions (Chapter 8). The diatom assemblage changes from the

sediment cores represent highly complex records, only some of which are consistent with other

proxies of seabird-influence. The conclusions drawn from this study suggest that changes in

diatom assemblages are not robust indicators of seabird-derived nutrients to High Arctic ponds;

rather, proxies tracking trophic sources of nutrients (i.e. stable isotopes) and estimates of overall

production (i.e. chlorophyll a) are likely more effective indicators. However, adequate

chronological control is of the utmost importance when using proxies based on concentration data

such as these.

Future directions

Much progress has been made in High Arctic limnology over the past several decades,

and the field is now at a very exciting point. While the continued collection and surveying of

modern limnological data would certainly supplement and therefore improve our understanding

of Arctic limnology, several other avenues of future limnological and paleolimnological research

also deserve attention.

With the advent of the excellent and much needed guide to High Arctic diatom taxonomy

and autecology provided by Antoniades et al. (in press), the seeds for a much more thorough

understanding of diatom autecology have been sown. Including the data from this thesis, both

limnological and diatom ecological information have been collected, using identical techniques,

from over 400 lakes and ponds from ten of the largest islands in the Canadian Arctic Archipelago.

Although complex taxonomic issues (i.e. splitting and grouping, revised names) make the

compilation of these data a daunting task, the resulting data would do much to further refine

diatom-based limnological and paleolimnological research in the High Arctic. Moreover, this

“super-calibration set” would rival any in the world by its size, scope, and, perhaps most

importantly, methodological consistency.

Furthering our understanding of Arctic limnology is critical, in light of the importance of

Arctic freshwater ponds and lakes as “hotspots of biodiversity”, as well as their sensitivity to

environmental change. Unlike some terrestrial ecosystems (Henry & Molau 1997), long-term

experimental data on the effects of warming in Arctic freshwater lakes are completely lacking

(ACIA 2005). Experimental data of this type would be particularly useful in High Arctic lakes

and ponds, where initial, shorter term studies have identified distinctive relationships relative to

temperate regions.

Perhaps the most intriguing research question arising from this thesis is that of the

complex relationship between diatom assemblages and nutrient enrichment in High Arctic lakes

and ponds: If other primary producers respond to nutrients, why is there no clear species

assemblage shift in diatom communities? Questions aimed at teasing apart the diatom-nutrient

relationship could be addressed in a series of controlled experiments in High Arctic ponds. Some

avenues that would be worth following include: a) Do diatom communities (composition or

biomass) change in response to known nutrient enrichment? b) Does the diatom response depend

on interactions with temperature and/or grazers? c) Is the diatom response, or lack thereof,

ecologically meaningful with respect to Arctic food webs? Such data would be useful in more

fully understanding the aquatic ecosystems of the High Arctic, and ultimately lead to a better

appreciation of the threats facing these sentinels of environmental change.

References

ACIA, 2004. Impacts of a warming Arctic. Cambridge University Press, Cambridge, UK.

ACIA, 2005. Arctic climate impact assessment. Cambridge University Press, New York.

Antoniades, D., M. S. V. Douglas, & J. P. Smol, 2003. Comparative physical and chemical limnology of two Canadian High Arctic regions: Alert (Ellesmere Island, NU) and Mould Bay (Prince Patrick Island, NWT). Archiv für Hydrobiologie 158: 485-516.

Antoniades, D., M. S. V. Douglas, & J. P. Smol, 2005. Quantitative estimates of recent environmental changes in the Canadian High Arctic inferred from diatoms in lake and pond sediments. Journal of Paleolimnology 33: 349-360.

Antoniades, D., P. B. Hamilton, M. S. V. Douglas & J. P. Smol, in press. Freshwater diatoms from the Canadian High Arctic. Iconographica Diatomologica.

Bonilla, S., V. Villeneuve, & W. F. Vincent, 2005. Benthic and planktonic algal communities in a High Arctic Lake: Pigment structure and contrasting responses to nutrient enrichment. Journal of Phycology 41: 1120-1130.

Douglas, M. S. V.& J. P. Smol, 1999. Freshwater diatoms as indicators of environmental change in the High Arctic. In E. F. Stoermer & J. P. Smol (eds), The Diatoms: Applications for the Environmental and Earth Sciences. Cambridge University Press, Cambridge: 227-244.

Douglas, M. S. V., & J. P. Smol, 2000. Eutrophication and recovery in the High Arctic: Meretta lake (Cornwallis Island, Nunavut, Canada) revisited. Hydrobiologia 431: 193-204.

Douglas, M. S. V., J. P. Smol, & W. Blake Jr, 1994. Marked post-18th century environmental change in high-arctic ecosystems. Science 266: 416-419.

Douglas, M. S. V., J. P. Smol, J. M. Savelle, & J. M. Blais, 2004. Prehistoric Inuit whalers affected arctic freshwater ecosystems. Proceedings of the National Academy of Sciences of the United States of America 101: 1613-1617.

Edlund, S. A., & B. T. Alt, 1989. Regional congruence of vegetation and summer climate patterns in the Queen Elizabeth Islands, Northwest Territories, Canada. Arctic 42: 3-23.

Henry, G. H. R., & U. Molau, 1997. Tundra plants and climate change: the International Tundra Experiment (ITEX). Global Change Biology 3: 1-9.

Hobbie, J. E., B. J. Petersen, N. Bettez, L. Deegan, W. J. O’Brien, G. W. Kling, G. W. Kipphut, W. B. Bowden, & A. E. Hershey, 1999. Impact of global change on the biogeochemistry of an Arctic freshwater system. Polar Research 18: 207-214.

Lim, D.S.S., 2004. Limnology and diatom paleoecology of diatoms on Banks Island, N.W.T., and Devon Island, Nunavut, Canadian Arctic. University of Toronto. PhD Thesis.

Lim, D. S. S., M. S. V. Douglas, & J. P. Smol, 2005. Limnology of 46 lakes and ponds on Banks Island, NWT, Canadian Arctic Archipelago. Hydrobiologia 545: 11-32.

Lim, D. S. S., M. S. V. Douglas, J. P. Smol, & D. R. S. Lean, 2001. Physical and chemical limnological characteristics of 38 lakes and ponds on Bathurst Island, Nunavut, Canadian High Arctic. International Review of Hydrobiology 86: 1-22.

Michelutti, N., M. S. V. Douglas, D. C. G. Muir, X. Wang, & J. P. Smol, 2002. Limnological characteristics of 38 lakes and ponds on Axel Heiberg Island, High Arctic Canada. International Review of Hydrobiology 87: 385-399.

Michelutti, N., M. H. Hermanson, J. P. Smol, P. J. Dillon, & M. S. V. Douglas, in press. Delayed response of diatom assemblage changes to sewage inputs in an Arctic lake. Aquatic Sciences.

Rae, R., & W. F. Vincent, 1998. Effects of temperature and ultraviolet radiation on microbial foodweb structure: potential responses to global change. Freshwater Biology 40: 747-758.

Sand-Jensen, K., T. Riis, S. Markager, & W. F. Vincent, 1999. Slow growth and decomposition of mosses in Arctic lakes. Canadian Journal of Fisheries and Aquatic Sciences 56: 388-393.

Schindler, D. W., & J. P. Smol, 2006. Cumulative effects of climate warming and other human activities on freshwaters of Arctic and subarctic North America. Ambio 35: 160-168.

Smol, J. P., 1983. Paleophycology of a high arctic lake Cape Herschel, Ellesmere Island. Canadian Journal of Botany 61: 2195-2204.

Smol, J. P., 1988. Paleoclimate proxy data from freshwater arctic diatoms. Verh. int. Ver. Limnol. 23: 837-844.

Smol, J. P., & M. S. V. Douglas, 2007. Crossing the final ecological threshold in high Arctic ponds. Proceedings of the National Academy of Sciences of the United States of America 104: 12395-12397.

Smol, J. P., A. P. Wolfe, H. J. B. Birks, M. S. V. Douglas, V. J. Jones, A. Korhola, R. Pienitz, K. Ruhland, S. Sorvari, D. Antoniades, S. J. Brooks, M. A. Fallu, M. Hughes, B. E. Keatley, T. E. Laing, N. Michelutti, L. Nazarova, M. Nyman, A. M. Paterson, B. Perren, R. Quinlan, M. Rautio, E. Saulnier-Talbot, S. Siitoneni, N. Solovieva, & J. Weckstrom, 2005. Climate-driven regime shifts in the biological communities of arctic lakes. Proceedings of the National Academy of Sciences of the United States of America 102: 4397-4402.

Vezina, S., & W. F. Vincent, 1997. Arctic cyanobacteria and limnological properties of their environment: Bylot Island, Northwest Territories, Canada (73 degrees N, 80 degrees W). Polar Biology 17: 523-534.

Villeneuve, V., W. F. Vincent, & J. Komárek, 2001. Community structure and microhabitat characteristics of cyanobacterial mats in an extreme high Arctic environment: Ward Hunt Lake. Nova Hedwigia 123: 199-224.

Vincent, W. F., and J. E. Hobbie, 2000. Ecology of Arctic lakes and rivers. In: M. Nuttal, and T. V. Callaghan (eds), The Arctic: Environment, People, Policy. Harwood, Amsterdam: 197-231.

Wolfe, A. P., C. A. Cooke, & W. O. Hobbs, 2006. Are current rates of atmospheric nitrogen deposition influencing lakes in the Eastern Canadian Arctic? Arctic Antarctic and Alpine Research 38: 465-476.

APPENDICES

Appendix 1. Detailed water chemistry methods used for all water sampling and temperature measurement during field work. Note that unless otherwise noted, all water samples were taken from the nearshore area where the average water depth was ~70 cm. All bottles were rinsed three times with lakewater prior to filling. Air temperature: Air temperatures were recorded twice daily (approximately 08:00h and 20:00h) from a thermometer kept inside a Stevenson screen. Water temperature: Water temperatures were measured at the time of sampling for water chemistry using a hand-held alcohol thermometer held ~10 cm below the water surface. pH: Samples for pH were taken from the lake water using a Nalgene bottle, rinsed 3 times with sample water. pH was measured back at the base camp at the end of each sampling day. Water samples were kept in the dark in a backpack from time of sampling until time of measurement. We used three hand-held pHEP 4 pH meters (Hanna) to measure each sample. All pH meters were calibrated to 7 and 10 pH buffer solution daily. We used the average pH from the three meters as our pH value, although often all three meters gave the same reading. Specific conductivity: We measured specific conductivity at the base came each day, using a subsample of the same water obtained for pH measurement. Specific conductivity was measured using a YSI model 33 conductivity meter. Chlorophyll a: Samples for chlorophyll a were filtered onto glass fibre filters (Whatman GF/C; diameter = 4.25 cm) under low vacuum pressure. Filters were placed immediately in Petri dishes, wrapped in aluminum foil, and then kept in the dark and as cold as possible until return to PCSP. Once at PCSP, filters were kept frozen until analysis at CCIW, Burlington, ON. Filtered nutrients (including total phosphorus filtered, all nitrogen fractions), dissolved organic carbon and dissolved inorganic carbon: water was filtered through Sartorius-type cellulose acetate filters (Brand: AcetatePlus Supported, Plain, pore size: 0.45 µm, filter diameter 4.7 cm). Filters were placed immediately in Petri dishes, wrapped in aluminum foil, and then kept in the dark and as cold as possible until return to PCSP. Once at PCSP, filters were kept frozen until analysis at CCIW, Burlington, ON. Particulate carbon and nitrogen: sample water is filtered through pre-ignited glass fibre filters (Whatman GF/C 2.5 cm). Filters were placed immediately in Petri dishes, wrapped in aluminum foil, and then kept in the dark and as cold as possible until return to PCSP. Once at PCSP, filters were kept frozen until analysis at CCIW, Burlington, ON.

Appendix 2. Output tables from various spatial neighbourhood models tested using VARCAN (Peres-Neto et al. 2006, see Chapter 3 for more details). Only the most restricted neighbourhood model indicated that spatial variables explained a significant portion of the diatom species distribution. The number of spatial variables tested for each model is given in the heading. The environmental variables included in the model were the same for each: elevation, surface area, pH, specific conductivity, temperature, total phosphorus, total phosphorus filtered, total dissolved nitrogen, chlorophyll a, DOC, SiO2, SRP, Al.

Delauney: 12 spatial variables retained

% explained % exp - adjustment probabilityenv 0.315 0.098 0.024env x space 0.098 0.084space 0.211 -0.013 0.898residual 0.375 0.831

Gabriel graph: 17 spatial variables retained

% explained % exp - adjustment probabilityenv 0.311 0.127 0.018env x space 0.103 0.055space 0.325 0.042 0.391residual 0.261 0.775

Sphere of Influence: 16 spatial variables retained

% explained % exp - adjustment probabilityenv 0.318 0.138 0.006env x space 0.095 0.044space 0.315 0.056 0.182residual 0.271 0.761

Relative neighbours: 17 spatial variables retained

% explained % exp - adjustment probabilityenv 0.325 0.158 0.002env x space 0.088 0.024space 0.334 0.065residual 0.253 0.752 0.178

50 000m: 8 spatial variables retained

% explained % exp - adjustment probabilityenv 0.32 0.124 0.004env x space 0.094 0.058space 0.181 0.067residual 0.405 0.751 0.026

Appendix 3. Diatom species response scatterplots versus specific conductivity from Melville Island, Nunavut/NWT. Roman numerals next to species names indicate the best HOF model. See Chapter 3 for details on HOF models.

