UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Relative influence of temperature and disturbance on vegetation dynamics in the Low Arctic : an investigation… Lantz, Trevor Charles 2008

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

Item Metadata

Download

Media
24-ubc_2008_fall_lantz_trevor.pdf [ 2.57MB ]
Metadata
JSON: 24-1.0066494.json
JSON-LD: 24-1.0066494-ld.json
RDF/XML (Pretty): 24-1.0066494-rdf.xml
RDF/JSON: 24-1.0066494-rdf.json
Turtle: 24-1.0066494-turtle.txt
N-Triples: 24-1.0066494-rdf-ntriples.txt
Original Record: 24-1.0066494-source.json
Full Text
24-1.0066494-fulltext.txt
Citation
24-1.0066494.ris

Full Text

RELATIVE INFLUENCE OF TEMPERATURE AND DISTURBANCE ON VEGETATION DYNAMICS IN THE LOW ARCTIC: AN INVESTIGATION AT MULTIPLE SCALES  by  TREVOR CHARLES LANTZ  B.Sc., University of Alberta, 1998 M.Sc., University of Victoria, 2001  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Forestry)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) July 2008  © Trevor Charles Lantz, 2008  Abstract Climate change will affect Arctic plant communities directly, by altering growth and recruitment, and indirectly, by increasing the frequency of natural disturbance. Since the structure of northern vegetation influences global climate, understanding both temperature and disturbance effects on vegetation is critical. Here, I investigate the influence of temperature and disturbance on Low Arctic vegetation at several spatio-temporal scales in the Mackenzie Delta Region, N.W.T. To disentangle the relative impact of temperature and disturbance on forest-tundra and tundra ecosystems, I sampled microenvironmental variability, plant community composition, and green alder abundance, growth, and reproduction on disturbed (burns and thaw slumps) and undisturbed sites across a regional temperature gradient. Disturbed areas showed increases in alder productivity, catkin production, and seed viability, as well as differences in plant community composition and microenvironment. The magnitude of plot-level responses to disturbance compared to variation across the temperature gradient suggests that in the short-term, increasing the frequency of disturbance may exert a stronger influence on tundra ecosystems than changes in temperature. At the plot level, increases in alder seed viability and recruitment at warmer sites point to the fine-scale mechanisms by which shrub abundance will change. To examine the relative influence of temperature and biophysical variables on landscape-level patterns of shrub dominance, I mapped Low Arctic vegetation using aerial photos. At this broader scale, correlations between temperature and the areal extent of shrub tundra suggest that warming will increase the dominance of shrub tundra. To assess the magnitude of changes in temperature and thaw slump activity, I analyzed climate records and mapped retrogressive thaw slumps using aerial photographs. An increase in thaw slump activity in recent decades, coincident with higher temperatures, suggests that continued warming will change the area affected by thermokarst disturbances like slumps. Taken together, my research indicates that the effects climate change will be magnified by shifts in the frequency of disturbance, initiating changes to Arctic vegetation with significant implications for global climate. My work also shows that to fully understand the influence of patch-landscape feedbacks on Arctic vegetation dynamics, the effects of disturbance must be examined across longer temporal and broader spatial scales.  ii  Table of Contents ABSTRACT....................................................................................................................... ii TABLE OF CONTENTS ................................................................................................ iii LIST OF TABLES .......................................................................................................... vii LIST OF FIGURES ......................................................................................................... iii ACKNOWLEDGEMENTS ............................................................................................ xi DEDICATION................................................................................................................ xiii CO-AUTHORSHIP STATEMENT ............................................................................. xiv CHAPTER 1. GENERAL INTRODUCTION ............................................................... 1 BACKGROUND ................................................................................................................ 1 TEMPERATURE EFFECTS.............................................................................................. 2 DISTURBANCE EFFECTS............................................................................................... 4 DISSERTATION OUTLINE.............................................................................................. 5 LITERATURE CITED ....................................................................................................... 9 CHAPTER 2. RESPONSE OF GREEN ALDER (ALNUS VIRIDIS SUBSP. FRUTICOSA) AND PLANT COMMUNITY COMPOSITION TO FIRE ACROSS TREELINE IN THE MACKENZIE DELTA REGION, NWT. ................................ 15 INTRODUCTION ............................................................................................................ 15 Objective 1: Regional temperature, green alder abundance and seed viability. ........... 17 Objective 2 – Relative effects of temperature and fire on green alder and plant community composition................................................................................................ 17 METHODS ....................................................................................................................... 19 Study Area .................................................................................................................... 19 Field Sampling .............................................................................................................. 19 Statistical Analysis........................................................................................................ 21 Objective 1: Regional temperature, green alder abundance and seed viability....... 21 Objective 2 – Relative effects of temperature and fire on green alder and plant community composition. ........................................................................................... 22 iii  RESULTS ......................................................................................................................... 24 Objective 1: Regional temperature, green alder abundance and seed viability. ........... 24 Objective 2 – Relative effects of temperature and fire on green alder and plant community composition................................................................................................ 25 DISCUSSION ................................................................................................................... 37 Regional temperature, green alder abundance and seed viability................................. 37 Relative effects of temperature and fire on green alder and plant community composition................................................................................................................... 39 Green Alder: ............................................................................................................. 39 Plant community composition: ................................................................................. 39 Implications................................................................................................................... 40 ACKNOWLEDGEMENTS.............................................................................................. 42 LITERATURE CITED ..................................................................................................... 43 CHAPTER 3. LONG-TERM CHANGES IN MICROENVIRONMENT, PLANT COMMUNITY COMPOSITION, AND SHRUB GROWTH IN RETROGRESSIVE THAW SLUMPS............................................................................................................. 50 INTRODUCTION ............................................................................................................ 50 METHODS ....................................................................................................................... 53 Study Area .................................................................................................................... 53 Slump Sampling............................................................................................................ 53 Abiotic Data .................................................................................................................. 55 Vegetation Sampling..................................................................................................... 56 Statistical Analysis........................................................................................................ 57 RESULTS ......................................................................................................................... 59 Regional Temperature Conditions ................................................................................ 59 Snow Pack, Ground Temperature and Active Layer Depth ......................................... 60 Nutrient Availability and Soils ..................................................................................... 62 Plant community composition ...................................................................................... 63 Green alder autecology ................................................................................................. 66 DISCUSSION ................................................................................................................... 74 Snow Pack, Ground Temperature and Active Layer Depth ......................................... 74 Nutrient Availability and Soils ..................................................................................... 75 Persistence of altered microenvironmental conditions on stable slumps...................... 76 Biotic Effects ................................................................................................................ 77 Community Composition........................................................................................... 77 Green Alder............................................................................................................... 79 Implications................................................................................................................... 80 ACKNOWLEDGEMENTS.............................................................................................. 81  iv  LITERATURE CITED ..................................................................................................... 83 CHAPTER 4. INCREASING RATES OF RETROGRESSIVE THAW SLUMP ACTIVITY IN THE MACKENZIE DELTA REGION, N.W.T. ............................... 92 INTRODUCTION ............................................................................................................ 92 METHODS ....................................................................................................................... 93 RESULTS ......................................................................................................................... 96 ACKNOWLEDGEMENTS............................................................................................ 103 LITERATURE CITED ................................................................................................... 104 CHAPTER 5. VARIABILITY IN THE VEGETATION OF THE SHRUB TUNDRA ECOTONE IN THE MACKENZIE DELTA REGION, NWT: IMPLICATIONS FOR GLOBAL CHANGE. .......................................................................................... 106 INTRODUCTION .......................................................................................................... 106 Objective 1: Describe latitudinal changes in the proportion and patch sizes of shrub tundra and dwarf shrub tundra. ................................................................................... 107 Objective 2: Determine the relative importance of biophysical drivers of the shrub tundra ecotone............................................................................................................. 107 METHODS ..................................................................................................................... 108 Study Area .................................................................................................................. 108 Airphoto Selection and Image Manipulation.............................................................. 110 Rationale for Object-Based Classification.................................................................. 110 Object-Based Segmentation and Classification Procedure......................................... 111 Segmentation........................................................................................................... 113 Classification .......................................................................................................... 113 Accuracy Assessment .............................................................................................. 114 Regional Biophysical Variability................................................................................ 116 Statistical Analyses ..................................................................................................... 