Psammothidium chlidanos (II)

0

5

10

15

20

25

30Eucocconeis flexella (IV)

0

2

4

6

8Eucocconeis laevis (IV)

0

2

4

6

8

Psammothidium marginulatum (II)

0

10

20

30

40

50

60

70Achnanthidium minutissimum (IV)

0

5

10

15

20

25Rossithidium petersenii (I)

0

2

4

6

8

10

12

14

Psammothidium scotica (II)

Rel

ativ

e ab

unda

nce

(%)

0

2

4

6

8

10Psammothidium ventralis (II)

0

2

4

6

8

10

12

14

16Amphora inariensis (I)

0

2

4

6

8

Chamaepinnularia soehrensis (V)

0 300 600 900 1200 1500

02468

101214161820

Cymbella cleve-eulerae (IV)

specific conductivity (μS/cm)

0 300 600 900 1200 1500

0

1

2

3

4

5Encyonema minutum (I)

0 300 600 900 1200 1500

0

2

4

6

8

10

Appendix 3. Continued.

Diadesmis contenta (IV)

0

5

10

15

20

25

30

35Fragilaria capucina (V)

0

5

10

15

20

25

30Navicula chiarae (IV)

0

2

4

6

8

10

12

Navicula cryptocephala (I)

Rel

ativ

e ab

unda

nce

(%)

0

2

4

6

8

10Navicula gerloffi (I)

0

5

10

15

20

25

30

35Nitzschia frustulum (V)

0

10

20

30

40

50

Nitzschia inconspicua (I)

0 300 600 900 1200 1500

0

2

4

6

8

10

12

14Nitzschia perminuta (V)

specific conductivity (μS/cm)

0 300 600 900 1200 1500

0

5

10

15

20

25

30

35Nitzschia pusilla (I)

0 300 600 900 1200 1500

0

1

2

3

4

Appendix 4. Summary statistics, figures, and species optima of various weighted averaging models for diatom-inferred specific conductivity from surface sediments of Melville Island. Table 1a) all species >1% relative abundance in three sites or >10% relative abundance in one site (n=90), b) all species with a significant response to conductivity (n=55), and c) all species with a significant unimodal response to specific conductivity (n=30). All values shown are cross-validated using bootstrapping techniques. “RMSEP” is the Root Mean Squared Error of Prediction.

a)

WAInv WACla WATol Inv WATol Cla

r2boot 0.138 0.152 0.257 0.272

Average bias boot 44.560 48.941 74.910 84.390 RMSEP 250.110 249.923 249.336 250.231 b)

WAInv WACla WATol Inv WATol Cla

r2boot 0.156 0.176 0.467 0.488

Average bias boot 51.300 54.500 68.600 74.600 RMSEP 249.100 248.500 230.800 228.100 c)

WAInv WACla WATol Inv WATol Cla

r2boot 0.189 0.210 0.325 0.349

Average bias boot 3.380 5.040 18.300 32.130 RMSEP 217.300 312.500 201.000 250.920

Appendix 4. Figure 1. Scatterplot of: A) estimated versus observed specific conductivity, and B) residuals based on the weighted averaging technique with inverse deshrinking and tolerance downweighting.

observed cond (µS/cm)

WA

toli

nv e

stim

ated

con

d(µ

S/c

m)

observed cond (µS/cm)

WA

toli

nv c

ond

resi

dual

s

0 200 400 600 800 1000 1200

0

200

400

600

800

1000

1200r 2

boot = 0.257

0 200 400 600 800 1000 1200-1200

-1000

-800

-600

-400

-200

0

200RMSEP = 249

A B

Appendix 4. Table 2. Estimated specific conductivity optima for diatom species from the surface sediments of Melville Island. Species codes are the same as for Table 2 (Keatley et al. [3]).

Species Cond. Species Cond. Species Cond. code optimum code optimum code optimum

1 151 34 219 67 57 2 194 35 159 68 128 3 50 36 145 69 139 4 245 37 101 70 282 5 32 38 49 71 29 6 64 39 59 72 327 7 154 40 95 73 310 8 93 41 119 74 195 9 172 42 305 75 219

10 69 43 225 76 122 11 158 44 378 77 82 12 121 45 63 78 111 13 30 46 n/a 79 176 14 32 47 55 80 205 15 48 48 643 81 153 16 57 49 228 82 434 17 124 50 145 83 43 18 167 51 7 84 105 19 168 52 6 85 56 20 634 53 87 86 91 21 16 54 54 87 60 22 232 55 77 88 77 23 230 56 65 89 71 24 72 57 n/a 90 20 25 111 58 n/a 26 209 59 211 27 n/a 60 214 28 22 61 130 29 278 62 110 30 186 63 96 31 142 64 66 32 121 65 n/a 33 184 66 846

App

endi

x 5.

Raw

dia

tom

cou

nts f

rom

Mel

ville

Isla

nd su

rfac

e se

dim

ents

. M

elvi

lle Is

land

cal

ibra

tion

diat

oms

SITE

NAM

E (M

V- )

AB

CD

EF

GH

IJ

KL

MN

OP

QR

ST

UV

WX

YZ

AAAB

ACAD

AEAF

roAG

AHAI

AJAK

ALAM

ANAO

APAQ

ARAS

ATAc

hnan

thes

alta

ica

2Ac

hnan

thes

bro

enlu

nden

sis

00

700

00

30

00

00

40

00

00

00

00

00

00

00

00

00

60

20

00

00

00

00

00

Achn

anth

es h

olst

ii0

00

00

00

00

00

00

00

00

00

00

00

04

60

00

30

00

00

02

00

00

00

20

0Ac

hnan

thes

ingr

atifo

rmis

00

00

00

00

00

00

00

00

00

00

480

1318

00

200

010

00

00

00

00

07

00

08

190

Achn

anth

es k

riege

ri0

629

56

03

010

10

130

60

30

03

00

00

00

10

00

120

300

00

00

20

00

00

00

0Ac

hnan

thes

kry

ophi

la0

00

00

03

320

010

012

100

00

03

011

00

00

00

00

00

30

00

00

00

00

2612

00

0Ac

hnan

thes

mar

ginu

lata

108

7336

1617

082

1635

134

704

2519

910

712

2726

340

525

436

74

22

168

240

60

410

09

40

306

133

213

9Ac

hnan

thes

min

utis

sim

a11

021

4511

346

6528

043

756

116

780

018

293

2431

00

915

814

87

011

415

00

00

00

510

460

00

037

4043

Achn

anth

es p

eter

senn

i27

613

3413

40

80

1023

3910

773

34

815

20

2018

00

130

413

00

00

00

40

00

01

111

1933

Achn

anth

es s

cotic

a0

00

00

07

010

034

00

36

00

220

00

043

054

560

00

00

00

00

360

00

00

00

449

0Ac

hnan

thes

chl

idan

os8

95

2910

912

414

1110

3140

141

768

3143

2617

633

7812

62

4312

465

8513

336

4Ac

hnan

thes

cur

tissi

ma

12

Achn

anth

es d

aone

nsis

40Ac

hnan

thes

del

icat

ula

sp s

epte

ntrio

nalis

2Ac

hnan

thes

dev

ilens

is1

22

1114

Achn

anth

es fl

exel

la8

1410

72

3117

2317

5655

21

16

104

22

321

93

Achn

anth

es h

intz

ii2

Achn

anth

es la

evis

106

212

641

146

120

1963

88

10

06

58

130

68

100

120

00

00

00

00

00

00

00

00

40

Achn

anth

es v

entr

alis

22

22

00

996

764

200

68

212

02

1214

200

3738

3910

180

021

00

02

00

00

04

00

6120

96

Achn

anth

es h

elve

ticum

00

00

00

210

431

40

02

20

00

00

00

100

20

00

00

101

00

00

00

40

00

00

50

0Ac

hnan

thes

bio

retti

04

00

00

00

20

035

53

00

05

52

60

00

130

290

00

20

00

00

00

00

00

02

10

Achn

anth

es ri

cula

2Ac

hnan

thes

rupe

stris

126

335

278

Achn

anth

es ro

ssi

419

7Ac

hnan

thes

sac

cula

4Ac

hnan

thes

sp1

MVA

S2

Achn

anth

es s

p2 M

VAS

1Ac

hnan

thes

sto

lida

var s

chm

assm

anii

6Ac

hnan

thes

sub

atom

oide

s10

66

22

325

48Ad

lafi

bryo

phila

527

82

219

132

44

12

112

Amph

ora

inar

iens

is2

1226

82

29

26

112

62

138

Amph

ora

copu

lata

51

21

810

44

84

12

Amph

ora

duse

nii

42

122

32

22

6Am

phor

a fo

gedi

i26

710

Amph

ora

foge

dian

a29

25Am

phor

a ho

lsta

ica

44

24Am

phor

a lib

yca

52

47

Amph

ora

pedi

culu

s9

82

Amph

ora

spitz

berg

ensi

s3

66

222

6612

4Am

phor

a ve

neta

22

2C

alon

eis

silic

ula

tota

l0

60

60

122

64

12

103

120

00

142

00

00

06

00

02

00

00

82

00

014

00

00

04

Cal

onei

s cf

big

2C

alon

eis

aero

phila

26

109

62

144

1217

30C

alon

eis

hend

eyii

426

Cal

onei

s pu

lcha

2C

alon

eis

sp M

VK4

Cal

onei

s sm

all

190

00

60

110

190

00

176

00

20

02

00

20

06

00

00

00

20

00

00

362

00

00

00

Cav

inul

a co

cone

iform

is3

22

62

Cav

inul

a ja

ernf

eldi

i2

4759

22

74

Cav

inul

a ps

eudo

scut

iform

is23

Cha

mae

pinn

ular

ia g

andr

upii

00

00

00

014

00

00

00

00

00

00

00

00

00

00

00

00

028

00

094

010

00

00

00 0

Cha

mea

pinn

ular

ia s

ohre

nsis

228

102

6136

116

720

680

1610

063

80

00

00

04

00

00

02

01

00

00

00

00

00

00

00

04

App

endi

x 5.

con

tinue

d.

Mel

ville

Isla

nd c

alib

ratio

n di

atom

sSI

TE N

AME

(MV-

)A

BC

DE

FG

HI

JK

LM

NO

PQ

RS

TU

VW

XY

ZAA

ABAC

ADAE

AF ro

AGAH

AIAJ

AKAL

AMAN

AOAP

AQAR

ASAT

Cyc

lote

lla a

ntiq

ua3

22

1C

yclo

tella

ato

mus

(C g

lom

erat

a or

C p

seud

oste

llige

ra?)

892

Cyc

lote

lla b

odan

ica

var a

ff le

man

ica

2C

yclo

tella

oce

llata

1C

yclo

tella

trip

artit

a17

Cyc

lote

lla a

ff co

men

sis

(or C

aff

com

ta v

ar u

nipu

ncta

ta)

59C

ymbe

lla a

ffini

s5

16C

ymbe

lla a

mph

icep

hala

72

26

66

1614

64

218

2C

ymbe

lla a

ngus

tata

52

1015

107

64

256

211

44

22

66

22

3029

4C

ymbe

lla c

esat

ii4

97

22

1216

612

67

2C

ymbe

lla b

otel

lus

160

08

42

92

260

15

190

00

720

97

00

00

00

00

80

00

00

00

00

00

00

00

00

Cym

bella

cle

ve-e

uler

ae23

138

1915

219

88

016

1820

54

216

148

46

00

40

00

09

00

01

00

00

02

00

00

04

0C

ymbe

lla d

escr

ipta

24

00

215

260

40

06

390

30

22

00

00

00

00

00

00

00

00

00

00

00

00

00

00

Cym

bella

des

igna

ta2

104

1041

104

82

42

48

2C

ymbe

lla fo

gedi

i12

00

1224

01

02

08

105

02

016

00

02

00

00

00

00

00

00

00

00

00

00

00

029

5C

ymbe

lla g

auem

anii

24

Cym

bella

late

ns0

00

00

09

100

140

02

04

00

05

182

08

03

02

04

00

00

00

00

00

00

00

1039

8C

ymbe

lla la

nge-

bert

alot

ti4

84

62

82

22

87

Cym

bella

lapp

onic

a (a

ngus

tata

?)4

746

44

Cym

bella

min

uta

140

02

10

00

20

52

166

517

320

924

10

619

190

261

435

00

02

02

00

44

00

04

220

Cym

bella

obs

cura

22

Cym

bella

pro

xim

a3

Cym

bella

sile

siac

a9

48

57

47

04

04

12

12

160

26

20

00

134

313

017

170

00

20

05

04

00

00

03

0C

ymbe

lla s

p de

rmot

pla

te 3

5 #2

62

46

Cym

bella

sub

aequ

alis

1313

22

1110

112

44

22

2C

ymbe

lla s

ubsp

icul

a2

2C

ymbo

pleu

ra c

uspi

data

2C

ymbo

pleu

ra s

taur

onei

form

is2

22

22

16D

entic

ula

kuet

zing

ii21

30

244

1110

025

02

1129

00

06

00

90

00

00

00

00

00

00

00

00

02

00

00

00

0D

entic

ula

subt

ilis

15D

entic

ula

tenu

is0

00

69

00

00

00

00

00

00

00

00

00

00

00

00

00

00

00

00

092

00

00

00

0D

iade

smis

con

tent

a5

526

2713

923

362

1269

9874

558

7221

92

452

104

24

461

Dia

desm

is g

allic

a13

2D

iade

smis

par

acon

tent

a1

229

6D

iade

smis

per

pusi

lla2

22

66

2D

iato

ma

mon

ilifo

rmis

104

22

32

14D

iato

ma

tenu

is8

40

02

00

07

011

00

64

00

250

61

00

72

02

308

50

00

00

10

014

00

00

00

00

Dip

lone

is m

arge

nies

tria

ta6

62

25

912

1910

66

122

210

41

65

Dip

lone

is o

ccul

ata

124

24

1215

Dip

lone

is p

arm

a4

21

22

24

22

2En

cyon

ema

cf a

ff su

belg

inen

se2

Ency

onem

a sp

pla

te 3

5 #1

8-24

62

Ency

onem

a ve

ntric

osum

22

25Eu

notia

arc

us2

2Eu

notia

bilu

naris

24

22

62

1813

Euno

tia b

iluna

ris v

ar m

ucop

hila

2Eu

notia

bor

eote

nuis

00

00

00

00

08

00

00

00

00

00

00

00

00

00

00

4321

00

010

00

00

00

736

00

0Eu

notia

exi

gua

231

444

62

Euno

tia e

xigu

a fo

rmae

bid

ens

Hus

tedt

2Eu

notia

faba

3

Euno

tia fa

llax

var g

roen

land

ica

2Eu

notia

pra

erup

ta v

ar b

igib

ba4

Euno

tia p

raer

upta

40

04

20

00

85

214

140

011

00

30

00

20

226

40

12

212

00

00

00

02

07

10

06

Euno

tia p

seud

opec

tinal

is6

Euno

tia s

ubar

cuat

oide

s4

Falla

ceae

inso

ciab

ilis

21

Falla

ceae

pyg

mac

eae

43

Falla

ceae

sub

ham

ulat

a2

2Fa

llace

ae s

ublu

cidu

la1

App

endi

x 5.

con

tinue

d.

Mel

ville

Isla

nd c

alib

ratio

n di

atom

sSI

TE N

AME

(MV-

)A

BC

DE

FG

HI

JK

LM

NO

PQ

RS

TU

VW

XY

ZAA

ABAC

ADAE

AF ro

AGAH

AIAJ

AKAL

AMAN

AOAP

AQAR

ASAT

Frag

ilaria

alp

etris

2Fr

agila

ria c

apuc

ina

140

00

180

650

457

158

820

1035

4016

453

100

100

84

600

490

7226

032

140

00

00

02

120

316

515

Frag

ilaria

cap

ucin

a va

r aus

tria

ca24

30Fr

agila

ria c

apuc

ina

var g

raci

lis2

86

25

614

218

22

154

12Fr

agila

ria c

onst

ruen

s va

r ven

ter

212

Frag

ilaria

pin

nata

var

ven

ter(

?)14

6Fr

agila

ria p

inna

ta0

00

00

00

00

00

00

042

030

60

20

00

230

00

20

015

00

00

60

00

00

00

40

500

Frus

tulia

cra

ssin

ervi

a2

2G

eiss

leria

sch

oenf

eldi

i4

54

211

2G

eiss

leria

sp1

(MVN

)2

Gei

ssle

ria s

p D

erm

ot p

late

50

#19-

2311

4G

eiss

leria

sp

MVE

4G

eiss

leria

spe

c Fi

g16

plat

e 20

pag

e 14

86

Gom

phon

ema

angu

stat

um2

Gom

phon

ema

lage

rhei

mi

24

Gom

phon

ema

prod

uctu

m2

Gom

phon

ema

spp

aff e

xigu

um11

1G

omph

onem

a sp

aff

obsc

urum

16G

omph

onem

a sp

MVN

5G

omph

onem

a sp

MVA

O80

Han

tzsc

hia

amph

ioxy

s3

22

Hip

podo

ntic

a ar

ctic

a4

Lutic

ola

mut

ica

4Lu

ticol

a ni

valo

ides

2M

icro

cost

atus

kra

sske

i4

2M

ulle

ria g

ibbu

la2

Nav

icul

a ba

cillu

m6

42

22

Nav

icul

a cf

egr

egria

8N

avic

ula

cf rh

ynco

ceph

ala

1017

Nav

icul

a cf

sim

ilis

2N

avic

ula

chia

rae

220

40

3877

427

230

2517

4824

00

170

410

00

00

00

00

50

00

08

00

00

360

00

04

102

Nav

icul

a ch

iara

e/cr

ypto

ceph

ala

6647

Nav

icul

a ci

ncta

11

263

Nav

icul

a cl

emen

tis2

1N

avic

ula

com

plan

ata

1N

avic

ula

cryp

toce

phal

a2

217

420

231

88

743

132

536

225

1420

612

189

42

3620

Nav

icul

a di

gito

riadi

ata

4N

avic

ula

dolo

miti

ca (o

r Ach

nant

hes

dist

inct

a)6

Nav

icul

a eg

regr

ia8

214

1929

Nav

icul

a fr

agilo

ides

(?) s

mal

l, sk

inn y

23N

avic

ula

gerlo

ffi3

514

55

58

24

220

219

39

42

135

246

10N

avic

ula

genu

stria

ta6

Nav

icul

a gr

egar

ia3

116

1411

1416

62

Nav

icul

a hi

lliar

dii

102

42

4N

avic

ula

hilli

ardi

i und

ulat

e2

Nav

icul

a in

grat

a36

Nav

icul

a m

enis

culu

s6

234

Nav

icul

a m

utic

oide

s2

Nav

icul

a ph

ylle

pta

7322

0

App

endi

x 5.

con

tinue

d.

Mel

ville

Isla

nd c

alib

ratio

n di

atom

sSI

TE N

AME

(MV-

)A

BC

DE

FG

HI

JK

LM

NO

PQ

RS

TU

VW

XY

ZAA

ABAC

ADAE

AF ro

AGAH

AIAJ

AKAL

AMAN

AOAP

AQAR

ASAT

Nav

icul

a ps

eudo

scut

iform

is2

463

3315

339

72

321

3N

avic

ula

pseu

dote

nello

ides

(An

toni

ades

pla

te 4

9 #1

8)2

447

10N

avic

ula

pupu

la4

210

52

23

24

42

215

210

5N

avic

ula

pygm

aea

2N

avic

ula

rein

hard

ii15

3N

avic

ula

rhyn

coce

phal

a (d

erm

ot)

22

2N

avic

ula

salin

arum

62

Nav

icul

a cf

sal

inar

um b

ig6

Nav

icul

a sc

hmas

sman

ii19

Nav

icul

a sc

hmas

sman

ii bu

t ver

y sm

all a

nd n

ot a

s fa

t10

Nav

icul

a sm

all g

irdle

s M

VQ8

Nav

icul

a sp

ecie

s 1(

MVF

)4

Nav

icul

a sp

ecie

s M

VS2

Nav

icul

a sp

p48

DER

MO

T4

Nav

icul

a sp

34 p

late

51

derm

ot1

2N

avic

ula

sp15

pla

te 4

9 de

rmot

2N

avic

ula

sp24

-26

plat

e 49

der

mot

166

Nav

icul

a sp

18-2

3 pl

ate

49 d

erm

ot14

2N

avic

ula

spA

MVI

4N

avic

ula

spM

VAK

18N

avic

ula

spM

VD5

1N

avic

ula

spp

aff r

iedi

ana

25

25

Nav

icul

a sp

p af

f rei

nhar

dtii

92

2N

avic

ula

spp

aff v

enet

a4

63

94

29

616

217

Nav

icul

a sp

p af

f ven

erab

ilis

62

Nav

icul

a st

roem

ii5

2N

avic

ula

subh

amul

ata

6N

avic

ula

triv

alis

92

12

44

26

Nav

icul

a tu

scul

a23

27

12

22

22

22

102

Nav

icul

a vi

ridul

a va

r. lin

earis

2N

avic

ula

vulp

ina

1510

82

34

812

12

96

116

261

Nei

diop

sis

wul

fii2

83

2N

eidi

um a

ffine

14

26

182

84

16N

eidi

um a

ffine

var

long

icep

s2

13N

eidi

um a

mpl

iatu

m2

12

21

24

12

2N

eidi

um a

utsr

ium

3N

eidi

um b

ergi

i6

21

22

24

24

46

21

2N

eidi

um b

isul

catu

m2

Nei

dium

cal

vum

3N

eidi

um d

ecor

atum

19

Nei

dium

dec

orat

um v

ar b

ergi

i2

Nei

dium

dis

tinct

e-pu

ncta

tum

22

32

1N

eidi

um d

ubiu

m4

6N

eidi

um k

ozlo

wii

02

00

00

00

00

00

48

00

00

02

00

00

00

00

00

00

00

00

00

20

00

00

0N

eidi

um s

p1 M

VF1

0N

eidi

um s

p9 p

late

56

derm

ot2

Nei

dium

tem

perii

MVD

2N

itzsc

hia

amph

ibia

3

32

710

2N

itzsc

hia

accu

min

atum

8N

itzsc

hia

arch

ibal

di4

602

Nitz

schi

a cl

ausi

i3

23

89

1545

Nitz

schi

a co

mm

unta

ta2

22

14

104

Nitz

schi

a di

ssip

ata

var m

edia

63

84

44

115

22

Nitz

schi

a fr

ustu

lum

2211

1155

90

4811

170

3613

482

011

50

5414

80

50

318

20

30

167

00

218

00

02

305

00

04

019

18N

itzsc

hia

perm

inut

a35

5039

123

158

1718

610

244

3712

578

4824

380

3492

075

280

7076

120

3066

111

165

00

289

00

132

428

5826

00

6591

71

App

endi

x 5.

con

tinue

d.

Mel

ville

Isla

nd c

alib

ratio

n di

atom

sSI

TE N

AME

(MV-

)A

BC

DE

FG

HI

JK

LM

NO

PQ

RS

TU

VW

XY

ZAA

ABAC

ADAE

AF ro

AGAH

AIAJ

AKAL

AMAN

AOAP

AQAR

ASAT

Nitz

schi

a gr

acili

s4

10N

itzsc

hia

hom

burg

iens

is2

13

21

23

32

210

3846

1N

itzsc

hia

inco

nspi

cua

93

826

142

202

22

871

152

72

8112

22

6N

itzsc

hia

litto

ralis

6N

itzsc

hia

norm

anii

22

Nitz

schi

a ov

alis

1N

itzsc

hia

pale

a2

22

15N

itzsc

hia

pala

ecea

26

44

94

1912

Nitz

schi

a pa

laef

orm

is14

Nitz

schi

a pu

ra/p

alea

612

99

249

144

148

43

14

131

15

Nitz

schi

a pu

silla

87

1512

144

138

182

1215

1022

217

44

Nitz

schi

a re

cta

21

Nitz

schi

a su

blin

earis

2

172

44

44

419

Nitz

schi

a su

chla

ndii

148

210

276

7N

itzsc

hia

sp c

f aci

cula

ris6

Pinn

ular

ia a

ngus

tibor

ealis

2Pi

nnul

aria

bal

four

iana

247

37

1117

8Pi

nnul

aria

bic

eps

1Pi

nnul

aria

bor

ealis

var

lanc

eola

tat

22

Pinn

ular

ia b

ottn

ica

(lund

ii va

r bal

tica)

(gre

y bo

ok)

4Pi

nnul

aria

bra

ndel

iform

is2

22

Pinn

ular

ia b

rebi

sonn

i 2

22

31

4Pi

nnul

aria

bre

biso

nni v

ar b

icun

eata

2Pi

nnul

aria

bre

biso

nni s

ensu

lato

2Pi

nnul

aria

bre

vico

stria

ta2

Pinn

ular

ia c

f ang

lica

4Pi

nnul

aria

des

cres

cens

var

igno

rata

22

Pinn

ular

ia d

iver

gens

var

div

erge

ns18

Pinn

ular

ia fi

g 4

Plat

e61

2Pi

nnul

aria

spM

VT93

22

2Pi

nnul

aria

gru

now

ii2

22

102

2Pi

nnul

aria

inte

rmed

ia3

47

132

55

Pinn

ular

ia in

terr

upta

2Pi

nnul

aria

kra

mm

eri

10

20

47

02

60

30

52

90

00

610

00

65

43

162

20

40

03

05

52

23

02

20

3Pi

nnul

aria

lata

23

10Pi

nnul

aria

cf i

sost

auro

n va

r con

ifera

14Pi

nnul

aria

cf i

sost

auro

n va

r orie

ntal

is (p

142

Gre

y pi

nnul

aria

boo

k)13

35Pi

nnul

aria

cf m

inut

a6

Pinn

ular

ia m

icro

stau

ron

var b

rebi

sonn

i2

Pinn

ular

ia c

f mic

rost

auro

n4

Pinn

ular

ia o

bscu

ra2

4Pi

nnul

aria

str

epto

raph

e1

Pinn

ular

ia s

ubro

stra

ta2

Pinn

ular

ia c

f sub

rost

rata

7Pi

nnul

aria

cf p

laty

ceph

ala

(sm

all)

2Pi

nnul

aria

spp

aff

schr

oete

rae

22

Pinn

ular

ia s

p M

VK1

24Pi

nnul

aria

spM

VAL

28Pi

nnul

aria

sp

MVA

I3

Pinn

ular

ia v

asta

2Pl

acon

eis

clem

entis

24

6Pl

acon

eis

elgi

nens

is4

22

122

Plac

onei

s pl

acen

tula

22

4Pl

anot

hidi

um o

estr

upii

43

Plan

othi

dium

per

agal

li12

28

126

6

App

endi

x 5.

con

tinue

d.