116 Objective 1: Describe latitudinal changes in the proportion and patch sizes of shrub tundra and dwarf shrub tundra............................................................................... 116 Objective 2: Determine the relative importance of biophyscical drivers of the shrub tundra ecotone. ....................................................................................................... 118 RESULTS ....................................................................................................................... 119 Image Brightness and Accuracy Assessment ............................................................. 119 Objective 1: Describe latitudinal changes in the proportion and patch sizes of shrub tundra and dwarf shrub tundra. ................................................................................... 119 Objective 2: Determine the relative importance of biophysical drivers of the shrub tundra ecotone............................................................................................................. 123  v  DISCUSSION ................................................................................................................. 128 Latitudinal Changes in Shrub Abundance: Implications for Global Temperatures.... 128 Latitudinal Changes in Shrub Patch Size: Implications for Regional Models............ 129 Patch Size and Positive Shrub-Snow-Nutrient Feedbacks.......................................... 130 Relative importance of biophysical drivers of the shrub tundra ecotone.................... 131 Future Shrub Proliferation .......................................................................................... 132 Conclusions................................................................................................................. 133 ACKNOWLEDGEMENTS............................................................................................ 133 LITERATURE CITED ................................................................................................... 135 CHAPTER 6. CONCLUSIONS................................................................................... 141 SUMMARY.................................................................................................................... 143 UNCERTAINTY AND RESEARCH NEEDS............................................................... 145 Spatial Dynamics in a Warming World: Linking Plots and Landscapes.................... 145 Long-term Cumulative Impacts .................................................................................. 146 FUTURE CONSEQEUNCES OF A WARMING WORLD.......................................... 149 LITERATURE CITED ................................................................................................... 150  vi  List of Tables Table 2.1. Comparisons of green alder age distributions across treeline using the Kolmogorov-Smirnov statistic.......................................................................................... 33 Table 2.2. Results of the ANOVA for green alder response variables. ............................ 34 Table 2.3. Pair wise comparisons of plant community composition between effects using the ANOSIM procedure. ................................................................................................. 35 Table 2.4. Results of the SIMPER analysis showing the top five species or species group contribution to between group dissimilarity. .................................................................... 36 Table 3.1. ANOVA Table for Abiotic Response Variables.............................................. 71 Table 3.2. Pair wise comparisons of plant community composition between site types using the ANOSIM procedure. ....................................................................................... 72 Table 3.3. Results of ANOVA for Alder Autecology Response Variables...................... 73 Table 5.1. Biophysical characteristics of each sample plot. ......................................... 117 Table 5.2. Classification accuracy assessments............................................................. 124 Table 5.3. Comparison of linear and non-linear models using Adjusted R2, AIC, and AIC weights. . ........................................................................................................................ 125 Table 5.4. Spearman rank correlation coefficients for comparisons among the proportion of shrub tundra and biophysical variables. ................................................................... 126 Table 5.5. Results of the hierarchical variance partitioning analysis. .......................... 127  vii  List of Figures Figure 1.1. Generalized effects and feedbacks of climate and disturbance on tundra ecosystems. ......................................................................................................................... 3 Figure 1.2. Conceptual framework for the dissertation. Black arrows and numbers indicate chapters where specific effects are investigated. ................................................ 7 Figure 2.1. Map of the study region showing settlements, control sub-regions, (green) burned subregions (red stippling), water (blue), and the boundaries of the treeline zones sensu Timoney et al. 1992. ............................................................................................... 18 Figure 2.2. Mean growing season temperature and latitude for 10 sites in the Mackenzie delta uplands, June-August 2005. .................................................................................... 22 Figure 2.3. Green alder characteristics measured on undisturbed controls across the entire subarctic-low arctic gradient treeline gradient.................................................................. 27 Figure 2.4. Green alder age distributions across the treeline ecotone. . .......................... 28 Figure 2.5. Green alder characteristics measured on undisturbed controls (light gray bars) and burned sites (dark gray bars) across the northern and southern transition zone. ..... 29 Figure 2.6. Non-metric multidimensional scaling ordinations of plant community composition based on Bray-Curtis similarity matrix. ...................................................... 31 Figure 2.7. 2004 aerial photo showing the northern extent of a 1968 fire in the northern transition zone. ................................................................................................................ 32 Figure 3.1. Photo of an active retrogressive thaw slump showing visible signs of terrain slumping. ......................................................................................................................... 52 Figure 3.2. Map of the study region showing study sites, study sites with temperature loggers, vegetation physiogamy sensu Timoney et al. 1992. ......................................... 54 Figure 3.3. Mean growing season temperature and latitude for 8 sites in the Mackenzie delta uplands, June-August 2005 (r2 =0.933, P< 0.001). ................................................ 59 Figure 3.4. Abiotic variables measured on active slumps (Active), stable slumps (Stable) and undisturbed controls (Control) in the low arctic and the subarctic transition zone. .. 61 Figure 3.5. Starting date of active layer freezeback (at 100 cm depth) at 4 sites in the study area (from north to south: Jimmy Lake, Parsons Lake, Lucas Point, and Denis Lake). .............................................................................................................................. 62  viii  Figure 3.6. Abiotic variables measured on active slumps (Active), stable slumps (Stable) and undisturbed controls (Control) in the Low Arctic and the subarctic transition zone. . ........................................................................................................................................... 64 Figure 3.7. Mean percent cover of plant functional groups by site type. ......................... 65 Figure 3.8. Non-metric multidimensional scaling ordination of plant community composition based on Bray-Curtis similarity matrix. ..................................................... 67 Figure 3.9. Photographs showing typical plant community structure on slumps and undisturbed sites in Mackenzie Delta uplands. ................................................................ 68 Figure 3.10. Growth and reproduction of green alder measured on stable slumps (stable) and undisturbed controls (control) in the Low Arctic and the subarctic transition zone... ........................................................................................................................................... 69 Figure 3.11. Growth and reproduction of green alder measured on stable slumps (stable) and undisturbed controls (control) in the Low Arctic and the subarctic transition zone: (A) proportion of plots with alder (B) catkins / m2, (C) percent germination. ............... 70 Figure 4.1. Retrogressive thaw slump adjacent to a tundra lake, Richards Island, Mackenzie Delta region. .................................................................................................. 93 Figure 4.2. Map of the retrogressive thaw slumps in the upland tundra study region east of the Mackenzie River Delta. ......................................................................................... 95 Figure 4.3. Air temperature time series for the central Mackenzie Delta region from 1926 to 2006. ........................................................................................................................... 98 Figure 4.4. Mean rates of slump growth in the Mackenzie Delta region. (A) Average annual rates of slump growth estimated from the change in areal extent of disturbance from 1950 to 1973 (n=110) and from 1973 to 2004 (n=110) for all active slumps mapped on the 23 s study plots, and (B) average annual rates of headwall retreat from 1950 to 1973 (n=50) and 1973 to 2004 (n=50). ......................................................................... 100 Figure 5.1. Map of the study region showing the study area, settlements, water (blue), temperature loggers, and airphoto study plots. .............................................................. 109 Figure 5.2. Diagram showing the sequence of operations in the object-based classification each upland tundra plot. ................................................................................................ 112 Figure 5.3. Example image from helicopter surveys used to train image classifiers and in accuracy assessments. . .................................................................................................. 115  ix  Figure 5.4. To evaluate possible bias in brightness values within airphotos, the relationship between pixel brightness vs. distance from principle point was examined in three sample plots in the study area. ............................................................................. 120 Figure 5.5. Least squares regressions of: A) proportion of shrub-tundra vs. latitude (y=73.3495x - 0.5336x2 - 2519.9439, F2, 15 = 38.25, P<0.001 Adjusted R2=0.8142) .. 121 Figure 5.6. Least squares regressions of: A) proportion of dwarf shrub tundra vs. latitude (y=-73.3495x + 0.5336x2 + 2520.9439, F2, 15 = 38.25, P<0.001 Adjusted R2=0.8142) and B) mean dwarf shrub patch size vs. latitude (y=-759116x + 5512x2 + 26137774, F2, 15 = 15.97, P<0.001 Adjusted R2=0.6379)..................................................................... 122 Figure 6.1. Generalized effects and feedbacks of climate and disturbance on tundra ecosystems. ................................................................................................................... 