Mel

ville

Isla

nd c

alib

ratio

n di

atom

sSI

TE N

AME

(MV-

)A

BC

DE

FG

HI

JK

LM

NO

PQ

RS

TU

VW

XY

ZAA

ABAC

ADA

EAF

rock

AG

AHA

IAJ

AK

ALAM

ANA

OAP

AQ

ARAS

ATSe

llaph

ora

sp4

2Se

llaph

ora

baci

llim

42

22

28

25

26

Stau

rone

is s

p M

VAA

22

Stau

rone

is a

ncep

s va

r sib

eric

a2

Stau

rone

is a

ncep

s5

22

41

44

48

62

62

214

45

102

46

410

412

42

21

152

Stau

rone

is p

hoen

icen

tero

n2

25

62

Stau

rone

is p

rom

inul

a4

122

16

15St

auro

neis

sm

ithii

22

2St

auro

neis

und

ata

237

556

Stau

rone

is s

p M

VAI

4St

epha

nodi

scus

min

utilu

s1

Step

hano

disc

us v

estib

ulus

3Su

rirel

la a

ngus

tata

25

22

102

Surir

ella

min

uta

22

199

2Ta

bella

ria fl

occu

losa

22

1720

24

145

310

32

453

Tabe

llaria

floc

culo

sa v

ar lo

ng2

Tryb

lione

lla d

ebili

s2

22

TOTA

L51

679

058

878

451

877

595

157

670

745

985

884

176

842

841

959

863

562

959

786

960

00

630

697

687

613

596

357

522

670

530

548

7062

760

050

755

644

970

632

177

459

569

477

958

132

4cy

sts

112

2012

310

415

67

1219

354

5812

5716

181

277

43

293

174

2418

22

18

269

79

28

4175

317

4fu

nky

frag

men

tsye

sye

s 6

yes

App

endi

x 6.

Tra

ce m

etal

wat

er c

hem

istry

dat

a fr

om n

orth

ern

Elle

smer

e Is

land

. A

ll un

its a

re µ

g/L.

IDO

ffic

ial N

ame

Al

As

BB

aB

eC

dC

oC

rC

uFe

Ga

La

Li

Mn

Mo

Ni

PbR

bSb

SrT

iU

VZ

nEP

AC

raig

Lak

e25

.10.

2721

.130

.00

0.00

60.

099

0.06

70.

166

1.31

56.0

0.01

30.

099

6.1

2.43

0.87

31.

430.

185

0.69

0.05

532

3.0

0.00

20.

7760

0.26

72.

57EP

BA

pple

by L

ake

1.6

0.53

24.1

32.2

0<

0.00

05<

0.00

050.

030

0.02

60.

3712

.40.

003

0.00

511

.29.

520.

086

0.42

0.02

60.

430.

023

348.

00.

001

0.11

400.

061

0.19

EPC

Bra

inar

d La

ke3.

00.

305.

426

.20

<0.

0005

<0.

0005

0.03

60.

028

0.46

17.3

0.00

20.

010

2.7

9.47

0.07

70.

980.

028

0.21

0.01

817

2.0

0.00

10.

1320

0.06

30.

72EP

D3.

70.

380.

85.

26<

0.00

050.

002

0.02

30.

016

0.54

25.9

0.00

20.

007

1.1

1.49

0.16

20.

360.

045

0.08

0.04

463

.3<

0.00

10.

0793

0.02

10.

86EP

E7.

31.

589.

36.

790.

002

0.02

00.

103

0.05

91.

2536

.10.

005

0.00

919

.62.

030.

620

0.84

0.05

90.

950.

170

80.5

0.00

21.

0400

0.17

02.

83EP

F17

7.0

0.24

1.0

4.50

0.01

00.

027

0.65

60.

203

0.66

374.

00.

044

0.15

22.

330

.60

0.17

81.

920.

558

0.50

0.05

513

7.0

0.00

30.

0969

0.09

61.

55EP

G30

.30.

041.

127

.70

0.00

60.

003

0.05

80.

100

0.46

48.2

0.00

80.

017

0.8

3.55

0.13

80.

270.

071

0.11

0.01

558

.40.

001

0.36

200.

031

0.71

EPH

6.9

0.48

2.1

30.5

00.

002

0.00

10.

089

0.04

12.

0431

.80.

002

0.00

73.

81.

700.

904

1.10

0.06

80.

180.

091

779.

00.

002

1.12

000.

033

0.45

EPI

2.7

0.15

1.5

3.55

0.00

10.

001

0.04

00.

024

0.68

18.9

0.00

10.

004

0.1

1.92

0.19

30.

310.

061

0.03

0.01

421

.9<

0.00

10.

1040

0.01

30.

71EP

JW

ard

Hun

t Lak

e3.

80.

221.

22.

220.

001

0.01

00.

013

0.04

00.

355.

80.

002

0.00

60.

11.

420.

031

0.14

0.59

90.

030.

010

5.2

<0.

001

0.04

530.

019

0.81

EPK

Lake

A12

.80.

363.

913

.20

0.00

30.

006

0.02

00.

053

0.42

23.2

0.00

50.

013

0.2

4.51

0.04

50.

150.

023

0.07

0.00

414

.0<

0.00

10.

0282

0.07

50.

42EP

L10

.60.

226.

519

.20

0.00

10.

003

0.01

50.

051

0.48

11.6

0.00

40.

022

0.3

1.66

0.08

20.

180.

047

0.07

0.00

832

.6<

0.00

10.

0632

0.20

21.

05EP

M12

.20.

391.

271

.80

0.00

30.

008

0.04

70.

078

1.18

236.

00.

006

0.03

10.

219

.00

0.08

30.

250.

253

0.07

0.02

131

.8<

0.00

10.

1070

0.09

61.

54EP

NLa

ke C

26.

90.

010.

95.

48<

0.00

05<

0.00

050.

011

0.02

50.

2512

.10.

002

0.00

80.

10.

860.

045

0.16

0.06

50.

020.

012

8.5

<0.

001

0.02

130.

024

0.40

EPO

8.6

0.01

0.7

1.26

0.00

2<

0.00

050.

029

0.02

70.

1321

.80.

001

0.02

50.

31.

110.

015

0.21

0.07

30.

050.

011

7.3

0.00

10.

1800

0.03

20.

52EP

P69

.40.

344.

47.

100.

007

0.01

80.

136

0.10

61.

0752

8.0

0.02

20.

334

1.6

28.2

00.

060

0.73

0.51

20.

240.

048

111.

00.

001

0.06

590.

167

3.60

EPQ

20.3

1.30

1.4

4.19

0.00

30.

003

0.17

70.

062

1.31

486.

00.

007

0.02

81.

216

.50

0.16

10.

540.

273

0.05

0.12

371

.5<

0.00

10.

1290

0.07

12.

04EP

R20

.00.

062.

831

.70

0.00

2<

0.00

050.

020

0.05

80.

2225

.60.

008

0.02

70.

71.

350.

046

0.21

0.10

50.

110.

016

139.

00.

001

0.08

310.

042

0.86

EPS

13.6

0.15

2.2

4.81

0.00

30.

007

0.01

00.

066

0.27

12.7

0.00

80.

019

0.4

0.35

0.19

50.

210.

071

0.12

0.03

440

.50.

001

0.13

200.

038

1.16

EPT

11.8

0.11

2.4

6.77

0.00

10.

001

0.00

50.

069

0.24

6.6

0.00

60.

007

0.3

0.10

0.15

90.

170.

054

0.05

0.02

528

.70.

001

0.13

900.

031

2.26

EPU

Car

olyn

Lak

e35

.20.

321.

73.

690.

016

0.05

40.

030

0.08

70.

2460

.10.

020

0.02

91.

11.

040.

118

0.17

0.11

90.

100.

038

56.0

0.00

10.

3410

0.04

60.

94EP

VN

an L

ake

6.0

1.31

5.2

8.95

0.00

20.

003

0.03

80.

027

1.00

16.8

0.00

40.

006

9.2

0.73

0.30

20.

460.

075

0.40

0.19

210

7.0

0.00

11.

0900

0.03

23.

15EP

W52

.10.

611.

18.

230.

008

0.01

20.

055

0.12

00.

4590

.00.

016

0.04

80.

81.

190.

066

0.33

0.22

70.

110.

049

47.1

0.00

10.

2720

0.06

11.

40EP

X8.

40.

333.

125

.70

0.00

30.

008

0.01

40.

056

0.33

12.9

0.00

30.

010

1.6

0.58

0.04

00.

280.

048

0.06

0.06

552

.30.

001

0.19

800.

036

0.11

EPY

2.2

0.58

41.5

17.2

00.

002

0.02

40.

031

0.06

30.

574.

20.

005

0.00

216

.90.

373.

120

0.61

0.05

40.

510.

067

6270

.00.

006

1.58

000.

147

4.48

EPZ

2.3

0.26

3.2

19.6

00.

001

0.00

80.

017

0.01

60.

6427

.90.

003

0.00

31.

40.

620.

037

0.30

0.11

20.

120.

016

68.7

0.00

10.

1790

0.01

70.

71EP

AA

Ket

tle L

ake

14.0

1.27

1.5

9.64

0.00

20.

010

0.06

10.

038

0.68

45.8

0.00

80.

012

9.4

5.23

0.51

20.

530.

086

0.99

0.12

722

4.0

0.00

20.

2350

0.10

60.

91EP

AB

4.0

0.15

1.3

8.21

0.00

20.

093

0.01

20.

023

0.85

12.9

0.00

30.

004

0.3

0.76

0.23

10.

390.

059

0.05

0.01

712

.30.

001

0.01

680.

017

0.75

EPA

C26

.50.

471.

78.

070.

003

0.02

30.

062

0.09

80.

5547

.90.

010

0.01

31.

61.

340.

250

0.57

0.07

00.

130.

099

85.5

0.00

10.

7860

0.06

50.

86EP

AD

73.1

1.39

2.1

41.1

00.

006

0.03

11.

240

0.19

62.

4320

1.0

0.02

10.

022

3.7

37.5

00.

874

2.16

0.26

60.

280.

115

108.

00.

002

2.25

000.

113

4.18

EPA

E18

.41.

0231

.232

.00

0.00

20.

039

0.06

20.

128

1.77

62.2

0.00

70.

008

38.6

12.6

00.

598

2.08

0.02

80.

990.

051

472.

00.

002

2.97

000.

072

3.25

mea

n22

.30.

486.

116

.67

0.00

30.

017

0.10

30.

069

0.75

83.0

0.00

80.

032

4.4

6.44

0.33

20.

600.

139

0.25

0.05

332

1.8

0.00

10.

4754

0.07

31.

48m

edia

n11

.80.

332.

18.

950.

002

0.00

80.

036

0.05

80.

5525

.90.

005

0.01

21.

21.

660.

159

0.36

0.07

10.

110.

038

68.7

0.00

10.

1390

0.06

10.

91m

ax17

7.0

1.58

41.5

71.8

00.

016

0.09

91.

240

0.20

32.

4352

8.0

0.04

40.

334

38.6

37.5

03.

120

2.16

0.59

90.

990.

192

6270

.00.

006

2.97

000.

267

4.48

min

1.6

0.01

0.7

1.26

<0.

0005

<0.

0005

0.00

50.

016

0.13

4.2

0.00

10.

002

0.1

0.10

0.01

50.

140.

023

0.02

0.00

45.

2<

0.00

10.

0168

0.01

30.

11

EP19

Lake

Haz

en17

.50.

191.

78.

220.

002

0.00

20.

022

0.06

10.

1637

.50.

010

0.03

81.

01.

300.

136

0.44

0.06

00.

170.

026

70.5

0.00

10.

1120

0.04

50.

49

App

endi

x 7.

Tra

ce m

etal

wat

er c

hem

istry

dat

a fr

om th

e oa

sis r

egio

n no

rther

n of

Lak

e H

azen

, Elle

smer

e Is

land

. A

ll un

its a

re µ

g/L.

IDO

ffic

ial N

ame

Al

As

BB

aB

eC

dC

oC

rC

uFe

Ga

La

Li

Mn

Mo

Ni

PbR

bSb

SrT

iU

VZn

EP1

Skel

eton

Lak

e22

.70.

323.

611

.80

0.00

10.

013

0.05

50.

063

0.70

29.4

0.00

80.

010

3.8

5.45

0.16

10.

530.

140

0.22

0.00

717

8.0

<0.

001

0.29

700.

090

2.30

EP2

1.3

1.31

4.1

14.0

00.

003

0.03

50.

049

0.05

60.

3672

.30.

006

0.01

23.

110

.90

0.08

70.

430.

043

0.15

0.00

217

3.0

<0.

001

0.09

870.

078

1.52

EP3

20.5

0.33

3.9

15.1

00.

003

0.00

60.

034

0.04

50.

2645

.40.

005

0.00

73.

77.

640.

124

0.41

0.03

10.

190.

002

192.

00.

001

0.13

200.

077

1.42

EP4

8.2

1.12

1.3

16.5

00.

003

0.01

80.

158

0.14

60.

8767

.70.

005

0.02

66.

610

.60

0.25

30.

660.

159

0.99

0.06

419

0.0

0.00

10.

0830

0.25

83.

01EP

54.

41.

549.

227

.70

0.00

30.

003

0.17

00.

153

0.65

149.

00.

004

0.02

012

.14.

840.

349

0.74

0.33

11.

150.

047

295.

00.

002

0.17

500.

225

3.85

EP6

4.0

0.10

4.3

6.04

0.00

20.

004

0.11

10.

057

1.28

143.

00.

001

0.09

01.

311

.20

0.06

70.

940.

058

0.06

0.00

462

.7<

0.00

10.

0071

0.04

13.

63EP

79.

31.

5211

9.0

9.32

0.00

10.

004

0.09

50.

069

0.80

43.6

0.00

50.

009

47.6

6.20

0.30

20.

640.

082

1.02

0.05

440

6.0

0.00

10.

2920

0.26

00.

83EP

82.

01.

7840

.639

.00

0.00

7<

0.00

050.

202

0.23

90.

6520

6.0

0.00

60.

009

24.7

3.50

0.17

91.

330.

071

1.72

0.07

271

7.0

0.00

10.

6660

0.39

22.

21EP

913

.82.

0076

.250

.40

0.00

10.

037

0.38

80.

097

1.71

359.

00.

006

0.01

773

.83.

526.

250

2.97

0.22

62.

350.

157

2410

.00.

009

12.5

000

0.15

26.

38EP

1042

.10.

4611

.341

.50

0.00

30.

007

0.09

60.

146

1.38

86.5

0.01

90.

123

10.2

9.24

0.69

61.

340.

153

0.66

0.04

967

6.0

0.00

51.

5000

0.10

50.

91EP

1119

.60.

571.

022

.30

0.00

10.

006

0.15

00.

045

1.17

231.

00.

007

0.08

73.

618

.00

0.31

80.

790.

248

0.37

0.03

037

6.0

0.00

20.

4500

0.04

46.

66EP

122.

21.

2516

.432

.20

0.00

10.

018

0.21

80.

062

0.88

430.

00.

002

0.01

251

.68.

681.

380

1.80

0.06

71.

890.

080

1630

.00.

002

1.73

000.

066

2.00

EP13

9.6

1.84

6.4

39.0

00.

005

0.00

50.

198

0.11

20.

8917

8.0

0.00

60.

034

13.7

14.5

00.

334

1.13

0.11

50.

740.

093

400.

00.

001

0.20

100.

234

1.41

EP14

4.2

1.66

3.7

27.1

00.

002

0.01

20.

353

0.11

90.

9718

3.0

0.00

20.

016

11.2

20.3

01.

060

1.20

0.09

10.

680.

092

436.

00.

001

0.74

000.

125

2.01

EP15

6.9

0.85

1.1

5.89

0.00

10.

009

0.09

60.

082

0.71

68.2

0.00

50.

025

4.7

4.17

0.34

70.

440.

186

0.42

0.03

921

9.0

0.00

20.

0887

0.25

54.

34EP

164.

72.

039.

017

.40

0.00

20.

007

0.20

00.

095

0.94

92.4

0.00

70.

018

33.7

6.49

0.38

51.

420.

068

2.02

0.07

810

20.0

0.00

30.

3880

0.16

22.

32EP

1712

.71.

107.

524

.80

0.00

20.

022

0.13

50.

243

0.57

73.5

0.00

30.

014

16.5

1.93

0.12

71.

330.

159

0.49

0.06

761

2.0

0.00

10.

5940

0.16

72.

25EP

185.

91.

3315

.836

.20

0.00

20.

001

0.23

60.

093

0.62

117.

00.

010

0.01

718

.925

.50

0.16

11.

510.

108

1.64

0.05

939

1.0

0.00

90.

2310

0.14

92.

56EP

208.

51.

5722

.318

.40

0.00

30.

009

0.20

80.

126

0.97

610.

00.

007

0.02

618

.86.

400.

383

2.47

0.14

11.

620.

148

769.

00.

008

0.59

800.

212

2.78

EP21

3.0

0.94

19.2

19.5

00.

001

0.00

80.

207

0.10

60.

8427

4.0

0.00

40.

009

13.4

8.46

0.54

41.

860.

104

1.17

0.07

157

6.0

0.00

40.

4520

0.08

63.

66EP

2223

.70.

522.

74.

040.

002

0.04

30.

034

0.04

90.

5841

.50.

007

0.07

50.

59.

140.

023

0.32

0.17

20.

150.

024

84.8

<0.

001

0.02

810.

046

2.07

EP23

31.4

0.13

1.6

5.16

0.00

20.

005

0.04

40.

062

0.60

52.1

0.00

80.

146

0.3

4.50

0.02

40.

310.

161

0.14

0.01

078

.30.

001

0.01

930.

036

1.67

EP24

37.1

0.07

1.4

4.30

0.00

30.

001

0.04

00.

079

0.26

53.4

0.01

00.

169

0.2

2.92

0.03

50.

270.

173

0.12

0.01

740

.40.

001

0.03

090.

055

0.81

mea

n12

.91.

0616

.621

.20

0.00

20.

012

0.15

10.

102

0.81

156.

80.

006

0.04

216

.38.

870.

591

1.08

0.13

40.

870.

055

518.

80.

002

0.92

620.

144

2.63

med

ian

8.5

1.12

6.4

18.4

00.

002

0.00

70.

150

0.09

30.

8092

.40.

006

0.01

811

.27.

640.

302

0.94

0.14

00.

680.

054

391.

00.

001

0.29

200.

125

2.25

max

42.1

2.03

119.

050

.40

0.00

70.

043

0.38

80.

243

1.71

610.

00.

019

0.16

973

.825

.50

6.25

02.

970.

331

2.35

0.15

724

10.0

0.00

912

.500

00.

392

6.66

min

1.3

0.07

1.0

4.04

0.00

1<

0.00

050.

034

0.04

50.

2629

.40.

001

0.00

70.

21.

930.

023

0.27

0.03

10.

060.

002

40.4

<0.

001

0.00

710.

036

0.81

App

endi

x 8.

Raw

dia

tom

cou

nts f

rom

MV

-AT

core

. M

VAT

core

cou

nts

Dep

th (c

m)

0.25

0.75

1.25

1.75

2.25

2.75

3.25

3.75

4.25

4.75

5.25

5.75

6.25

6.75

7.25

7.75

8.25

8.75

9.25

9.75

10.2

510

.75

11.2

511

.75

12.2

512

.75

13.5

14.5

15.5

16.5

17.5

18.5

19.5

20.5

Achn

anth

es b

iore

tti0

02

22

Achn

anth

es c

hild

anos

72

Achn

anth

es d

aone

nsis

1Ac

hnan

thes

flex

ella

1239

115

79

812

1413

71

33

82

82

25

07

12

53

41

22

83

50

Achn

anth

es c

f hel

vetic

a (th

in)

01

Achn

anth

es g

risch

una

6Ac

hnan

thes

krie

geri

42

Achn

anth

es la

evis

85

1111

113

1720

110

198

58

103

34

72

21

32

14

63

59

2Ac

hnan

thes

mar

ginu

lata

75

24

92

1210

27

105

136

613

24

14

32

42

Achn

anth

es m

inut

issi

ma

113

419

293

219

332

219

240

306

303

245

151

104

138

6114

488

116

7294

104

8065

8363

6093

9288

107

4698

5080

81Ac

hnan

thes

pet

erse

nii

634

176

104

611

411

1112

123

1511

1916

98

92

73

56

59

66

73

Achn

anth

es s

accu

la2

11

Achn

anth

es s

ubat

omoi

des

13

92

44

48

710

210

63

166

55

811

84

75

912

62

23

43

Achn

anth

es v

entra

lis0

41

22

17

42

23

23

45

12

3Am

phor

a in

arie

nsis

37

62

27

88

11

33

41

26

45

1Am

phor

a sp

17

42

Cal

onei

s si

licul

a2

90

05

09

95

27

1214

02

21

20

00

45

60

41

60

21

26

4C

alon

eis

sp1

11

Cyc

lote

lla a

ntiq

ua0

10

10

11

11

1C

yclo

tella

sp1

(8.5

-9)

12

Cym

bella

arc

tica

1C

ymbe

lla c

esat

ii5

113

64

1110

1317

220

1435

910

1719

1421

156

1511

610

277

1623

59

68

7C

ymbe

lla c

f gra

cilis

thin

12

Cym

bella

cf b

otel

lus

22

2C

ymbe

lla c

f. st

auro

neifo

rmis

00

42

2C

ymbe

lla d

escr

ipta

012

40

02

24

186

215

60

08

66

410

70

74

313

63

14

810

56

Cym

bella

gau

eman

ii0

00

21

Cym

bella

late

ns0

1314

418

63

24

63

24

23

3C

ymbe

lla m

esia

na3

Cym

bella

min

uta

1114

163

395

2116

2910

247

373

63

146

17

24

49

132

410

7C

ymbe

lla c

f obs

cura

2233

64

208

65

25

411

2512

2712

213

198

4216

2511

1613

1220

07

177

1523

Cym

bella

cf p

roxi

ma

21

1C

ymbe

lla o

bscu

ra0

00

2C

ymbe

lla p

erpu

silla

00

00

00

2C

ymbe

lla s

ilesi

aca

03

15

43

14

23

62

1C

ymbe

lla s

p11

Cym

bella

cle

ve-e

uler

ae c

mpl

x1

1610

521

66

96

10

00

40

03

02

20

20

21

20

32

01

23

1D

iato

ma

tenu

is0

00

10

22

04

15

42

1Eu

notia

arc

us2

21

41

26

12

41

11Eu

notia

exi

gua

01

23

3Eu

notia

mon

odon

24

Euno

tia p

raer

upta

213

96

176

87

36

03

02

87

74

316

46

1111

924

1312

1716

2228

1720

Frag

ilaria

cap

ucin

a va

r cap

ucin

a39

7723

4746

3344

4471

6411

516

520

417

523

827

825

525

422

328

317

116

622

017

911

520

615

117

023

857

180

119

211

81Fr

agila

ria c

apuc

ina

var g

raci

lis19

4120

25

28

74

144

42

61

42

12

Frag

ilaria

exi

gua

00

00

12

1Fr

agila

ria p

inna

ta2

App

endi

x 8.

con

tinue

d.

MVA

T co

re c

ount

sD

epth

(cm

)0.

250.

751.

251.

752.

252.

753.

253.

754.

254.

755.

255.

756.

256.

757.

257.

758.

258.

759.

259.