142 Figure 6.2. Map showing the distribution of the major disturbances in the study area. 147  x  Acknowledgements I have been looking forward to writing this section for a long time now. Over the last four and a half years I have been extremely fortunate to work with, and amongst, a group of extraordinary individuals. I hope that I can return the generosity I have been shown. During my first field season in Inuvik I was incredibly lucky to cross paths with Dr. Steve Kokelj. Within an hour of first meeting Steve, he was sharing his maps, loaning me equipment, and was passing on important logistical advice. His insistence that I think more carefully about the ecological importance of “frozen mud” has had an enormous influence on the direction of this dissertation, as well as that of ongoing research. Thanks Steve. I would also like to thank all the tenacious individuals who endured the delta bugs in the field or worked long hours in the lab. Without their efforts this research simply would not have been possible. Thanks to: Sarah Bogart, Marc Cassis, Doug Joe Esagok, Celina Gabriel, Sarah Gergel, Larry Greenland, Michael Jansen, Robert Jenkins, Alexis Johnson, Julian Kanigan, Steve Kokelj, Stephanie Mills, Peter Morse, Mike Palmer, Rory Tooke, Rufus Kinmiak, Erika Tizya-Tramm, Matt Tomlinson, Anika Trimble, and Peter Verstrasen. Northern field research also depends on the contribution of a huge number of individuals who don’t get to go out into the field. Thanks to: Andrea Chan, Rosemarie Cheng, Kate Del Bel, Deb Feduik, Norman Hodges, Gayle Kosh, Sandy Lapsky, Jerry Maedel, Susan Rootman, Candice Staley, Tracey Teasdale, and Tom Wray. The funding support critical to this research was provided by a range of organizations. Thanks to: the Arctic Institute of North America (Grant-in-Aid), the Aurora Research Institute (Research Fellowships), Canon USA and the AAAS (Canon National Parks Science Scholarship), Global Forest Research (Research Grant GF-182004-210), Indian and Northern Affairs Canada (Water Resources Division and the Northern Science Training Program), the Killiam Trusts (Predoctoral Fellowship), Natural Resources Canada (Polar Continental Shelf Program), the Natural Sciences and  xi  Engineering Research Council of Canada (PGS-B Scholarship and Northern Internship to T.C Lantz, and grants to S.E. Gergel and G.H.R Henry). My progress over the course of this project was frequently punctuated by leaps of understanding that emerged out of conceptual and technical discussions with friends and colleagues. Thanks to: Christopher Bater, Rebecca Best, Big Joe Bennett, Chris Burn, Alaine Camfield, Ben Gilbert, Kendra Holt, Laurie Beth Marczak, Isla Meyers-Smith, Yulia Stange, Shanley Thompson, and Carmen Wong. I would also like to thank my supervisory committee for their support, insight, promptness, and collegiality over the last four and a half years. I would particularly like to acknowledge Sarah Gergel’s interest in, and dedication to, this project. Thanks to Gladys and Cory Alexie (Fort Macpherson) for sharing their home and cranberry recipes in the summer of 2004, and to Mike Craig, Jared Egerington, Les Kutny, and Pippa Secombe-Hett for logistical assistance in Inuvik. Thanks to everyone in the Gergel Lab (Selina Agbayani, Kate Kirby, Jessica Morgan, Yulia Stange, Shanley Thompson, and Matt Tomlinson,) for their encouragement, good humour and help with rock and roll wigs. Playing music on a regular basis at UBC provided a wonderful diversion that always brought laughter and clarity on dimly lit afternoons. Thanks to: Peter Arcese, Joe Bennett, Rebecca Best, Sierra Curtis-Mclean, Ian Dalmeyer, Brad Fedy, Sarah Gergel, and Laurie Beth Marczak. Thanks to Kendra Holt for her support and patience throughout the final stages of this process. Finally, thanks to Victor Lin and Nicholas Alexsiuk for long ago wisdom, and Alestine Andre and Ruth Welsh for planting the seed.  xii  Dedication  to Helen and Oliver Lantz ( … for everything)  xiii  Co-authorship Statement Chapter 2 was co-authored with Sarah Gergel and Greg Henry. I formulated the research questions, designed and conducted the research and analysis. My co-authors assisted with revisions. Chapter 3 was co-authored with Steve Kokelj, Sarah Gergel and Greg Henry. I formulated the research questions, designed and conducted the research and analysis. My co-authors assisted with revisions. Chapter 4 was co-authored with Sarah Gergel, Steve Kokelj, and Nicholas Coops. With the exception of the classification of ground truth photos, which was completed by Steve Kokelj, I formulated the research questions, designed and conducted the research and carried out the analyses. My co-authors assisted with revisions. Chapter 5 was co-authored with Steve Kokelj. The research questions explored in this Chapter emerged out of discussions with Steve as a part of an NSERC Northern Research Internship with the Water Resources Division of Indian and Northern Affairs Canada. The development of research questions and revisions were shared between Steve and myself. Slump mapping was conducted by Peter Verstrasen at Sub-arctic Surveys (Yellowknife, N.W.T). I conducted the majority of the data analysis and led manuscript preparation.  xiv  Chapter 1. General Introduction Global warming connects us all. Use what is happening in the Arctic as a vehicle to connect us all, so that we may understand that the planet and its people are one. The Inuit hunter who falls through the depleting and unpredictable sea ice is connected to the cars we drive, the industries we rely upon, and the disposable world we have become. Sheila Watt-Cloutier (President and Chair of the Inuit Circumpolar Council, 1996-2006) BACKGROUND Rising global temperatures are predicted to significantly alter the structure and function of world’s ecosystems (Cramer et al. 2001, Fischlin and Midgley 2007, Notaro et al. 2007). At high latitudes average temperatures have risen at nearly double the rate of the rest of the planet (Hassol 2004, Johannessen et al. 2004, Parry and Intergovernmental Panel on Climate Change. Working Group II. 2007), and these increases have been linked to a number of profound changes in Arctic ecosystems. Some of these changes include: warming permafrost, drying of lakes, altered natural disturbance regimes, reductions in the extent of sea ice, shifts in community composition, treeline movement, and the increased proliferation of tall shrubs (Osterkamp and Romanovsky 1999, Lloyd and Fastie 2002, Smol et al. 2005, Riordan et al. 2006, Tape et al. 2006, Perovich et al. 2007, Smol and Douglas 2007a, b, Lantz and Kokelj 2008). Such changes are often used to highlight the Arctic’s role as a sentinel, alerting the globe to the imminence of climate impacts in all biomes (Meadows 2001, Foley 2005, Struzik 2007). This metaphor of the Arctic as the ‘canary’ in the biosphere’s ‘coalmine’ is often used to emphasize the need to act quickly in order to avoid the negative ecological and social consequences that will be associated with a changing climate (Hassol 2004, Parry and Intergovernmental Panel on Climate Change. Working Group II. 2007). Many of the observed changes at high latitudes represent climate forcings of a magnitude similar to that predicted by doubling of atmospheric CO2 (Chapin et al. 2005, Perovich et al. 2007). The strength of these climate feedbacks suggest that avoiding the consequences of a warmer world may be unlikely. Although some alterations may produce negative feedbacks, the weight of current evidence suggests that the net effect of 1  change will exacerbate warming. The three most important links between terrestrial Arctic ecosystems and the global climate system are: 1) albedo, 2) partitioning of energy fluxes mediated by permafrost dynamics, and 3) terrestrial emissions of CO2, CH4 and N2O. Vegetation, and its interactions with permafrost, can change the direction of all of these feedbacks (McGuire et al. 2006). Thus, our ability to predict how vegetation will respond to a changing climate represents a fundamental challenge in global change ecology. Increasing Arctic temperatures are likely to exert direct and indirect effects on both vegetation and natural disturbance regimes (Henry and Molau 1997, Shaver et al. 2000, Chapin et al. 2006, Figure 1.1: A, B). Like temperature changes, natural disturbance can dramatically alter Arctic vegetation structure and community composition (Chapin et al. 1995, Shaver et al. 2000, Forbes et al. 2001, Racine et al. 2004, Figure 1.1: C). However, the long-term effects and interactions of temperature and disturbance are not well understood. As disturbance becomes more frequent in the Arctic it is likely to exert a stronger influence on vegetation dynamics. Changes in vegetation structure, driven by temperature increases or disturbance may drive secondary alterations to the frequency of natural disturbance (Chapin 2003, Higuera et al. 2008, Figure 1.1C) that will have strong feedbacks to the global climate system by altering vegetation and releasing greenhouse gases (Thompson et al. 2004, Chapin et al. 2005, Figure 1.1A, B). Consequently, disentangling the effects of natural disturbance from interactions between the direct and indirect effects of increased temperatures is a critical component of understanding the global consequences of climate change.  TEMPERATURE EFFECTS Over the last two decades considerable research effort has been directed at predicting the response of Arctic plant communities to climate change through plot level manipulations of temperature, nutrients, moisture, and light (Chapin et al. 1995, Henry and Molau 1997, Arft et al. 1999, Shaver et al. 2000, Dormann and Woodin 2002, Walker et al. 2006). Temperature increases are expected to affect Arctic plants directly by altering photosynthetic and other physiological rates, and indirectly by influencing 2  other factors such as nutrient availability and snow pack duration (Chapin 1983, Chapin et al. 1995, Henry and Molau 1997, Shaver et al. 2000). Short term changes in growth, phenology, reproduction, species composition, and productivity in response to perturbations simulating climate change indicate that Arctic ecosystems are very sensitive to changes in climate (Chapin et al. 1995, Jonasson et al. 1999, Shaver et al. 2000, Graglia et al. 2001, Dormann and Woodin 2002). Results from these experiments also highlight the importance of the indirect effects of increasing temperature, particularly those altering resource availability (Chapin and Shaver 1985, 1996, Arft et al. 1999, Graglia et al. 2001, Dormann and Woodin 2002, Hollister et al. 2005a, Hollister et al. 2005b, Walker et al. 2006). .  Figure 1.1. Generalized effects and feedbacks of climate and disturbance on tundra ecosystems. (A) Direct and indirect effects of temperature on vegetation and abiotic conditions will feedback to regional climate. (B) Increasing temperatures will alter the frequency of natural disturbance. (C) Disturbance will alter tundra ecosystems, which may lead to additional changes in the frequency of natural disturbance.  Manipulative experiments have greatly improved our understanding of the factors influencing Arctic vegetation, but differences among ecosystems, functional groups and individual species (Chapin et al. 1995, Chapin and Shaver 1996, Arft et al. 1999, Shaver et al. 2000, Walker et al. 2006) complicate efforts to scale up predictions to broader temporal and spatial scales. In many cases, short-term responses have also differed from 3  longer-term findings (Shaver et al. 2000, Hollister et al. 2005a). Consequently, there is a need to test the generalizations generated by this work across broader spatial and temporal scales (Shaver et al. 2000, Dunne et al. 2004, Hollister et al. 2005a). Comparing experimental results with individual and community responses across regional environmental gradients that integrate longer term responses to key drivers is an important way to test the broad scale applicability of predictions derived from experimental manipulations A growing number of broad scale observations also indicate that vegetation is changing (Myneni et al. 1997, Silapaswan et al. 2001, Lloyd and Fastie 2002, Jia et al. 2003, Stow et al. 2004, Tape et al. 2006) and understanding the factors that may constrain or facilitate change at broad scales is vital component of realistic regional predictions.  DISTURBANCE EFFECTS One factor that is likely to exert a strong influence on the fate of Arctic vegetation is natural disturbance (Forbes et al. 2001, McGuire et al. 2006). For the purposes of this dissertation I define disturbance as a discrete event in time and space (100-109 m2) that “disrupts the ecosystem, community or population structure and changes the resources, substrate availability or physical environment” (Pickett and White 1985: 7) In the Low Arctic disturbances can increase the opportunity for colonization by exposing new substrates, and altering nutrient availability and permafrost conditions (Mackay 1970, Wein and Bliss 1973, Truett and Kertell 1992, Forbes et al. 2001, Chapin 2003, Kokelj and Burn 2003, Yoshikawa et al. 2003, Kokelj et al. 2005, Smithwick et al. 2005). Growing evidence suggests that the frequency of natural disturbance is increasing (Murphy et al. 2000, Jorgenson and Osterkamp 2005, Kasischke and Turetsky 2006, Lantz and Kokelj 2008) indicating that its effects will become more significant over time. By increasing the potential for establishment, elevating nutrient availability, improving microclimate, and reducing the strength of competition in the short-term, disturbance is likely to facilitate rapid ecosystem-level responses to warming. Observations of shortterm vegetation responses to disturbance include increased growth, reproduction and altered community structure (Hernandez 1973, Chapin and Shaver 1981, Cargill and 4  Chapin 1987, Walker et al. 1987, Landhäusser and Wein 1993, Vavrek et al. 1999, de Groot and Wein 2004, Racine et al. 2004) The magnitude of these effects suggests that disturbance and disturbancetemperature interactions will have much a larger effect that temperature alone. While functional recovery from disturbance can be very rapid in the Low Arctic, longer term successional trajectories are not well understood Dynamics inferred from observations of undisturbed tundra communities suggest that complete recovery from disturbance (i.e. a return to antecedent conditions) may only occur on centennial or millennial time scales (Billings and Peterson 1980, Peterson and Billings 1980, Cargill and Chapin 1987, Svoboda and Henry 1987, Walker et al. 1994). It is also possible that secondary succession in a warmer world will lead to communities with no modern analogue (Rupp et al. 2000, Forbes et al. 2001, Williams and Jackson 2007). However, a lack of data on long-term effects and disturbance-temperature interactions, as well as uncertainty about the cumulative impact of multiple disturbances, make this difficult to assess. Thus, investigations of the long-term consequences of disturbance and disturbance- temperature interactions are necessary to fully assess the ecosystem level consequence of disturbance.  