7510

.25

10.7

511

.25

11.7

512

.25

12.7

513

.514

.515

.516

.517

.518

.519

.520

.5G

irdle

s 1

10G

omph

onem

a la

gerh

eim

i0

2G

omph

onem

a m

inut

um0

41

45

22

23

14

42

67

21

24

21

4G

omph

onem

a sp

110

Nav

icul

a ch

iara

e0

23

14

32

212

24

11

2N

avic

ula

cf rh

ynco

ceph

ala

00

2N

avic

ula

cryp

toce

phal

a1

57

21

22

13

2N

avic

ula

dilu

vian

a1

Nav

icul

a ni

valis

1N

avic

ula

prot

ract

a0

00

00

02

Nav

icul

a pu

pula

022

1012

1616

54

53

512

1615

611

2119

1212

107

1112

813

107

163

153

208

Nav

icul

a st

roem

ii2

21

Nav

icul

a tu

scul

a0

00

03

14

31

15

4N

avic

ula

vulp

ina

05

91

42

24

22

31

21

11

Nav

icul

a sp

ecie

s 2

4N

av s

p1(0

.5-1

)2

Nav

sp1

2N

av s

p21

Nav

sp3

1N

av s

p41

Nei

dium

alp

inum

2N

eidi

um a

mpl

iatu

m2

21

22

44

21

58

154

46

21

45

57

79

42

813

75

Nei

dium

sp1

12

Nei

dium

bis

ulca

tum

1N

itzsc

hia

cf a

urar

iae

2N

itzsc

hia

inco

nspi

cua

332

1510

45

76

1010

102

64

14

45

22

27

44

24

Nitz

schi

a cf

per

min

uta

(hea

vy s

triae

)3

Nitz

schi

a fru

stul

um15

106

815

30

66

40

625

100

30

00

211

20

18

220

36

22

23

6N

itzsc

hia

perm

inut

a12

014

781

7216

178

6162

3835

5523

7641

4440

4421

2636

2614

2219

2313

1626

224

1814

2631

Nitz

schi

a pe

rmin

uta

long

04

04

00

00

00

00

00

30

00

00

00

00

00

00

20

00

00

Nitz

schi

a gi

rdle

s (1

.5-2

)8

2Pi

nnul

aria

ang

usta

13

71

63

74

46

146

84

86

88

1010

16

147

182

93

101

Pinn

ular

ia b

alfo

uria

na2

Pinn

ular

ia d

escr

esce

ns v

ar ig

nora

ta1

21

Pinn

ular

ia d

iver

gent

issi

ma

var s

ubro

stra

ta2

11

84

94

31

4Pi

nnul

aria

div

erge

ns v

ar d

iver

gens

13

2Pi

nnul

aria

gib

ba v

ar m

esog

ongy

la3

1Pi

nnul

aria

inte

rmed

ia3

2Pi

nnul

aria

stre

ptor

aphe

12

11

73

65

33

Pinn

sp1

7Pi

nn s

p24

1Pi

nn s

p31

Pinn

ular

ia in

terr

upta

00

21

12

11Pi

nn s

mal

l (6.

5-7)

4pi

nn s

p a

2pi

nn s

p b

(9.5

-10)

2sp

ecie

s a

1St

auro

neis

anc

eps

03

22

62

16

22

44

13

42

31

35

7St

auro

neis

pho

enic

ente

ron

03

57

35

51

19

21

33

29

23

15

76

41

Stau

rone

is c

f obt

usa

12

21

Surir

ella

sp1

00

1Ta

bella

ria fl

occu

losa

25

18

69

22

51

12

21

14

21

6TO

TAL

411

1003

587

468

769

448

485

567

589

497

468

432

658

399

561

555

596

476

487

575

415

374

455

367

296

512

380

431

538

176

444

342

470

312

Cys

ts56

107

94

114

44

24

43

App

endi

x 9.

Raw

dia

tom

cou

nts f

rom

the

Skel

eton

Lak

e (E

P1) s

edim

ent c

ore.

Sk

elet

on L

ake

(EP1

) cor

e 29

5-JP

S-03

Inte

rval

(cm

)0.

251.

252.

253.

254.

255.

256.

257.

258.

259.

2510

.25

11.2

512

.25

13.2

514

.25

15.2

516

.25

17.2

518

.25

19.2

520

.25

21.2

522

.25

23.2

524

.25

25.2

526

.25

27.2

528

.25

29.2

530

.25

31.2

532

.25

Achn

anth

es b

ahus

iens

is5

1Ac

hnan

thes

chi

ldan

os

3Ac

hnan

thes

cur

tissi

ma

13

Achn

anth

es in

grat

iform

is1

Achn

anth

es k

riege

ri1

2Ac

hnan

thes

min

utis

sim

a1

2Ac

hnan

thes

pet

erse

nii

23

24

26

21

85

33

Achn

anth

es v

entra

lis6

1C

alon

eis

silic

ula

22

Cyc

lote

lla a

ntiq

ua2

1C

ymbe

lla c

esat

tii2

2C

ymbe

lla tu

mid

ula

1D

entic

ula

kuet

zing

ii1

2D

iato

ma

mon

ilifor

mis

12

12

Dia

tom

a te

nuis

22

Frag

ilaria

bre

vist

riata

9413

216

315

4720

7616

1813

3339

1016

1511

927

917

1719

1925

2823

6110

145

3848

2739

Frag

ilaria

con

stru

ens

var v

ente

r com

plex

815

965

1170

503

637

397

825

778

653

847

704

763

1053

426

442

543

942

853

727

559

418

925

584

609

988

541

487

1115

547

643

614

489

636

Frag

ilaria

pin

nata

82

65

96

1026

97

1628

5053

9512

350

3728

2547

6763

2346

4254

4613

4121

1017

Frag

ilaria

pse

udoc

onst

ruen

s7

52

32

23

112

144

115

56

012

516

875

6526

101

212

Nav

icul

a ch

iara

e2

21

Nav

icul

a cr

ypto

ceph

ala

2N

avic

ula

gerlo

ffi1

Nav

icul

a pu

pula

11

2N

avic

ula

vulp

ina

12

1N

itzsc

hia

perm

inut

a2

16

2N

itzsc

hia

pusi

lla4

1114

13

23

42

11

2Pi

nnul

aria

bre

biso

nni

6Pi

nnul

aria

stre

ptor

aphe

23

Step

hano

disc

us m

ediu

s3

SUM

935

1123

1363

524

694

427

916

827

687

873

765

832

1133

497

556

688

1015

924

770

603

504

1019

706

667

1138

684

634

1279

608

728

697

527

704

Cys

ts9

48

11

11

21

12

23

42

Appendix 10. Raw counts from EP2 sediment core. EP2 core 297-JPS-03 Interval (cm)

0.25 1.25 2.25 3.25 4.25 5.25 6.25 7.25 8.25 9.25 10.25 11.25 12.25 13.25 14.25 15.25 16.25Achnanthes childanos 4 2 3Achnanthes flexella 1 4 2 1 1 1 1Achnanthes laevis 2Achnanthes marginulata 2 1 2Achnanthes minutissima 9 11 23 33 45 61 105 48 5 2 11 7 1 6Achnanthes petersenii 1 1 5 14 12 9 16 28 24 10 3 5 2 1Achnanthes scotica 1Achnanthes ventralis 2Amphora inariensis 2Brachysira zellensis 4 3 2 4 2 4 4 2Caloneis silicula 6 5 16 13 22 18 26 32 9 2 7 15Cyclotella antiqua 6 2 3 2 10 1 1Cyclotella stelligera 1Cymbella angusta 4 7 6 11 8 29 21 11 2 2 2 4 5Cymbella austriaca 2 2Cymbella bottellus 1 4 4 7 2 4 1Cymbella cesattii 3 6 5 10 10 2Cymbella cleve-eulerae 1 2 3 10 1 4 4 9 2Cymbella fogedii 2 1 6 10 6 12 9 1 1 2Cymbella microcephala (Nav soehensis) 2 14Cymbella minutum 5 2 6 4 11 1 1 2Cymbella obscurum 1 8 2 2 1 1 4 1 2Cymbella proxima 1 1Cymbella silesiaca 2Cymbella subaequalis 2Cymbella tumidula 2 1 1 8 6 9Cymbopleura sp nov? Dermot P34 2 9 2 10 10 4 3 2 2 2 2 2Denticula kuetzingii 5 41 117 123 128 145 82 61 100 77 17 26 14 11 11 8 10Diadesmis contenta 3 1 2 2Diatoma moniliformis 1 1 1Epithemia sorex 5 4 5 4 4 1Eunotia praerupta 1 14 7 8 3 14 6 4 2Fragilaria brevistriata 969 598 474 246 86 75 21 4 16 52 60 66 78 104 83 56 62Fragilaria capucina 1 1 2 4 6 17 18 5 3Fragilaria capucina var gracilis 1 2Fragilaria construens var venter 3 2 2 17 39 40 27 64 60 48 70Fragilaria pinnata 2 10 8 52 72 82 91 96 175 337 568 475 365 541 418 471 465Fragilaria pseudoconstruens football 7 2 19Gomphonema parvulum 2 1 9 2Gomphonema tricatum 1 2Navicula cf angusta 1Navicula cf incerta 2 4 2 1Navicula chiarae 3Navicula cryptocephala 4 1 1Navicula gerloffi 1 1 1 1Navicula pseudoscutiformis 2 1 2 6 2 2Navicula soehernsis 4 38 37 46 13 7 7 6 2 1 4Navicula tuscula 2 3Navicula vulpina 2 3 2 1Nitzschia alpina 2Nitzschia frustulum 16 19 28 4 1Nitzschia inconspicua 5 46 88 96 59 83 54 56 21 1 19 2 2 6Nitzschia palaceae 8 15 10 4 8 5Nitzschia perminuta 9 31 29 65 89 95 100 110 83 18 20 20 20 16 4 12Nitzschia pura 1 1Nitzschia pusilla 2Pinnularia balfouriana 4 2Pinnularia streptoraphe 2Stauroneis anceps 1SUM 988 676 728 629 598 602 585 587 833 758 735 658 578 783 599 596 671Cysts 1 0 2 3 6 6 6 5 10 6 1 1 1 5 3 0

1

1

2

2

3

Appendix 11. 210Pb data (based on alpha spectroscopy from MyCore Scientific, Inc. Deep River, ON) and CRS dates with errors for the Skeleton Lake sediment core.

midpoint year standard deviation Total 210 Pb

(cm) (AD) (years) (pCi/g) 0.25 2001 3.38 1.874780.75 1998 3.55 1.6030241.25 1994 3.805 1.5392161.75 1987 4.32 1.6545822.25 1980 5.01 1.2705132.75 1973 5.91 1.2623663.25 1963 7.63 1.1907383.75 1946 11.985 1.1490624.75 1877 133.14 0.979254

Appendix 12. Total Hg and total Pb for the Skeleton Lake sediment core.

0

5

10

15

20

25

30

35

40

45

50

0 5 10 15 20 25 30 35

depth in core (cm)

conc

entra

tion

Hg &

Pb

(tota

l in

seds

)

Hg (ppb)Pb (ppm)

0

20

40

60

80

100

120

140

160

180

0 5 10 15 20 25 30 35

depth in core (cm)

conc

entra

tion

Hg &

Pb

(exp

ress

ed r

elat

ive

to g

or

gani

c C

)

Hg (ppb)Pb (ppm)

0

2

4

6

8

10

12

0 5 10 15 20 25 30 35

depth in core (cm)

conc

entr

atio

n Hg

& P

b (e

xpre

ssed

rela

tive

to %

or

gani

c C

)

Hg (ppb)Pb (ppm)

App

endi

x 13

. R

aw c

ount

s fro

m C

ape

Ver

a su

rfac

e se

dim

ents

(mul

tiple

yea

r dat

a).

Site

CV1

CV2

CV3

CV4

CV5

CV5

CV6

CV6

CV7

CV8

CV9

CV9

CV9

aC

V10

CV1

1C

V12

CV1

3C

V14

CV1

5C

V16

CV1

7C

V18

CV2

0C

V22

CV2

3C

V24

CV1

3C

V31

Year

2004

2004

2004

2004

2004

2005

2004

2005

2004

2004

2004

2005

2005

2004

2004

2005

2005

2005

2005

2005

2005

2005

2005

2005

2005

2005

2006

2006

Achn

anth

es b

roen

lund

sis

3Ac

hnan

thes

chl

idan

os23

2Ac

hnan

thes

flex

ella

1050

Achn

anth

es k

riege

ri2

817

615

15Ac

hnan

thes

kry

ophi

la1

Achn

anth

es la

evis

21

820

Achn

anth

es m

argi

nula

ta3

175

22

252

1117

1116

Achn

anth

es m

inut

issi

ma

108

412

415

3151

6139

6Ac

hnan

thes

pet

erse

nii

22

53

62

819

16Ac

hnan

thes

sco

tica

2Ac

hnan

thes

ven

tral

is2

Aneu

mas

tus

tusc

ulus

2Am

phor

a co

pula

ta2

2Am

phor

a du

seni

i24

51

2Am

phor

a sp

itzbe

rgen

sis

5Am

phor

a ve

neta

92

Cal

onei

s si

licul

a2

3C

alon

eis

silic

ula

sp16

pla

te 2

09

2C

alon

eis

silic

ula

sp18

pla

te 2

04

Cal

onei

s sp

cf s

ilicu

la s

p9-1

4 pl

ate

2011

84

Cha

mae

pinn

ular

ia g

andr

upii

6C

occo

neis

sp

362

2C

occo

neis

sp2

1C

occo

neis

spC

V18

1C

ratic

ula

ambi

gua

525

Cyc

lote

lla a

ntiq

ua1

Cyl

indr

othe

ca g

raci

le1

Cym

bella

ang

usta

ta11

930

3782

12C

ymbe

lla a

rctic

a3

Cym

bella

bot

ellu

s7

25

811

29

231

1320

188

Cym

bella

cf e

xcis

a2

22

Cym

bella

cle

ve-e

uler

ae13

937

1524

197

1396

3333

1929

5510

2440

479

Cym

bella

des

crip

ta16

5C

ymbe

lla d

esig

nata

22

158

810

2C

ymbe

lla e

lgin

ense

2C

ymbe

lla fo

gedi

i12

2729

6498

3819

279

186

5311

820

2974

2213

1020

2029

Cym

bella

late

ns5

312

19

12

17

4C

ymbe

lla m

icro

ceph

ala

46

223

108

947

Cym

bella

min

uta

2727

235

1211

3014

73

747

1118

1032

212

716

Cym

bella

pau

cist

riatu

m2

10C

ymbe

lla p

roxi

ma

11

3C

ymbe

lla s

ilesi

aca

22

213

95

118

29

42

63

41

141

Cym

bella

sile

siac

a C

V14

164

Cym

bella

sub

aequ

alis

711

112

Cym

bella

sta

uron

eifo

rmis

92

Cym

bopl

eura

cus

pida

ta2

App

endi

x 13

. con

tinue

d.

Site

CV1

CV2

CV3

CV4

CV5

CV5

CV6

CV6

CV7

CV8

CV9

CV9

CV9

aC

V10

CV1

1C

V12

CV1

3C

V14

CV1

5C

V16

CV1

7C

V18

CV2

0C

V22

CV2

3C

V24

CV1

3C

V31

Year

2004

2004

2004

2004

2004

2005

2004

2005

2004

2004

2004

2005

2005

2004

2004

2005

2005

2005

2005

2005

2005

2005

2005

2005

2005

2005

2006

2006

Den

ticul

a ku

etzi

ngii

148

429

1119

Den

ticul

a su

btili

s2

Den

ticul

a te

nuis

911

Der

mot

sp1

7 pl

ate

4931

Dia

desm

is c

onte

nta

898

131

4D

iade

smis

par

acon

tent

a2

Dia

tom

a te

nuis

211

1717

58

24

266

653

415

212

Dip

lone

is m

arge

nies

tria

ta16

05

Dip

lone

is p

arm

a2

Ency

onem

a sp

4 pl

ate

283

112

Frag

ilaria

cap

ucin

a14

812

61

938

13

72

4033

Frag

ilaria

con

stru

ens

var v

ente

r1

Gei

ssle

ria s

choe

nfel

dii

212

Gei

ssle

ria s

p1

Gei

ssle

ria s

p19-

23 p

late

50

2G

eiss

leria

sp1

3,14

pla

te 5

04

Gom

phon

ema

lapp

onic

a de

rmot

10G

omph

onem

a pr

oduc

tum

18H

anna

ea a

rcus

102

41

Mic

roco

stat

us k

rass

kei

175

3330

25

9126

71

114

21N

avic

ula

chia

rae

214

115

755

169

Nav

icul

a ci

ncta

318

Nav

icul

a cr

ypto

ceph

ala

811

212

2N

avic

ula

cryp

toce

phal

a 2

4N

avic

ula

gerlo

ffi5

Nav

icul

a gr

egar

ia19

Nav

icul

a m

enis

culu

s2

Nav

icul

a ph

ylle

pta

102

42

72

102

517

Nav

icul

a ps

eudo

tene

lloid

es2

Nav

icul

a sa

linar

um5

34

214

18N

avic

ula

spp

aff r

einh

ardt

ii9

Nav

icul

a sp

p af

f ven

eta

26

26N

avic

ula

vulp

ina

2315

614

144

531

2512

48

11N

eidi

um a

mpl

iatu

m1

2N

eidi

um d

istin

cte-

punc

tatu

m6

23

1N

eidi

um k

ozlo

wii

2N

itzsc

hia

diss

ipat

a va

r med

ia22

12

27

17

5N

itzsc

hia

frus

tulu

m18

710

910

831

.316

.742

.550

248

326

416

221

228

288.

133

1.4

33.1

442

.79

368.

644

2.9

383.

738

3.6

044

6.8

286.

115

533

2.9

183

42N

itzsc

hia

perm

inut

a35

670

.419

720

38.

3312

011

061

.511

523

.752

.949

.214

5.9

174.

662

.86

44.2

181

.37

73.1

313

0.3

76.4

254

417

.19

82.8

810

412

7.1

7166

Nitz

schi

a ho

mbu

rgie

nsis

112

4N

itzsc

hia

pale

a15

34N

itzsc

hia

pala

ecea

125

1421

112

Nitz

schi

a pu

ra3

1834

702

154

314

Nitz

schi

a su

chla

ndii

204

Pinn

ular

ia b

rebi

sonn

i sp9

pla

te61

1014

Pinn

ular

ia h

umili

s8

Sella

phor

a ba

cillu

m1

Stau

rone

is a

ncep

s1

12

Stau

rone

is p

hoen

icen

tero

n1

1TO

TAL

721

614

510

469

460

394

795

684

468

476

633

574

450

553

687

617

066

072

167

454

654

558

364

168

067

947

537

0

Appendix 14. Raw diatom counts from CV5 core . CV5 core1

0.5 1.5 2.5 3.5 4.5 5.5 6.5 7.5 8.5 9.5Achnanthes broenlundensis 2 1Achnanthes chlidanos 13 14 4 3 2Achnanthes flexella 1 1Achnanthes girdles 4 14 6 2 10 28Achnanthes kreigeri 9Achnanthes kryophila 2Achnanthes marginulata 10 5Achnanthes minutissima 1 4 2 2 5 4 12Achnanthes petersenni 3 6 2 10 7 8 9 4 3Achnanthes scotica 2Adlafia bryophila 2 2Amphora copulata 2 2 2Amphora dusenii 2 4 6 1 1 2 2Caloneis sp cf silicula dermot 9-14 plate 20 4Caloneis sp cf silicula 2 4 2Cocconeis sp 3Cyclotella pseudostelligera 2 2Cymbella amphicephela 3 2 2Cymbella gauemanni 1 1Cymbella lange-bertalotti 2Cymbella latens 1 3 9 10 11 8 3 2Cymbella microcephala 7 1 4 2 1 2Cymbella silesiaca 42 41 22 27 22 20 34 52 21 32Cymbella large girdle 3Cymbella fogedii 118 83 80 64 91 75 77 63 37 42Cymbella minuta 100 23 18 18 33 36 43 64 19 24Cymbella bottelus 20 0 1 0 2 2 0 0 3 7Cymbella cleve-eulerae 8 6 8 13 7 16 6 13 8 12Cymbopleura cuspidata 4 3Diatoma tenuis 1 3 8 7 10 4 4 7 7Fragilaria capucina 4 4Hannaea arcus 2Hippodonta arctica 2Luticola mutica var ventricosa 2 2 2Microstarus krasskei 14 15 9 11 17 14 8 1Navicula chiarae 34 21 14 10 3 8 1 3 8 5Navicula cryptocephala 2 2 2 2 4Navicula phyllepta 7 19 19 4 15 16 7 6 10 3Navicula salinarum 15 13 6 6 10 2 3 2Navicula sp aff veneta 14 11 2 4 2Navicula sm girdles 2Navicula vulpina 1 5 2 8 9 6 4 3 11 9Navicula sp 29-38 plate 47 dermot 8Neidium distincte-punctatum 2 2 2Nitzschia homburgensis 2Nitzschia frustulum 44 85 114 169 238 236 175 148 76 241Nitzschia perminuta 95 68 88 67 204 223 201 168 226 64Nitzschia paleacea 3Nitzschia pura 3 10 4 1 2 2 3Pinnularia grunowii 2Pinnularia krammeri 1 3Pinnularia sp 2-4 plate 63 dermot 4Pinnularia vasta 1 2Stauroneis anceps 3 6 7Stauroneis phoenicenteron 8 2 2 2 1TOTAL 561 454 441 470 690 697 596 562 463 517cysts 7 5 3 9 6 4 4 4 2 2

Appendix 15. Raw diatom counts from CV6 core 1.

CV6 core10.25 1.25 2.25 3.25 4.25 5.25 6.25 7.25 8.25 9.25 10.25 11.25

Achnanthes flexella 4Achnanthes girdles 13 8 2 4 10 4Achnanthes kreigeri 2Achnanthes marginulata 2Achnanthes minutissima 4 2 6 3 2 1 7 15 7 10 1Achnanthes petersenni 2 2 2 3Achnanthes subatomoides 2Caloneis sp cf silicula 2 2 2 2 4Cyclotella pseudostelligera 6 1Cymbella angustata 1 8 4 2 3 3 2 7 6 2Cymbella amphicephela 2Cymbella botellus 4 3 12 2 27 23 82 15 69 77Cymbella cleve-eulerae 13 8 4 4 4 2 7 2 3 12 11 2Cymbella designata 7 6 4 2 1Cymbella latens 6 5 10 7 3 2 9 2Cymbella microcephala 8 8 14 17 5 5 6 14 12 6 38 13Cymbella proxima 2 2Cymbella silesiaca 2 9 3 6 7 2 2 2 5 7Cymbella fogedii 34 34 10 21 14 15 10 12 6 15 11 0Cymbella minuta 4 4 4 7 0 8 8 3 3 12 4Cymbella subaequalis 2 4 6 6 2 7 4 2Cymbella stauroneiformis 13 13 13 10 6 2 18 2Denticula kuetzigii 2 4Diadesmis contenta 2Diatoma tenuis 3 1 2 2 2 2 1 1Fragilaria capucina 6 12 4 5 9 4 13 1Fragilaria construens var venter 19Fragilaria pseudoconstruens 3Microstarus krasskei 3 4 2 1Navicula chiarae 23 11 16 11 13 2 3 7 2 23 3Navicula phyllepta 2Navicula sm girdles 2 7 2 4 4 2 2Navicula vulpina 11 12 24 19 33 26 58 45 61 20 54 26Neidium ampliatum 2 2 5 5 1 7 1 7Nitzschia dissipata var media 3Nitzschia frustulum 569 504 389 302 393 373 178 250 163 493 188 125Nitzschia perminuta 99 165 83 80 52 58 35 60 21 73 49 29Nitzschia sp cf acicularis 2Stauroneis phoenicenteron 1 3 1 1 2 1weird cv6 1TOTAL 795 809 606 509 578 538 364 460 406 747 471 369cysts 1 1 4 1 6 3 1 5 2 12 8 1

4

63

3

2

32

2

6

2

0

Appendix 16. Raw diatom counts from CV7 core 1.

CV7 core10.25 1.25 2.25 3.25 4.25 5.25 6.25-18.25 19.25

Achnanthes marginulata 7 11 4 12 2 NOAchnanthes minutissima 1 DIATOMSAchnanthes petersenni 2 2Amphora inariensis 2Caloneis silicula 2 9 4 2 4Caloneis silicula big 4Cocconeis sp 24Cymbella angustata 2 2Cymbella botellus 16 12 30 29 14 57 12Cymbella cleve-eulerae 67 31 30 19 27 30 4Cymbella designata 2 3 6 2Cymbella lange-bertalotti 6Cymbella latens 22 4 5 7 5 8Cymbella silesiaca 13 9 4 1Cymbella fogedii 41 21 63 49 25 24 4Cymbella minuta 13 21 30 21 14 4 0Cymbopleura cuspidata 2Diadesmis contenta 2 3Diploneis marginestriata 4 4 2 1Diploneis parma 6 8 4 2 1Fragilaria capucina 27 14 4 6 6 5Geissleria sp 19-23 plate 50 dermot 2Luticola mutica var ventricosa 2Microstarus krasskei 46 12 2 28 2Navicula chiarae 104 34 49 23 19 15 10Navicula cryptocephala 2Navicula phyllepta 4 6 14 23 24 2 2Navicula pseudotenelloides 4 6 16Navicula sp aff veneta 48 9 13 24 7 8 2Navicula sm girdles 2 2 8 2 6Navicula vulpina 6 12 9 10 2 4Navicula sp 29-38 plate 47 dermot 40 9 3 15Neidium ampliatum 5 4Neidium bergii 4 3Neidium distincte-punctatum 5 2Neidium kozlowii 2 2Nitzschia homburgensis 2 1Nitzschia frustulum 73 53 176 46 82 23 10Nitzschira perminuta 233 72 133 140 103 135 30Nitzschia palea 10 2 2 4Nitzschia pura 6Nitzschia sp cf acicularis 2unidentified (looks like peanut in the shell) 6Pinnularia biceps 2Pinnularia vasta 4 3 2 8Pinnularia large girdle CV7 4Sellaphora bacillum 4 2 7 2Stauroneis anceps 2 1TOTAL 766 333 639 519 395 341 0 121cysts 15 3 5 7 6 8

8

Appendix 17. Raw diatom counts from CV9 core 1.