DISSERTATION OUTLINE I examine the relative importance of regional temperature and disturbance on tundra and forest-tundra vegetation at several spatio-temporal scales using a suite of complementary techniques. This research explores the drivers of plant community composition, green alder population dynamics, microenvironment variability, and regional vegetation structure. I also explore the influence of temperature change on disturbances such as retrogressive thaw slumps and examine the long-term effects of fire and thermokarst disturbance on vegetation succession and microenvironment. This research was carried out in the Mackenzie Delta Region, N.W.T, Canada, primarily in the Pleistocene uplands east of the Mackenzie River Delta. A strong temperature gradient coupled with relatively uniform topography and parent materials (Burn 1997, Aylsworth et al. 2000) make the Delta uplands an excellent area to attempt to disentangle the relative influence of disturbance and temperature on vegetation. Dominant disturbances in this 5  area include: fires (Wein 1975), retrogressive thaw slumps (Lantz and Kokelj 2008), and thermokarst lake drainage (Marsh et al. 2005). Anthropogenic disturbances including trails from seismic exploration and drilling mud sumps and gravel pads associated with gas exploration and extraction in the 1970s are also common (Mackay 1970, Hernandez 1973, Kemper 2005, Johnstone and Kokelj In Press). The proposed construction of Mackenzie Gas Pipeline from the Beaufort Coast through the Delta uplands will also increase the frequency of anthropogenic disturbance (Holroyd and Retzer 2005). Green alder (Alnus viridis ssp. frtuticosa (Ruprect) Nyman) was selected as a focal species because of its wide distribution in the Mackenzie Delta uplands. In many ecosystems this species is among the first to colonize after disturbance and is a particularly successful invader of mineral soils and newly exposed substrates along streams, floodplains, coastlines, lake edges, and retreating glaciers (Haeussler et al. 1990, Matthews 1992, Beaudry et al. 2000). In the upland terrain adjacent to the Mackenzie River Delta, green alder is not confined to disturbed sites, but is also common in both dry and moist terrain where it is frequently one of the most dominant species. Although the niche occupied by green alder in the delta uplands is likely not representative of the entire Low Arctic, its dominance on disturbed and undisturbed sites in the study area make it an excellent focal species to contrast the relative influence of temperature and disturbance on tall shrubs. It has also been identified as the species primarily responsible for increased shrub cover visible on air photos of Alaska (Tape et al. 2006). Rationale for specific studies and additional background information is presented in the introductory sections of the chapters that follow. A conceptual framework for the dissertation is presented in Figure 1.2.  6  Figure 1.2. Conceptual framework for the dissertation. Black arrows and numbers indicate chapters where specific effects are investigated. (Grey arrows show feedbacks not explicitly explored here).  My specific objectives are as follows: Chapter 2. To disentangle the relative influence of temperature and fire on the vegetation of the forest tundra ecotone, I compared: green alder abundance, growth, and reproduction, and plant community composition on burned and unburned field plots across a regional temperature gradient.  Chapter 3. To separate the relative effects of temperature and thermokarst disturbance in the Low Arctic I contrasted: 1) microenvironmental conditions, 2) green alder abundance, growth, and reproduction, and 3) plant community composition on retrogressive thaw slumps and undisturbed tundra across a regional temperature gradient.  Chapter 4. To explore the link between temperature changes and the frequency of thermokarst disturbance I examined historical temperature records and used air photos  7  from 1950, 1973, and 2004 to estimate changes in the activity of retrogressive thaw slumps across the landscape.  Chapter 5. To explore potential landscape-level drivers of shrub dominance in the Low Arctic, I mapped vegetation across the shrub tundra-dwarf shrub tundra transition and investigated broad scale correlations between shrub dominance and biophysical variables governing the location of this ecotone.  In Chapter 6 I synthesize results and conclusions and highlight areas of uncertainty and future research needs.  By examining the effects and interactions of temperature and disturbance across a regional environmental gradient over broad-spatial scales and long time frames, this research broadens our understanding of the role disturbance will play in a warmer world. My multi-scale approach, addressing fine scale plot-level mechanisms within the context of the regional landscape, is intended to yield a perspective that informs attempts to extrapolate across spatial scales.  8  LITERATURE CITED Arft, A. M., M. D. Walker, J. Gurevitch, J. M. Alatalo, M. S. Bret-Harte, M. Dale, M. Diemer, F. Gugerli, G. H. R. Henry, M. H. Jones, R. D. Hollister, I. S. Jonsdottir, K. Laine, E. Levesque, G. M. Marion, U. Molau, P. Molgaard, U. Nordenhall, V. Raszhivin, C. H. Robinson, G. Starr, A. Stenstrom, M. Stenstrom, O. Totland, P. L. Turner, L. J. Walker, P. J. Webber, J. M. Welker, and P. A. Wookey. 1999. Responses of tundra plants to experimental warming: Meta-analysis of the international tundra experiment. Ecological Monographs 69:491-511. Aylsworth, J. M., M. M. Burgess, D. T. Desrochers, A. Duk-Rodkin, T. Robertson, and J. A. Traynor. 2000. Surficial geology, subsurface materials, and thaw sensitivity of sediments. Pages 41-48 in L. D. Dyke and G. R. Brooks, editors. The phsyical environment of the Mackenzie Valley, Northwest Territories: A base line for the assessment of environmental change. Geological Survey of Canada Bulletin 547. Geological Survey of Canada, Ottawa, ON. Billings, W. D., and K. M. Peterson. 1980. Vegetational Change and Ice-Wedge Polygons through the Thaw-Lake Cycle in Arctic Alaska. Arctic and Alpine Research 12:413-432. Beaudry, L., R. Coupé, C. Delong, and J. Pojar. 2000. Plant Indicator Guide for Northern British Columbia: Boreal, Sub-Boreal, and Subalpine Biogeoclimatic Zones (BWBS, SBS, SBPS, and northern ESSF). B.C. Ministry of Forests Research Branch, Victoria, B.C. Burn, C. R. 1997. Cryostratigraphy, paleogeography, and climate change during the early Holocene warm interval, western Arctic coast, Canada. Canadian Journal of Earth Sciences 34:912-925. Cargill, S. M., and F. S. Chapin. 1987. Application of Successional Theory to Tundra Restoration - a Review. Arctic and Alpine Research 19:366-372. Chapin, F. S. 1983. Direct and Indirect Effects of Temperature on Arctic Plants. Polar Biology 2:47-52. Chapin, F. S. 2003. Effects of plant traits on ecosystem and regional processes: a conceptual framework for predicting the consequences of global change. Annals of Botany 91:455-463. Chapin, F. S., M. D. Robards, H. P. Huntington, J. E. Johnstone, S. E. Trainor, G. P. Kofinas, R. W. Ruess, N. Fresco, D. C. Natcher, and R. L. Naylor. 2006. Directional changes in ecological communities and social-ecological systems: A framework for prediction based on Alaskan examples. American Naturalist 168:S36-S49. Chapin, F. S., and G. R. Shaver. 1981. Changes in Soil Properties and Vegetation Following Disturbance of Alaskan Arctic Tundra. Journal of Applied Ecology 18:605-617. Chapin, F. S., and G. R. Shaver. 1985. Individualistic Growth-Response of Tundra PlantSpecies to Environmental Manipulations in the Field. Ecology 66:564-576. Chapin, F. S., and G. R. Shaver. 1996. Physiological and growth responses of arctic plants to a field experiment simulating climatic change. Ecology 77:822-840.  9  Chapin, F. S., G. R. Shaver, A. E. Giblin, K. J. Nadelhoffer, and J. A. Laundre. 1995. Responses of Arctic Tundra to Experimental and Observed Changes in Climate. Ecology 76:694-711. Chapin, F. S., M. Sturm, M. C. Serreze, J. P. McFadden, J. R. Key, A. H. Lloyd, A. D. McGuire, T. S. Rupp, A. H. Lynch, J. P. Schimel, J. Beringer, W. L. Chapman, H. E. Epstein, E. S. Euskirchen, L. D. Hinzman, G. Jia, C. L. Ping, K. D. Tape, C. D. C. Thompson, D. A. Walker, and J. M. Welker. 2005. Role of land-surface changes in Arctic summer warming. Science 310:657-660. Cramer, W., A. Bondeau, F. I. Woodward, I. C. Prentice, R. A. Betts, V. Brovkin, P. M. Cox, V. Fisher, J. A. Foley, A. D. Friend, C. Kucharik, M. R. Lomas, N. Ramankutty, S. Sitch, B. Smith, A. White, and C. Young-Molling. 2001. Global response of terrestrial ecosystem structure and function to CO2 and climate change: results from six dynamic global vegetation models. Global Change Biology 7:357-373. de Groot, W. J., and R. Wein. 2004. Effects of fire severity and season of burn on Betula glandulosa growth dynamics. International Journal of Wildland Fire 13:287-295. Dormann, C. F., and S. J. Woodin. 2002. Climate change in the Arctic: using plant functional types in a meta-analysis of field experiments. Functional Ecology 16:417. Dunne, J. A., S. R. Saleska, M. L. Fischer, and J. Harte. 2004. Integrating experimental and gradient methods in ecological climate change research. Ecology 85:904-916. Fischlin, A., and G. F. Midgley. 2007. Ecosystems, their properties, goods and services. Pages 211-272 in M. L. Parry, O. F. Canziani, J. P. Palutikof, P. J. v. d. Linden, and C. E. Hanson, editors. Climate change 2007: Impacts, adaptation and vulnerability. Contribution of Working Group II to the Assessment Report Four of the Intergovernmental Panel of Climate Change (IPCC). Cambridge University Press Cambridge, UK. Foley, J. A. 2005. Tipping Points in the Tundra. Science 310:627-628. Forbes, B. C., J. J. Ebersole, and B. Strandberg. 2001. Anthropogenic disturbance and patch dynamics in circumpolar arctic ecosystems. Conservation Biology 15:954969. Graglia, E., S. Jonasson, A. Michelsen, I. K. Schmidt, M. Havstrom, and L. Gustavsson. 2001. Effects of environmental perturbations on abundance of subarctic plants after three, seven and ten years of treatments. Ecography 24:5-12. Haeussler, S., D. Coates, and J. Mather. 1990. Autecology of common plants in British Columbia : a literature review. Forestry Canada, Victoria, B.C. Hassol, S. J. 2004. Arctic Climate Impact Assessment: Impacts of Warming Climate. Cambridge University Press, Cambridge. Henry, G. H. R., and U. Molau. 1997. Tundra plants and climate change: the International Tundra Experiment (ITEX). Global Change Biology 3(Suppl. 1):1-9. Hernandez, H. 1973. Natural Plant Recolonization of Surficial Disturbances, Tuktoyaktuk Peninsula Region, Northwest Territories. Canadian Journal of Botany-Revue Canadienne De Botanique 51:2177-2196. Higuera, P. E., L. B. Brubaker, P. M. Anderson, T. A. Brown, A. T. Kennedy, and F. S. Hu. 2008. Frequent Fires in Ancient Shrub Tundra: Implications of Paleo-records for Arctic Environmental Change. PLoS ONE 3: e0001744. . 10  Hollister, R. D., P. J. Webber, and C. Bay. 2005a. Plant response to temperature in Northern Alaska: Implications for predicting vegetation change. Ecology 86:1562-1570. Hollister, R. D., P. J. Webber, and C. E. Tweedie. 2005b. The response of Alaskan arctic tundra to experimental warming: differences between short- and long-term responses. Global Change Biology 11:525-536. Holroyd, P., and H. Retzer. 2005. A Peak into the Future: The Potential Landscape Impacts from Gas Development in Northern Canada. The Pembina Institute for Appropriate Development, Calgary, AB. Jia, G. S. J., H. E. Epstein, and D. A. Walker. 2003. Greening of arctic Alaska, 19812001. Geophysical Research Letters 30:-. Johannessen, O. M., L. Bengtsson, M. W. Miles, S. I. Kuzmina, V. A. Semenov, G. V. Alekseev, A. P. Nagurnyi, V. F. Zakharov, L. P. Bobylev, L. H. Pettersson, K. Hasselmann, and A. P. Cattle. 2004. Arctic climate change: observed and modelled temperature and sea-ice variability. Tellus Series a-Dynamic Meteorology and Oceanography 56:328-341. Johnstone, J. F., and S. V. Kokelj. In Press. Environmental conditions and vegetation recovery at abandoned-drilling mud sumps in the Mackenzie Delta region, NWT, Canada. Arctic. Jonasson, S., A. Michelsen, I. K. Schmidt, and E. V. Nielsen. 1999. Responses in microbes and plants to changed temperature, nutrient, and light regimes in the arctic. Ecology 80:1828-1843. Jorgenson, M. T., and T. E. Osterkamp. 2005. Response of boreal ecosystems to varying modes of permafrost degradation. Canadian Journal of Forest Research-Revue Canadienne De Recherche Forestiere 35:2100-2111. Kasischke, E. S., and M. R. Turetsky. 2006. Recent changes in the fire regime across the North American boreal region - Spatial and temporal patterns of burning across Canada and Alaska. Geophysical Research Letters 33:-. Kemper, J. T. 2005. Short and long-term effects of winter seismic exploration on low arctic plant communities of the Kendall Island Migratory Bird Sanctuary, Northwest Territories. Masters Thesis. University of Alberta, Edmonton, Alberta. Kokelj, S. V., and C. R. Burn. 2003. Ground ice and soluble cations in near-surface permafrost, Inuvik, Northwest Territories, Canada. Permafrost and Periglacial Processes 14:275-289. Kokelj, S. V., R. E. Jenkins, D. Milburn, C. R. Burn, and 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. Landhäusser, S. M., and R. W. Wein. 1993. Postfire Vegetation Recovery and Tree Establishment at the Arctic Treeline - Climate-Change Vegetation-Response Hypotheses. Journal of Ecology 81:665-672. Lantz, T. C., and S. V. Kokelj. 2008. Increasing rates of retrogressive thaw slump activity in the Mackenzie Delta region, N.W.T., Canada. Geophysical Research Letters 35:L06502, doi:06510.01029/02007GL032433. Lloyd, A. H., and C. L. Fastie. 2002. Spatial and temporal variability in the growth and climate response of treeline trees in Alaska. Climatic Change 52:481-509. 11  Mackay, J. R. 1970. Disturbances to the tundra and forest tundra environment of the western Arctic. Canadian Geotechnical Journal 7:420-432. Marsh, P., M. Russell, and C. Onclin. 2005. The hydrology of lakes in the Western Canadian Arctic: Implications to proposed natural gas development. Pages 131140 in 15th International Northern Research Basins Symposium and Workshop Water Resources Engineering, Lund University, Luleå to Kvikkjokk, Sweden. Matthews, R. F. 1992. Alnus virisis ssp. crispa. Fire Effects Information System. in. U.S. Department of Agriculture, Forest Service. Rocky Mountain Research Station, Fire Sciences Laboratory. McGuire, A. D., F. S. Chapin, J. E. Walsh, and C. Wirth. 2006. Integrated regional changes in arctic climate feedbacks: Implications for the global climate system. Annual Review of Environment and Resources 31:61-91. Meadows, D. H. 2001. Polar Bears and Three-Year-Olds on Thin Ice. . in The Global Citizen. . Murphy, P., J. Mudd, and B. Stocks. 2000. Historical fire records in the North American boreal forest. Pages 275–288 in E. Kasischke and B. Stocks, editors. Fire, climate change and carbon cycling in the North American boreal forests. . SpringerVerlag, New York. Myneni, R. B., C. D. Keeling, C. J. Tucker, G. Asrar, and R. R. Nemani. 1997. Increased plant growth in the northern high latitudes from 1981 to 1991. Nature 386:698702. Notaro, M., S. Vavrus, and Z. Y. Liu. 2007. Global vegetation and climate change due to future increases in CO2 as projected by a fully coupled model with dynamic vegetation. Journal of Climate 20:70-90. Osterkamp, T. E., and V. E. Romanovsky. 1999. Evidence for warming and thawing of discontinuous permafrost in Alaska. Permafrost and Periglacial Processes 10:1737. Parry, M. L., and Intergovernmental Panel on Climate Change. Working Group II. 2007. Climate Change 2007 : impacts, adaptation and vulnerability : contribution of Working Group II to the fourth assessment report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, U.K. ; New York. Perovich, D. K., B. Light, H. Eicken, K. F. Jones, K. Runciman, and S. V. Nghiem. 2007. Increasing solar heating of the Arctic Ocean and adjacent seas, 1979-2005: Attribution and role in the ice-albedo feedback. Geophysical Research Letters 34:-. Peterson, K. M., and W. D. Billings. 1980. Tundra Vegetational Patterns and Succession in Relation to Micro-Topography near Atkasook, Alaska. Arctic and Alpine Research 12:473-482. Pickett, S. T., and P. S. White. 1985. The ecology of natural disturbance and patch dynamics. Academic Press, Orlando, Fla. Racine, C., R. Jandt, C. Meyers, and J. Dennis. 2004. Tundra fire and vegetation change along a hillslope on the Seward Peninsula, Alaska, USA. Arctic Antarctic and Alpine Research 36:1-10. Riordan, B., D. Verbyla, and A. D. McGuire. 2006. Shrinking ponds in subarctic Alaska based on 1950-2002 remotely sensed images. Journal of Geophysical ResearchBiogeosciences 111:-. 12  Rupp, T. S., F. S. Chapin, and A. M. Starfield. 2000. Response of subarctic vegetation to transient climatic change on the Seward Peninsula in north-west Alaska. Global Change Biology 6:541-555. Shaver, G. R., J. Canadell, F. S. Chapin, J. Gurevitch, J. Harte, G. Henry, P. Ineson, S. Jonasson, J. Melillo, L. Pitelka, and L. Rustad. 2000. Global warming and terrestrial ecosystems: A conceptual framework for analysis. Bioscience 50:871882. Silapaswan, C. S., D. L. Verbyla, and A. D. McGuire. 2001. Land cover change on the Seward Peninsula: The use of remote sensing to evaluate the potential influences of climate warming on historical vegetation dynamics. Canadian Journal of Remote Sensing 27:542-554. Smithwick, E. A. H., M. G. Turner, M. C. Mack, and F. S. Chapin. 2005. Postfire soil N cycling in northern conifer forests affected by severe, stand-replacing wildfires. Ecosystems 8:163-181. Smol, J. P., and M. S. V. Douglas. 2007a. 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., and M. S. V. Douglas. 2007b. From controversy to consensus: making the case for recent climate using lake sediments. Frontiers in Ecology and the Environment 5:466-474. 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, and 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. Stow, D. A., A. Hope, D. McGuire, D. Verbyla, J. Gamon, F. Huemmrich, S. Houston, C. Racine, M. Sturm, K. Tape, L. Hinzman, K. Yoshikawa, C. Tweedie, B. Noyle, C. Silapaswan, D. Douglas, B. Griffith, G. Jia, H. Epstein, D. Walker, S. Daeschner, A. Petersen, L. M. Zhou, and R. Myneni. 2004. Remote sensing of vegetation and land-cover change in Arctic Tundra Ecosystems. Remote Sensing of Environment 89:281-308. Struzik, E. 2007. Rapid melt raises alarm. in The Edmonton Journal, Edmonton. Svoboda, J., and G. H. R. Henry. 1987. Succession in Marginal Arctic Environments. Arctic and Alpine Research 19:373-384. Tape, K., M. Sturm, and C. Racine. 2006. The evidence for shrub expansion in Northern Alaska and the Pan-Arctic. Global Change Biology 12:686-702. Thompson, C., J. Beringer, F. S. Chapin, and A. D. McGuire. 2004. Structural complexity and land-surface energy exchange along a gradient from arctic tundra to boreal forest. Journal of Vegetation Science 15:397-406. Truett, J. C., and K. Kertell. 1992. Tundra disturbance and ecosystem production: implications for impact assessment. Environmental Management 16:485-494. Vavrek, M. C., N. Fetcher, J. B. McGraw, G. R. Shaver, F. S. Chapin, and B. Bovard. 1999. Recovery of productivity and species diversity in Tussock tundra following disturbance. Arctic Antarctic and Alpine Research 31:254-258. 13  Walker, D. A., D. Cate, J. Brown, and C. Racine. 1987. Disturbnace and recovery of arctic Alaskan tundra terrain: a review of recent investigations. CRREL Report 87-11. US Army Corps of Engineers. Cold Regions Research and Engineering Laboratory, Hanover, NH. Walker, M. D., C. H. Wahren, R. D. Hollister, G. H. R. Henry, L. E. Ahlquist, J. M. Alatalo, M. S. Bret-Harte, M. P. Calef, T. V. Callaghan, A. B. Carroll, H. E. Epstein, I. S. Jonsdottir, J. A. Klein, B. Magnusson, U. Molau, S. F. Oberbauer, S. P. Rewa, C. H. Robinson, G. R. Shaver, K. N. Suding, C. C. Thompson, A. Tolvanen, O. Totland, P. L. Turner, C. E. Tweedie, P. J. Webber, and P. A. Wookey. 2006. Plant community responses to experimental warming across the tundra biome. Proceedings of the National Academy of Sciences of the United States of America 103:1342-1346. Walker, M. D., D. A. Walker, and N. A. Auerbach. 1994. Plant-Communities of a Tussock Tundra Landscape in the Brooks Range Foothills, Alaska. Journal of Vegetation Science 5:843-866. Wein, R. W. 1975. Vegetation recovery in arctic tundra and forest tundra after fire. ALUR 74-75-62, Indian and Northern Affairs Canada, Ottawa, ON. Wein, R. W., and L. C. Bliss. 1973. Changes in Arctic Eriophorum Tussock Communities Following Fire. Ecology 54:845-852. Williams, J. W., and S. T. Jackson. 2007. Novel climates, no-analog communities, and ecological surprises. Frontiers in Ecology and the Environment 5:475-482. Yoshikawa, K., W. R. Bolton, V. E. Romanovsky, M. Fukuda, and L. D. Hinzman. 2003. Impacts of wildfire on the permafrost in the boreal forests of Interior Alaska. Journal of Geophysical Research 107:8148. doi:10.1029/2001JD000438.  14  Chapter 2. Response of green alder (Alnus viridis subsp. fruticosa) and plant community composition to fire across treeline in the Mackenzie Delta Region, NWT. 1 INTRODUCTION One of the greatest challenges in understanding the impacts of a warming climate on the biosphere is predicting the direction and magnitude of feedbacks between regional vegetation and the global climate (Chapin et al. 2000a, Chapin et al. 2000b, Thompson et al. 2004, Chapin et al. 2005). Since vegetation in northern regions can exert a strong influence on both local and regional energy balance, changes in vegetation will have important feedbacks to the global climate (Chapin et al. 2000a, Chapin et al. 2005). There is a growing body of evidence that certain types of vegetation, particularly deciduous shrubs (e.g. Alnus, Betula, and Salix spp.) are becoming a more dominant component of northern ecosystems (Tape et al. 2006). Detailed analysis of repeat aerial photos of the north slope of Alaska show that tall shrub cover has increased considerably over the last 50 years (Sturm et al. 2001, Stow et al. 2004, Tape et al. 2006). Coarser scale remote sensing data, (Landsat TM and MODIS (Moderate-Resolution Imaging Spectroradiometer)) suggest that similar changes may have taken place across the western Arctic and northern hemisphere (Myneni et al. 1997, Silapaswan et al. 2001, Jia et al. 2003, Stow et al. 2004). Observed changes in shrub cover in the western Arctic may be linked to recent temperature increases (Silapaswan et al. 2001, Sturm et al. 2001, Jia et al. 2003, Tape et al. 2006), which have been among the most rapid in the world (Serreze et al. 2000, ACIA 2004, Johannessen et al. 2004). Experimental warming has also caused increases in deciduous shrub cover (Chapin et al. 1995, Bret-Harte et al. 2001, Walker et al. 2006). However, the mechanisms driving landscape level encroachment are not well understood. In high latitude ecosystems where growth and sexual reproduction in plants are limited 1  A version of this chapter will be submitted for publication. Lantz, T.C., S.G. Gergel, and G.H.R Henry. Response of green alder (Alnus viridis subsp. fruticosa) and plant community composition to fire across treeline in the Mackenzie Delta Region, NWT.  15  by short summer growing seasons, increases in temperature have the potential to favor bmany suppressed deciduous trees and tall shrubs by improving their capacity for growth and reproduction (Black and Bliss 1980, Sirois 1992, Zasada et al. 1992, Landhäusser and Wein 1993, Hobbie and Chapin 1998, Garcia et al. 2000, Bret-Harte et al. 2001, McLeod 2001, de Groot and Wein 2004). At the plot level, experimental evidence also shows that increased temperatures can promote growth of Arctic shrubs (Parsons et al. 1994, Chapin et al. 1995, Bret-Harte et al. 2001, Silapaswan et al. 2001, Bret-Harte et al. 2002, Walker et al. 2006). Across landscapes dominated by long-lived vegetation, the spread of tall shrubs may also be facilitated by disturbances such as wildfire (Bliss and Matveyeva 1992, Landhäusser and Wein 1993, Hobbie and Chapin 1998, Racine et al. 2004). Both models and empirical studies suggest that climate-driven increases in fire frequency are likely to have greater consequences for the distribution and abundance of woody species than increased temperatures alone (Brubaker 1986, Landhäusser and Wein 1993, Bergeron et al. 1998, Innes 1998, Rupp et al. 2000, Cullen et al. 2001, Johnstone and Chapin 2003, Turner et al. 2003, Stow et al. 2004). In the tundra and forest tundra, fires also expose new seedbeds and increase nutrient availability, providing new opportunities for species establishment (Zasada et al. 1983, Landhäusser and Wein 1993, Smithwick et al. 2005). In many northern ecosystems green alder (Alnus viridis ssp. frtuticosa (Ruprect) Nyman), is among the first species to colonize after disturbance and is a particularly successful invader of mineral soils and newly exposed substrates along streams, floodplains, coastlines, lake edges, and retreating glaciers (Haeussler et al. 1990, Matthews 1992, Beaudry et al. 2000). It has also been identified as the species primarily responsible for increased shrub cover visible on air photos of Alaska (Tape et al. 2006). Growing evidence that fire frequency is increasing (Flannigan and Vanwagner 1991, Weber and Flannigan 1997, Murphy et al. 2000, Hogg et al. 2002, McCoy and Burn 2005, Kasischke and Turetsky 2006) make understanding the relative effects of fire and temperature on green alder population dynamics and plant community composition critical to our understanding of vegetation change in the western Arctic and subarctic. As a result, the overall goal of this research is to evaluate the relative effects of climate and  16  fire on the population dynamics of green alder and forest-tundra plant community composition. In the upland terrain east of the Mackenzie Delta a strong summer temperature gradient driven by proximity to the Beaufort Sea creates a steep treeline ecotone (Ritchie 1984, Timoney et al. 1992, Burn 1997) across a landscape with relatively homogenous soils and subtle topography (Mackay 1963, Soil Landscapes of Canada Working Group 2007), making it an ideal location to explore the relative importance of temperature and disturbance. We have two specific objectives: Objective 1: Regional temperature, green alder abundance and seed viability. First we examine variability in summer air temperature and its relationships to green alder abundance, seed viability and population age structure across a gradient extending from the subarctic forest-tundra through the northern and southern transition zones to the low Arctic tundra (Figure 2.1). Objective 2 – Relative effects of temperature and fire on green alder and plant community composition. Second we examine green alder cover, growth and reproduction, and plant community composition on burned and unburned sites in the central portion of the treeline ecotone [northern and southern transition zones] (Figure 2.1). By sampling alder populations and plant community composition across a natural temperature gradient on both burned and unburned sites, we test the hypothesis that disturbance has a larger relative influence on green alder autecology and plant community composition than temperature alone. Specifically we expect that alder abundance, growth, sexual reproduction, and seedling establishment will be greater on burned sites compared to undisturbed controls, but will show little variation across the temperature gradient. We also expect that plant community composition will show greater differences between disturbed and undisturbed sites compared to differences between controls across a regional temperature gradient.  