CV9 core 10.5 1.5 2.5 3.5 4.5 5.5 6.5 7.5 8.5 9.5 10.5 11.5 12.5 13.5 14.5 15.5 16.5 17.5 18.5 19.5 20.5 21.5 22

Achnanthes marginulata 1Achnanthes petersenni 2Aneumastus tusculus 2Amphora dusenii 4 4Amphora inariensis 1 2Caloneis sp cf silicula 15 6 6 8 4 4 2 6 4 1 4 3 4 1 4 1 2Cocconeis sp 1 1 1 8 2 1 2Cymbella botellus 5 2 6 1 2 3Cymbella cleve-eulerae 2 1 3 6 4 6 19 14 11 15 5 14 9 12 12Cymbella latens 1 2 4 3 5 1 8 4 8 8 7 8Cymbella microcephala 5 7 13 5 5 7 6 2 5 1 7 6 3 6 2 4 23 6 15 3 1Cymbella fogedii total 60 102 85 83 84 92 59 49 50 56 33 46 28 33 10 22 9 15 21 22 11 11 11Cymbella minuta total 54 82 73 35 37 57 42 13 12 11 9 11 17 10 1 4 5 6 4 0 11 0 6Cymbella silesiaca 30 15 30 24 23 46 33 22 30 16 7 5 11 23 9 13 5 12 8 12 7 16 11Cymbella subaequalis 1Cymbella tynni 1Cymbella CV8 1Cymbopleura cuspidata 2Denticula kuetzigii 2 2Diadesmis contenta 2 10Diatoma tenuis 3 1 1Fragilaria brevistriata 2Fragilaria capucina 1 2 1Luticola muticaMicrostarus krasskei 8 19 14 4 2 2 2 1Navicula chiaraeNavicula cryptocephala 23 7 6 2Navicula menisculus 2Navicula muticola 1 1 1 1Navicula phyllepta 69 86 60 38 9 21 11 8 2 3 2 2 1 1 2Navicula salinarum 4 2Navicula trivalis 12 1Navicula tuscula 3 2Navicula vulpina 14 24 25 30 32 51 65 66 78 67 74 40 50 50 27 32 20 29 46 34 28 41 24Navicula large central area 5 1Neidium ampliatum 1 4 7 5 2 6 5 3 1 6 1 1 6Neidium distincte-punctatum 2 1 8 1 2 1 1 2 1 3Neidium kozlowii 1 1Nitzschia dissipata var media 3 2 5 3 4 5 2 4 5 8Nitzschia frustulum total 174 212 174 165 214 269 199 214 249 211 186 192 133 200 242 223 294 210 309 265 227 234 225Nitzschia perminuta total 61 80 32 20 18 60 29 40 22 53 27 34 51 45 28 14 38 49 21 38 11 64 22Nitzschia pusilla 4 2Pinnularia krammeri 1Sellaphora bacillum 1Stauroneis anceps 1Stauroneis phoenicenteron 4 4 2 4Big punctae CV8 1TOTAL 525 652 516 428 446 627 454 433 450 431 353 352 318 386 351 338 414 352 467 414 340 412 332cysts 6 5 3 2 2 3 4 1 3 1 2 4 7 5 4 5 21 6 8 18 2funky fragments 1 1 3 2 3 12 9 4 9 3 1 2

2

3

1

Appendix 18. Raw diatom counts from CV9a core 1.

CV9 core1 20050.25 1.25 2.25 3.25 4.25 5.25 6.25 7.25 8.25 9.25 10.25 11.25 12.25

Achnanthes marginulata 4 2 2 2 4Achnanthes minutissima 1Achnanthes ventralis 1Amphora inariensis 4Centric 1 1Cyclotella antiqua 1 1Cyclotella bodanica 2Cyclotella ocellata 1Cyclotella pseudostelligera 2 1 1 3 35Cymbella angustata 2Cymbella botellus 2Cymbella cleve-eulerae 2Cymbella fogedii 12 3 1 4Cymbella minuta 9 7 6 14 18 5 5 1 2 9 12 6Cymbella proxima 2 1 4 1 2Cymbella silesiaca 1 6 4 9 10 2 1 4Fragilaria capucina 2 2 1Fragilaria capucina var gracilis 12Fragilaria construens var venter 3 4 4 2 19Fragilaria pinnata 18Gomphonema productum 3Navicula trivalis 1Neidium ampliatum 2Nitzschia dissipata var media 5 2 8 5Nitzschia frustulum 473 573 504 581 440 309 293 256 46 46 87 77 45Nitzschia perminuta 76 138 148 146 150 150 182 89 54 11 59 11 10unidentified 1 1 1unidentified 2 1unidentified 3 3unidentified weird stick 2 1TOTAL 567 735 671 743 613 473 499 357 108 65 169 114 175cysts 32 17 14 28 6 7 4 2 1 297 6 8funky fragments 1 7 7 3 12 5

8

Appendix 19. Raw diatom counts from CV12 core 1.

CV12 broken broken broken broken broken0-0.5 1-1.5 2-2.5 3-3.5 4-4.5 5-5.5 6-6.5

Achnanthes girdle 4Achnanthes petersenii 2 2 1Amphora inariensis 4 2Amphora veneta 4 3Cocconeis sp 1 5 6 7 11Cocconeis (neodiminuta?) 3 2Cyclotella bodanica 1Big centric 3 1Cymbella cleve-eulerae 13 19 7 2 3Denticula kuetzigii 68 70 93 9 26 2 8Denticula tenuis 2Diatoma tenuis 4 10 16 12 3 1 2Diploneis didyma 4Diploneis big 3 1Fragilaria exigua 1 1Luticola mutica 3Microcostatus krasskei 6 6 1 1Navicula large centres 18 29 20Navicula rhyncocephala 7 27 5 2 1 1Navicula chiarae 1 2 2Navicula phyllepta 2 5 33 17 6 3Navicula vulpina 2Neidium large centres 33 73 95 72 7Nitzschia communtata 5 4 14 2 2Nitzschia frustulum 362 194 180 20 46 6 11Nitzschia palaecea 2Nitzschia perminuta 71 35 25 8 5 4Nitzschia pusilla 2TOTAL 536 362 444 153 209 130 150cysts 11 17 32 2 2 9 6funky fragments 3 14 13 10 27

1

App

endi

x 20

. Raw

dia

tom

cou

nts f

rom

CV

13 c

ore

1.

C

V13

0-0.

51-

1.5

2-2.

53-

3.5

4-4.

55-

5.5

6-6.

57-

7.5

8-8.

59-

9.5

10-1

0.5

11-1

1.5

12-1

2.5

13-1

3.5

Amph

ora

hols

taic

a4

44

22

Cal

onei

s si

licul

a2

Coc

cone

is s

p2

13

3C

ratic

ula

halo

phila

2C

yclo

tella

bod

anic

a1

Cym

bella

bot

ellu

s4

Cym

bella

cle

ve-e

uler

ae47

6574

3325

99

13

53

72

8C

ymbe

lla fo

gedi

i20

100

122

8962

5156

4147

5446

7858

43C

ymbe

lla m

inut

a12

749

544

1815

185

119

911

1811

Cym

bella

neo

cist

ula

62

Cym

bella

sile

siac

a14

3030

2012

37

717

154

76

Cym

bella

sm

all g

irdle

146

82

122

Dia

desm

is c

onte

nta

1D

iato

ma

tenu

is2

114

126

1930

2332

5820

5018

Frag

ilaria

cap

ucin

a4

Mic

roco

stat

us k

rass

kei

1N

avic

ula

rhyn

coce

phal

a2

23

21

Nav

icul

a vu

lpin

a11

1714

118

28

14

Nitz

schi

a fru

stul

um18

316

318

618

628

229

230

322

027

129

125

233

231

827

7N

itzsc

hia

perm

inut

a71

105

105

107

6042

3748

101

8164

9138

72TO

TAL

475

538

566

497

483

424

462

362

477

491

443

567

500

436

cyst

s11

227

62

115

156

716

67

Appendix 21. Raw diatom counts from CV20 core.

CV20 core 10.25 1.25 2.25 3.25 4.25 5.25 6.25 7.25

Cocconeis spCV20 1Cyclotella pseudostelligera 23 2 2Cymbella affinis 2Cymbella botellus 5 2 3 2Cymbella cleve-eulerae 25 30 36 21 23 29 20 25Cymbella fogedii 4 5 14 1 4Cymbella minuta 34 33 47 41 29 11 9 13Cymbella proxima 6 2 4 7 3Cymbella silesiaca 11 6 3 7 4 7 13 1Cymbella small girdle 6Fragilaria brevistriata 2Fragilaria capucina 2Fragilaria capucina var gracilis 1Fragilaria construens var venter 8 1 2Fragilaria pinnata 2 2Nitzschia frustulum 270 428 424 306 276 279 266 247Nitzschia perminuta 50 60 44 13 9 10 5 8TOTAL 430 579 572 389 350 342 325 328cysts 59 78 71 64 86 120 228 338

4

5

6

Appendix 22. Raw diatom counts from CV22 core.

CV22 core1 20050.25 1.25 2.25 3.25 4.25 5.25 6.25 7.25

Achnanthes flexella 12 2Achnanthes kriegeri 0 3 0 0 0 0 0 0Achnanthes laevis 0 1 3 0 2 0 1 0Achnanthes marginulata 0 6 0 0 0 0 0 0Achnanthes minutissima 58 23 26 15 10 13 16 5Achnanthes petersenni 11 3 4 3 0 3 20Amphora dusenii 2Amphora copulata 1 4Caloneis sp cf silicula 1Cocconeis sp 5 10 4 53 16 18 73Cocconeis small 8 7 4 3Cyclotella pseudostelligera 1Cymbella angustata 3 3 5 3Cymbella botellus 14 7 13 19 9 10 4 2Cymbella cleve-eulerae 13 22 5 4 9 13 12 2Cymbella descripta 7 5 2 6 3Cymbella designata 2 6 2Cymbella elginense 5Cymbella fogedii 17 11 6 4 3 13 7 2Cymbella microcephala 3 4 2 6 2 2Cymbella obscurum 1 1Cymbella proxima 1Cymbella silesiaca 2 2Cymbella subaequalis 7 4 2 1 2Cymbella tynni 2Cymbopleura cuspidata 2Denticula kuetzigii 8 1 1 7 6 4Denticula tenuis 31 21 8 5 7 2Diatoma tenuis 8 11 9 6 6 10 6 2Diploneis margenestriata 2Diploneis parma 1Fragilaria brevistriata 1Fragilaria capucina 4 2 3 4 5Fragilaria capucina var gracilis 1Large central 1 1Muelleria gibbula 1Navicula chiarae 10 6 2 12 2 2Navicula cryptocephala 2 4 2 6 4 3Navicula digitoradiata 4 17 5 2 2 1Navicula muticola 1 1Navicula phyllepta 1 5 4 3 4 2Navicula vulpina 3 14 76 14 117 23 32 130Neidium ampliatum 1 2 2Neidium distincte-punctatum 2 2Nitzschia clausii 2Nitzschia dissipata var media 8 6 6 4 4 16 11Nitzschia palaceae 7 5 2 4Nitzschia frustulum 314 251 159 177 66 189 143 29Nitzschia perminuta 383 208 210 173 57 176 180 46Nitzschia sublinearis 1 2Nitzschia suchlandii 6Orthoseira roseana 1Pinnularia krammeri 1Sellaphora bacillum 3 3Sp Dermot CV22 1 1unidentified 1 1TOTAL 924 649 582 474 382 530 479 371cysts 7 2 4 0 4 1 1 4funky fragments 14 4 49 2 15 35

0

5

2

3

5

2

21

Appendix 23. Raw diatom counts from CV24 core. CV24 core1

0.25 1.25 2.25 3.25 4.25 5.25Achnanthes marginulata 5Achnanthes minutissima 10 10 8 15 18 14Achnanthes petersenni 3Cymbella angustata 11 19 7 5 6 5Cymbella botellus 25 16 26 4 5 23Cymbella cleve-eulerae 26 12 15 8 5 9Cymbella descripta 1 2Cymbella designata 4Cymbella fogedii 9 13 12 7 6 13Cymbella microcephela 19 21 32 15 12 24Cymbella minuta 5 1 3Cymbella proxima 1Cymbella silesiaca 2 2 4Cymbella subaequalis 3 6 2 1 11Denticula kuetzingii 38 24 24 26 13 8Diatoma tenuis 1Encyonema sp18-24 plate 35 10 5 6 1 5Hannea arcus 2 3Navicula chiarae 1 4Navicula vulpina 5 8 6 6 14Nitzschia dissipata var media 1Nitzschia frustulum 311 366 263 209 267 268Nitzschia perminuta 79 76 47 65 32 37total valves 555 573 457 371 381 435cysts 1 4 9