17  Figure 2.1. Map of the study region showing settlements, control sub-regions (green), burned subregions (red stippling), water (blue), temperature loggers (orange), and the boundaries of the treeline zones sensu Timoney et al. (1992). Intensively studied portion of the treeline ecotone includes the northern and southern transition zones. Inset map at the bottom right shows the position of the study area in North America. 18  METHODS Study Area Our study area in northwestern Canada is approximately 25000 km2 and is located between the latitudes of 67°17’N and 69°30’N and the longitudes of 133°00’W and 135°00’W (Figure 2.1). This upland terrain to the east of the Mackenzie River Delta is characterized by low rolling topography (30-150 m) and thousands of small lakes (Mackay 1963). Pleistocene surficial deposits in this region are fine grained tills derived from carbonates and shales of fluvial and deltaic origin and are characteristically ice rich (Mackay 1963, Aylsworth et al. 2000). Sites are underlain by continuous permafrost, microtopography is characterized by earth hummocks, and soils are predominantly clayey silts overlain by organic layers of variable depth (Mackay 1963, Kokelj and Burn 2004). The climate of this region is defined by long cold winters, with mean air temperatures less than 0°C from October through April (Ritchie 1984, Burn 1997). Largely because of the influence of the Beaufort Sea, temperatures during the short summer growing season vary considerably across the region (Burn 1997). This steep temperature gradient drives one of the narrowest latitudinal tree line ecotones in Canada, with spruce cover changing from more than to 90 % to less than 0.01 % in less than 100 km (Black and Bliss 1980, Timoney et al. 1992, Figure 2.1). Field Sampling In the summers of 2004 and 2005, green alder abundance, growth and reproduction and plant community composition were sampled across the treeline temperature gradient on burned and unburned sites throughout the Mackenzie Delta region. Study sites were located along a north-south transect that extended from 67°17’30” (NE of Fort MacPherson) to 69°28’54” (northern Richards Island). Sites were grouped a priori according to their position along the tree line ecotone. Treeline zones include: 1) sub-arctic forest (>99% forest cover), 2) southern forest tundra zone (99%50% forest cover), 3) northern forest tundra zone (49%-1% forest cover), and 4) Low Arctic (forest cover <0.1%). Zone limits were derived from mapping completed by Timoney et al. (1992) and were chosen because they reflect structural differences linked 19  to ecosystem function (Epstein et al. 2004, Thompson et al. 2004). To document the variability in temperature along this transect temperature loggers were established at points distributed along the north-south transect. Temperature loggers were installed in radiation shields mounted 1.5 meters above the ground surface and recorded the temperature every hour from June to August, 2005 (Onset Computer Corporation, Pocasset MA HOBOTM, H08-030-08, RS1). In order to describe sites where the vegetation had considerable time to recover post-fire, we restricted our sampling to fires initiated between 1954 and 1968 (37-51 years (Wein 1975). In the central portion of the treeline ecotone we established between 3-5 transects, within four different fires (n=16). Two of these fires were located in the southern forest-tundra and two in the northern forest tundra zone (Figure 2.1). Although our sites were generally separated by distances greater than 1 km, they were grouped within sub-regions defined by the extent of either an individual fire or undisturbed area of similar size (Figure 2.1). Within burned or undisturbed areas, transects were established haphazardly on homogenous terrain and topography. At each site we established a 200 m line transect along a random azimuth. Sites were established haphazardly At each transect in the southern and northern transition zones we collected community composition data at ten random points by visually estimating percent cover in nested square quadrats: 1) tall shrubs: 5 m2 (n=10 /transect). 2) low shrubs; and 3) dwarf shrubs, 4) herbs and 5) mosses: 0.5 m2 (n=10 /transect)). To estimate the percent cover of large trees we recorded the density and aerial ground cover (using the mean maximum radius for side branches) of all trees (> 5 cm dbh) within a 100 m2 area around each plot centre. Nomenclature for vascular plants follows Porsild and Cody (1980) and Catling et al. (2005). On burned and unburned sites in the southern and northern transition zones we also measured the growth and reproduction of green alder. Alder subplots were selected by randomly choosing up to 3 shrub subplots (5 m2) where alder cover was > 0%. To ensure that all plots with alder cover > 0% were included in randomization, we intensively searched each plot for small alders prior to selection. In all plots sampled we excavated and mapped all stems that were rooted within the plot and obtained stem cross sections from above the top of the root collar. Subsequently we dried, sanded and used a 20  dissecting microscope to recorded stem ages by counting growth rings on a minimum of two radii. The age of alder seedlings were estimated by examining stem thin sections using a compound microscope. The high degree of consistency between counts on consecutive radii, and the frequent correspondence of narrow rings within a site (Yamaguchi 1991) suggest that these age estimates are accurate to within 2-3 years. However, since these chronologies have not been statistically cross-dated they represent minimum age estimates. Between August 27 and 30, 2005 we also obtained catkin samples from a number of individuals within each latitudinal zone and disturbance type. Catkins used to test seed viability were air dried at room temperature until they released their seeds. Subsequently these seed were used in germination trails where lots of 100 seeds were placed on moist filter paper in Petri dishes. Dishes were kept moist at room temperature under 12 hours of full spectrum light for three weeks. We used these data combined with variables recorded in each plot to compare the following alder response variables: 1) alder abundance (proportion of plots with alder), 2) vertical growth (stem height / stem age), 3) radial growth (stem basal diameter / stem age), 4) catkins / m2, 5) seed viability (number of germinants / total number of seeds *100) and 6) abundance of recruits (% of stems <5 years old / per plot). Within the intensively sampled portion of the study area (Figure 2.1) we measured all of these variables on burned and unburned sites. However, we also recorded alder abundance (proportion of plots with alder), seed viability and population age structure on undisturbed sites across the entire study area (Figure 2.1). Statistical Analysis Objective 1: Regional temperature, green alder abundance and seed viability. To explore the relationship between mean summer temperature and position along the treeline ecotone (latitude) we used linear regression analysis. Given the strength of the linear relationship between air temperature and latitude across the study area (Figure 2.2: F1, 8 =293, r2 =0.9734, P< 0.001) we used this regression to interpolate the mean summer temperature at sample sites located between our temperature loggers. Subsequently, we used these collected and interpolated data points as independent variables in regression models of the proportion of sample plots with alder present and 21  seed viability. To satisfy the assumptions of equal variance and normality, the proportion of sample plots with alder present and seed viability were arcsin-square-root transformed. To test if undisturbed standing age distributions in each treeline zone were significantly different we performed a Kolmogorov-Smirnov test for all pair wise combinations of treeline zone (SAS 2004).  MEAN SUMMER TEMPERATURE (ºC)  14  12  10  * *  8  6  r2=0.973 p<0.001 4 67.5  68.0  68.5  69.0  69.5  LATITUDE  Figure 2.2. Mean growing season temperature and latitude for 10 sites in the Mackenzie delta uplands, June-August 2005 (r2 =0.973, P< 0.001, y=219.4-3.06x). Data marked with asterisks (69 º30’N, 133º34’W and 69º11’N, 134º42’W) provided by Dr. Chris Burn (Carelton University). Error bars show the 95% confidence interval of the mean.  Objective 2 – Relative effects of temperature and fire on green alder and plant community composition. Green Alder: To test for significant differences and interactions between treeline zones and disturbance history within the intensively sampled central portion of the study area, we used the PROC MIXED procedure in SAS (2004). This technique uses restricted 22  maximum likelihood to estimate variance components in a general linear model with both fixed and random effects and is particularly useful for unbalanced and spatially nested datasets (Buckley et al. 2003, Littell 2006). In our models we treated treeline zones and disturbance history as fixed effects. Our replicate transects were generally separated by distances greater than 1 km, but were clustered within sub-regions defined by the extent of an individual fire or proximity to the remote field camps we used to access undisturbed sites (Figure 2.1). Observations were also grouped by plots nested within each transect. To include this potential spatial covariance in our models we incorporated all spatial factors including: 1) plots (within transects), 2) transects and 3) sub-region (groups of transects within a burn, or undisturbed area of similar size) as random effects. In all of our models we assumed a simple co-variance structure by using the variance components option for random effects. To assess the importance of random factors in our model we tested their significance by removing terms one at a time and comparing the difference between the log likelihoods of the reduced and complete models using a chi square test (Morrell 1998). Since individual transects were consistently the only random factor which had a significant effect on our models, all analyses presented include transects as the only random factor. To estimate the error degrees of freedom for all F tests of fixed effects we used the Kenward-Rogers approximation (SAS 2004). To meet the assumptions of normality and equal variance the following response variables were log transformed: alder abundance, vertical growth, radial growth, catkins / m2, and seed viability. Plant Community Composition: To explore differences in community composition at disturbed and undisturbed sites across in the northern and southern transition zones we used PRIMER to perform an NMDS ordination of a Bray-Curtis distance matrix calculated from percent cover data (Clarke 1993). We set PRIMER to repeat this analysis 20 times and selected the best two dimensional representation of the original distance matrix (i.e. least stress) (Legendre et al. 1998). To reduce noise and stress we log(1+x) transformed percent cover data. Subsequently, we used ANOSIM to test the null hypothesis that species composition did not differ between the four site types. ANOSIM is roughly analogous to a one-way ANOVA and uses ranked Bray-Curtis dissimilarities to test for significant differences in species composition between groups 23  (Legendre et al. 1998). The RANOSIM statistic ranges from zero to one and expresses the similarity between groups. Values of RANOSIM > 0.75 indicate well separated groups, values between 0.5 and 0.75 describe overlapping but distinguishable groups, and values < 0.25 are characteristic of groups that can barely be separated (Clarke and Gorley 2001). The significance of the RANOSIM statistic was calculated by performing 9999 randomizations of the original data. To identify the species making the largest contribution to differences between site types, we used the SIMPER function in PRIMER to calculate the percent contribution of each species and species groups to the Bray-Curtis dissimilarities between site types (Clarke and Gorley 2001).  RESULTS Objective 1: Regional temperature, green alder abundance and seed viability. Moving south to north across our study area in 2005 the average June-August temperature decreased from a mean of 12.5°C north of Tsiigehtchic (67°37N, 133°45W) to a mean of 6.8°C at Illisarvik (69°30N, 134°32). This represents a decrease in mean temperature of approximately 3°C for every degree of increasing latitude, or 6°C across the entire study area (Figure 2.2, F1, 8 =293, r2 =0.9734, P< 0.001). Within the intensively sampled northern and southern transition zones temperature changes by approximately 3°C. Across the entire study area green alder abundance and seed viability decreased with lower summer temperatures at more northern sites. A linear regression of the proportion of plots with alder present and mean growing season temperature showed that alder presence declines with reduced summer temperatures at higher latitudes (Figure 2.3a). Similarly, percent germination showed a significant linear increase with increasing average temperature (Figure 2.3b). In the northern transition zone alder age distributions were highly skewed with the majority of stems in the youngest age classes (Figure 2.4). Alder age distributions were similar in the Low Arctic, southern transition zone, and subarctic where they were largely even aged (Figure 2.4). Across all sites stems ranged in age from 1 to 166 years, but the older stems were more consistently found at more 24  northern latitudes of the Low Arctic and northern transition zone (Figure 2.4). Comparisons using the Kolmogorov-Smirnov statistic show that the standing age distribution in the northern transition zone were significantly different than age distributions in all other zones (P<0.001, Table 2.1). Age distributions among other zones were not significantly different (Table 2.1). When age data were grouped by the subzones shown in Figure 2.1 the same pattern was also evident. Objective 2 – Relative effects of temperature and fire on green alder and plant community composition. Green Alder: Alder showed large increases in cover, growth, catkin production and seed viability on burned sites (Figure 2.5, Table 2.2). Except for the percentage of recruits / plot, differences between burns and controls within a zone were consistently greater than differences among controls across treeline. For example, the nine-fold increase in seed viability on burns in northern transition zone was of greater magnitude than the effects of regional temperature on controls across treeline (Figure 2.5e). Seed viability, vertical growth, and the percentage of recruits were the only variables that differed between controls across treeline (Figure 2.5, Table 2.2). Both seed viability and vertical growth were higher on controls in the southern than northern transition zone (Figure 2.5b, 2.5d Table 2.2). Conversely, the percentage of green alder recruits was significantly higher on controls in the northern than southern transition zone (Figure 2.5f). The percentage of recruits was also greater on northern compared with southern burns (Figure 2.5f). Plant community composition: Plant community composition showed similar responses to fire regardless of position along the treeline ecotone (Figure 2.6, Table 2.3). In the northern transition zone burned and unburned sites had significantly different species compositions (Figure 2.6, Tables 2.2-2.3, RAnosim = 0.58). This difference was driven primarily by an abundance of green alder and diamondleaf willow (Salix pulchra Cham.) on burned sites. These differences also made it easy to distinguish burned and unburned on aerial photos many decades following fire (Figure 2.7). On unburned control sites, willow and alder were less abundant, and dwarf birch (Betula glandulosa  25  Michx.), fruticose lichens, and northern Labrador tea (Ledum decumbens (Ait.) Lodd) generally dominated.  26  1.2  A Proportion of Plots with Alder  1.0  0.8  0.6  0.4  r2=0.33 p<0.001  0.2  0.0 6  8  10  12  14  35  B Percent Germination  30  25  20  15  10  r2=0.86 p<0.001  5  0 6  7  8  9  10  11  12  13  14  Mean Summer Temperature (°C)  Figure 2.3. Green alder characteristics measured on undisturbed controls across the entire subarctic-low arctic gradient treeline gradient. A) Proportion of plots with alder present versus average summer temperature. Data were arcsin square root transformed for regression analysis (F1, 33 = 16.08, P<0.001 r2=0.33, n=34). B) Percent seed viability vs. average summer temperature. Data were arcsin square root transformed for regression analysis (F1, 16 = 94.5, P<0.001, r2=0.86, n=17). 27  35  Low Arctic 30  Percent  25  Low Arctic Subzones  20  1  2  3  2  3  2  3  2  3  15 10 5 0 100  Northern Transition Zone  Percent  80  Northern Transition Subzones 60  1 40  20  0 50  Southern Transition Zone  Percent  40  Southern Transition Subzones 30  1 20  10  0 50  Sub-arctic Forest  Percent  40  Sub-arctic Forest Subzones 30  1 20  10  0 5  15  25  35  45  55  65  Age  75  85  95  105 115 125  5  15 25 35 45 55 65 75 85 95 105 115 125 5  Age  15 25 35 45 55 65 75 85 95 105 115 125  Age  5  15 25 35 45 55 65 75 85 95 105 115 125  Age  Figure 2.4. Green alder age distributions across the treeline ecotone. Panel of histograms at the far left shows the percentage of stems in a given age class in each treeline zone. Error bars show the 95% confidence interval of the zone mean. To the right of the histogram for each treeline zone are histograms showing the percentage of stems in a given age class when plot data are grouped by the subzones shown in Figure 2.1.  28  25  400  A  20  b  2  Catkins / m  Alder Cover (%)  b  300  b 15  a  10  a  0 10  200  a  100  5  a 0 40  B  E  b  d  8  30  b  Seed Viability (%)  Vertical Growth ( cm / year)  D  b  6  c 4  a  b  b 20  10  2  a 0 0.14  0 60  C  b b  0.10 0.08  a a  0.06  F a  50 Recruits / Plot (%)  Radial Growth (cm / year)  0.12  0.04  40  a  30 20  b  b  Control  Burn  10  0.02 0.00  0 Control  Burn  Northern Transition Zone  Control  Burn  Southern Transition Zone  Control  Burn  Northern Transition Zone  Southern Transition Zone  Figure 2.5. Green alder characteristics measured on undisturbed controls (light gray bars) and burned sites (dark gray bars) across the northern and southern transition zone. Reponses variables include: (A) percent alder cover within shrub sub-plots, (B) vertical growth (cm/year), (C) radial growth (cm/year), (D) catkins / m2, (E) percent seed viability, (F) recruits (% stems <5 years old) / plot. Bars represent mean for each site type and error bars are 95% confidence intervals of the mean (untransformed). Bars sharing the same letter are not significantly different (P≤0.05, Mixed Model ANOVA and Tukey Adjusted LSD). 29  Undisturbed controls in the northern transition zone had significantly different species composition compared with burned sites in the southern transition zone (Figure 2.6, Tables 2.2-2.3, RAnosim = 0.47). This difference also corresponded to an abundance of diamondleaf willow and lack of fruticose lichens on burned sites in the southern transition zone (Table 2.4). Undisturbed control sites in the northern and southern transition zone also had significantly different species compositions (Figure 2.6, Tables 2.2-2.3, RAnosim = 0.56). In the southern transition zone, undisturbed controls were characterized by an abundance of black spruce (Picea mariana (Mill.) B.S.P.), mosses, and green alder, while undisturbed sites in the northern transition zone are typified by fruticose lichens and dwarf birch (Table 2.4). 38-51 years following fire, sites in the southern transition zone were characterized by a similar suite of species on burns and controls. These sites all shared an abundance of green alder, diamondleaf willow, and cranberry (Vaccinium vitis-idaea L.). Pair wise comparisons between burned and control sites in the southern transition and burned sites in northern transition zone using the ANOSIM procedure all generated RAnosim statistics ≤ 0.25 (Table 2.3). These differences are also reflected in the NMDS ordination of sites, which shows that undisturbed sites in the northern transition zone are clearly separated from the other sites types, while all other site types overlap considerably (Figure 2.6).  30  3  A  2  1  0  -1 Control Burn Control Burn  -2  - Northern TZ - Northern TZ - Southern TZ - Southern TZ  -3 -2 3  -1  0  1  2  B  2  1  0  -1  Control Burn Control Burn  -2  - Northern TZ - Northern TZ - Southern TZ - Southern TZ  -3 -2  -1  0  1  2  Figure 2.6. Non-metric multidimensional scaling ordinations of plant community composition based on Bray-Curtis similarity matrix. A) Colored ellipses in the upper panel show the differences in species composition and variability between site types . Ellipses were defined using the mean NMDS scores (x and y) for each site type ± standard deviation. B) Symbols in the lower panel represent individual plots sampled in burns and undisturbed controls in the northern and southern transition zones (abbreviated TZ). 31  Figure 2.7. 2004 aerial photo showing the northern extent of a 1968 fire in the northern transition zone. Dark green vegetation on the bottom portion of the image is dominated by dense alder and willow stands. In the unburned area on the upper portion of the image the cover of alder and willow is scattered.  32  Table 2.1. Comparisons of green alder age distributions across treeline using the Kolmogorov-Smirnov statistic. Pair wise Comparison Northern Transition Zone vs. Low Arctic Northern Transition Zone vs. Southern Transition Zone Northern Transition Zone vs. Subarctic Forest Low Arctic vs. Southern Transition Zone Low Arctic vs. Subarctic Forest Southern Transition Zone vs. Subarctic Forest  P  Kolmogorov-Smirnov  D Statistic  <0.001  0.243  0.5577  <0.001  0.204  0.612  <0.001  0.196  0.504  0.235  0.087  0.19  0.553  0.062  0.126  0.238  0.2384  0.21  33  . Table 2.2. Results of the ANOVA for green alder response variables.  Abundance and Growth  Response Variable Alder Percent Cover  Vertical Growth  Radial Growth  Reproduction  Catkins / m2  Percent Seed Viability  Recruits / Plot (%)  Effect Treeline Zone Disturbance Zone * Disturbance Treeline Zone Disturbance Zone * Disturbance Treeline Zone Disturbance Zone * Disturbance Treeline Zone Disturbance Zone * Disturbance Treeline Zone Disturbance Zone * Disturbance Treeline Zone Disturbance Zone * Disturbance  P 0.8504 0.0242 0.4788 0.0009 <0.0001 0.1560 0.4500 0.0106 0.9727 0.4501 0.0106 0.9727 0.3573 0.0167 0.0143 0.0268 0.4549 0.6498  F Value 0.04 5.16 0.50 14.22 30.96 2.13 0.59 7.41 0.001 0.59 7.41 0.00 0.91 7.66 8.14 5.39 0.57 0.21  df 186 186 186 25 26 26 31 31 31 31 31 31 12 12 12 31 31 31  34  Table 2.3. Pair wise comparisons of plant community composition between effects using the ANOSIM procedure. RANOSIM values for sites that can be readily distinguished based on their species composition are shown in bold. P< 0.001 for all comparisons.  Burn  Control  Burn  Control  0.25  0.56  0.22  Burn  0.23  0.47  Control  Southern Transition Zone  0.58  Control  Burn  Northern Transition Zone  Southern Transition Zone  Northern Transition Zone  35  Table 2.4. Results of the SIMPER analysis showing the top five species or species group contribution to between group dissimilarity. Species or Abundance Abundance Species Group Site type 1 Site type 2 Control (Northern T.Z.) and Burn (Northern T.Z.) Fruticose Lichens 0.85 0.08 Betula glandulosa 0.94 0.45 Alnus viridis 0.36 0.84 Salix pulchra 0.08 0.74 Ledum decumbens 0.89 0.60 Burn (Northern T.Z.) and Control (Southern T.Z.) Picea mariana 0.01 0.9 Mosses 0.26 0.93 Alnus viridis 0.84 0.56 Salix pulchra 0.74 0.53 Vaccinium vitis-idaea 0.6 0.75 Burn (Northern T.Z.) and Burn (Southern T.Z.) Betula glandulosa 0.45 1.19 Alnus viridis 0.84 0.47 Vaccinium vitis-idaea 0.6 0.72 Salix pulchra 0.74 0.89 Moss 0.26 0.69 Control (Northern T.Z.) and Burn (Southern T.Z.) Salix alaxensis 0.08 0.89 Fruticose Lichens 0.85 0.26 Betula glandulosa 0.94 1.19 Vaccinium vitis-idaea 0.89 0.72 Mosses 0.53 0.69 Control (Northern T.Z.) and Control (Southern T.Z.) Picea mariana 0.01 0.9 Fruticose Lichens 0.85 0.52 Mosses 0.53 0.93 Betula glandulosa 0.94 0.49 Alnus viridis 0.36 0.56 Burn (Southern T.Z.) and Control (Southern T.Z.) Betula glandulosa 1.19 0.49 Mosses 0.69 0.93 Picea mariana 0.36 0.9 Alnus viridis 0.47 0.56 Salix pulchra 0.89 0.53  % Dissimilarity  Cumulative % Dissimilarity  8.78 8.73 8.07 7.77 7.63  8.78 17.51 25.58 33.35 40.98  9.5 9.09 7.84 6.8 6.73  9.5 18.59 26.43 33.23 39.96  11.58 9.53 8.49 8.18 7.75  11.58 21.11 29.6 37.78 45.53  9.71 8.7 8.03 7.69 7.19  9.71 18.41 26.44 34.13 41.32  9.72 8.63 8.16 7.12 6.09  9.72 18.35 26.51 33.63 39.72  10.26 9.03 8.9 7.57 7.11  10.26 19.29 28.19 35.76 42.87  36  DISCUSSION Regional temperature, green alder abundance and seed viability The decrease in alder abundance and seed viability with colder summer conditions at northern sites suggest that low growing season temperatures influence alder abundance by limiting the availability of viable seed. This is consistent with observational evidence that low temperatures limit reproduction in many trees and shrubs at high latitudes. In black spruce, white spruce (Picea glauca (Moench) Voss) and larch (Larix laricina (DuRoi) Koch) both seed production and viability decline with reduced temperature near their northern limits (Elliott 1979, Black and Bliss 1980, Sirois 2000, Meunier et al. 2007). Black and white spruce populations at their range margins also often persist through vegetative growth (Black and Bliss 1980, McLeod 2001). Observations of low pollen and seed viability in populations of dwarf birch at their range limit on Baffin Island and reports of low seed germination in Salix spp. and green alder in Arctic Alaska suggest that low temperature, combined with high mortality of recruits driven by low temperature, competition and predation may limit the northern extent of the majority of tall shrubs in the Low Arctic (Bliss 1958, Weis and Hermanutz 1993, Hobbie and Chapin 1998). In the Low Arctic, the southern transition zone, and the subarctic forest, even aged alder age distributions suggest that recruitment is infrequent in these populations and may occur when warm temperatures increase seed viability and there are available microsites (Gilbert and Payette 1982). Conversely, in the northern transition zone over 50% of stems originated in the last two decades. These skewed age distributions suggest that recent alder recruitment in the northern transition zone may have occurred in response to increases in mean summer conditions. Summer temperatures have increased by over two degrees from 1926-2006 (Lantz and Kokelj 2008). Using the observed relationship between temperature and recruitment to estimate recruitment rates expected under historical temperatures offers some useful points of comparison. In 2005, the average summer temperature (June-August) in the northern transition zone was 10.3˚C. This is considerably colder than the 12.3˚C average for period from 1981-2000, likely making 2005 a low recruitment year. Temperatures in the northern 37  transition zone from 1930-1950 were similar to current temperatures in the Low Arctic and to the unusually cold conditions in the northern transition zone in 2005. Thus, it is likely that recruitment in the northern transition zone from 1930-1950 would have been similar to the contemporary pattern of recruitment in the Low Arctic. Since seed viability would likely also have been similarly low (<3%) during the 1930-1950 period, it seems unlikely that alder populations in the northern transition zone would have been able to maintain the levels of recruitment observed in last two decades. Using historical temperature averages to estimate seed viability in 1930-1950 and 1970-2000 suggests that in the northern transition zone the probability of successful germination has increased by approximately 7 percent. This probable increase in seed viability offers the most reasonable explanation for the high proportion of alder recruits in the northern transition zone. Gilbert and Payette (1982) also attributed green alder population expansion in the early nineteenth century to changes in seed germination caused by climate warming in northern Quebec during that period. This explanation is also consistent with observations of increased shrub cover in at sites across the western Arctic (Tape 2004, Tape et al. 2006, Thorpe et al. 2002). Despite warmer temperatures, we likely did not observe a parallel increase in recruitment in the southern transition zone and sub-arctic forest because germination and survival in these forested sites may be limited by competition with spruce and other trees. Spruce cover in the southern transition zone is approximately 15%, but in the northern transition zone it is closer to 1%. A decrease in the number of alder recruits on warmer southern sites suggests that site specific factors (e.g., competition, microsite availability and nutrients) may have a larger impact at southern sites than temperature. Experimental evidence of others also shows that competition, and substrate and nutrient availability can limit recruitment in spite of the presence of viable seed (Hobbie and Chapin 1998). It is likely that reduced recruitment on disturbed and control sites in the southern transition zone reflect competition with spruce and other woodland species. An increase in alder vertical growth on controls in the southern transition zone is also indicative of higher competition for light than at northern sites.  38  Relative effects of temperature and fire on green alder and plant community composition. Green Alder: In the southern boreal forest green alder is often among the first species to colonize after disturbance and is a particularly successful invader of mineral soils and newly exposed substrates (Haeussler et al. 1990, Matthews 1992, Beaudry et al. 2000). Our results show that near the edge of its range, green alder abundance and reproduction are also strongly affected by disturbance. For other deciduous shrubs, burning improves growth and regeneration by increasing shoot growth and production and creating favourable microsites (Gilbert and Payette 1982, Zasada et al. 1983, de Groot and Wein 2004). Intense wildfires can also increase air and soil temperatures (Thonicke et al. 2001, Hart et al. 2005), deepen active layers (Wein and Bliss 1973, Fetcher et al. 1984, Mackay 1995, Vavrek et al. 1999, Yoshikawa et al. 2003), enhance nutrient availability by increasing nutrient mineralization (Smithwick et al. 2005), and release ions from the base of the active layer (Mackay 1963, Kokelj and Burn 2005). Consequently, the success of early successional species such as green alder on burned sites is likely related to both opportunity and an improved competitive ability when resources are abundant. Conversely, similar levels of alder percent cover, radial growth, and catkin production between treeline zones indicate that alder growth was not limited by regional temperature differences. This is reinforced by similar patterns of growth and recruitment on disturbed sites in different zones as well as experimental manipulations that show nutrient additions generally have a much larger effect on Betula spp. than warming (Chapin and Shaver 1996, van Wijk et al. 2004). Plant community composition: Similar species composition on burned and unburned sites across treeline suggests that tundra in the northern transition zone may be on a successional trajectory to communities resembling forest-tundra. Approximately 40 years post-fire, disturbed plant communities in the northern and southern transition zones most closely resemble undisturbed ‘forest-tundra like’ sites in the southern transition zone. Other observations of tundra plant community composition 22 years after a fire near Inuvik suggest that post-fire increases of tall shrubs and deciduous trees represent the initial stages of the northward movement of treeline in response to increases in regional temperature (Landhäusser and Wein 1993). Our observation that disturbed sites in the northern transition zone have not reverted to tundra, but resemble disturbed and 39  undisturbed sites in the southern transition zone are consistent with this interpretation. Burned sites in both zones had fewer evergreen shrubs including cranberry and northern Labrador tea. A number of other studies have documented post-fire increases in deciduous shrubs, coincident with decreases in stress tolerant evergreen shrubs (Tirmenstein 1990, Landhäusser and Wein 1993, Vavrek et al. 1999, Racine et al. 2004, Gucker 2005). In an examination of species composition of burns of different ages in the Inuvik area Black and Bliss (1980) reported the presence of a tall shrub phase which is replaced by spruce early succession. This is consistent with similar descriptions of succession in forest tundra in Alaska and Quebec (Foote 1983, Morneau and Payette 1989). Although mean total spruce cover (P. glauca and P. mariana) on northern burns (~10%) was less than on southern burns (~15%) and controls (~16%), its presence in both areas suggests tall-shrub communities may eventually succeed to spruce woodland as envisioned by Landhausser and Wein (1993). Implications Temperature limitation of green alder reproduction coupled with evidence of an increase in alder recruitment in the northern transition zone suggests that warming has altered green alder abundance. Changes in shrub abundance reported in other parts of the western Arctic (Tape et al. 2006) are also likely related to temperature driven increases in the probability of shrub recruitment. However, the magnitude of fire effects on reproduction indicate that disturbance may have an even greater role in mediating the spread of green alder than temperature. A single alder plant on a burned site in the northern transition zone contributes approximately 39 times the quantity of viable seed than an individual at an undisturbed site. In contrast, a 1.5˚C increase in mean summer temperature (the average temperature difference between the northern and southern transition zones) increased the availability of viable seed by only 3.3 times. As fire frequency increases at northern sites (Flannigan and Vanwagner 1991, Weber and Flannigan 1997, Murphy et al. 2000, McCoy and Burn 2005, Kasischke and Turetsky 2006, Kochtubajda et al. 2006) it is likely to facilitate tall shrub expansion by providing both available substrates and sources of viable seed. Interestingly, fossil pollen and charcoal evidence from the Brooks Range indicates that transition from herb to shrub 40  tundra in the early Holocene was concurrent with an increased fire return interval (Higuera 2006). Evidence that burned tundra sites at the southern limit of the tundra (northern transition zone) have followed a post-fire successional trajectory similar to burned forest-tundra sites also suggest that the northern movement of the boreal forest will be facilitated by increases in the frequency of disturbance. The magnitude of fire effects on alder seed production and viability raise the possibility that burns may influence undisturbed tundra by acting as sources of viable seed. Given that cold temperatures limit seed viability on undisturbed sites (Hobbie and Chapin 1998, Meunier et al. 2007), the spatial arrangement of disturbance is likely to influence landscape level patterns of recruitment. Although the plot level effects of fire are of greater magnitude, their impact at a broader spatial scale is also likely to be constrained by their size. Consequently, increases in the number and size of fires emphasize the need to model the effects of temperature and fire on recruitment in a spatially explicitly manner. Understanding the combined effects of disturbance and climate change is a critical component of efforts to predict the response of the Low Arctic to global change. In the short term, changes in tall shrub abundance in the Low Arctic will likely modify regional warming (Chapin et al. 2000a, Epstein et al. 2004, Sturm et al. 2005), alter snow pack (Sturm et al. 2005, Grogan and Jonasson 2006, Johnstone and Kokelj In Press) change active layer depth and nutrient dynamics (Rhoades et al. 2001, Schimel et al. 2004, Sturm et al. 2005, Grogan and Jonasson 2006), and modify important animal habitat (Walsh et al. 1997). In the long-term, if succession on burned sites leads to encroachment of forest-tundra into the Low Arctic, increasing fire frequency will decrease albedo, alter soil carbon sinks and drive large feedbacks to the global climate system (Chapin et al. 2000a, Chapin et al. 2005). Consequently, understanding the interaction between temperature changes and natural disturbance regimes on the short and long-term responses of vegetation will be critical in predicting the future trajectory of northern ecosystems.  41  ACKNOWLEDGEMENTS The authors would like to acknowledge the following individuals for the contribution to the development of this paper. For assistance in the field and laboratory we thank: Sahara Borgarti, Celina Gabriel, Alexis Johnson, Michael Janssen, and Stephanie Mills. We would also like to thank for Ben Gilbert and Nicholas Coops for helpful comments on drafts of this manuscript. Funding support was received from: Arctic Institute of North America (Grantin-Aid), Aurora Research Institute (Research Fellowship), Canon USA and the AAAS (Canon National Parks Science Scholarship), Global Forest Research (Research Grant GF-18-2004-211), Indian and Northern Affairs Canada (Water Resources Division and the Northern Science Training Program), Killiam Trusts (Predoctoral Fellowship), Natural Resources Canada (Polar Continental Shelf Program), Natural Sciences and Engineering Research Council of Canada (PGS-B and Northern Internship to T.C Lantz, an operating Grant to G.H.R Henry, and a Discovery Grant to Sarah Gergel).  42  LITERATURE CITED ACIA. 2004. Arctic Climate Impact Assessment: Impacts of Warming Climate. . Cambridge University Press, Cambridge. Aylsworth, J. M., M. M. Burgess, D. T. Desrochers, A. Duk-Rodkin, T. Robertson, and J. A. Traynor. 2000. Surficial geology, subsurface materials, and thaw sensitivity of sediments. Pages 41-48 in L. D. Dyke and G. R. Brooks, editors. The phsyical environment of the Mackenzie Valley, Northwest Territories: A base line for the assessment of environmental change. Geological Survey of Canada Bulletin 547. Geological Survey of Canada, Ottawa, ON. Beaudry, L., R. Coupé, C. Delong, and J. Pojar. 2000. Plant Indicator Guide for Northern British Columbia: Boreal, Sub-Boreal, and Subalpine Biogeoclimatic Zones (BWBS, SBS, SBPS, and northern ESSF). B.C. Ministry of Forests Research Branch, Victoria, B.C. Bergeron, Y., O. Engelmark, B. Harvey, H. Morin, and L. Sirois. 1998. Key issues in disturbance dynamics in boreal forests: Introduction. Journal of Vegetation Science 9:464-468. Black, R. A., and L. C. Bliss. 1980. Reproductive Ecology of Picea mariana (Mill) Bsp, at Tree Line near Inuvik, Northwest Territories, Canada. Ecological Monographs 50:331-354. Bliss, L. C. 1958. Seed germination in arctic and alpine species. Arctic 11:180-188. Bliss, L. C., and N. V. Matveyeva. 1992. Circumpolar Arctic Vegetation. Pages 59-89 in F. S. Chapin, editor. Arctic ecosystems in a changing climate : an ecophysiological perspective. Academic Press, San Diego. Bret-Harte, M. S., G. R. Shaver, and F. S. Chapin. 2002. Primary and secondary stem growth in arctic shrubs: implications for community response to environmental change. Journal of Ecology 90:251-267. Bret-Harte, M. S., G. R. Shaver, J. P. Zoerner, J. F. Johnstone, J. L. Wagner, A. S. Chavez, R. F. Gunkelman, S. C. Lippert, and J. A. Laundre. 2001. Developmental plasticity allows Betula nana to dominate tundra subjected to an altered environment. Ecology 82:18-32. Brubaker, L. B. 1986. Responses of Tree Populations to Climatic-Change. Vegetatio 67:119-130. Buckley, Y. M., D. T. Briese, and M. Rees. 2003. Demography and management of the invasive plant species Hypericum perforatum. II. Construction and use of an individual-based model to predict population dynamics and the effects of management strategies. Journal of Applied Ecology 40:494-507. Burn, C. R. 1997. Cryostratigraphy, paleogeography, and climate change during the early Holocene warm interval, western Arctic coast, Canada. Canadian Journal of Earth Sciences 34:912-925. Catling, P. M., W. J. Cody, and G. Mitrow. 2005. A compilation of additions to the flora of the continental portions of the Northwest Territories and Nunavut. in BEN. Chapin, F. S., W. Eugster, J. P. McFadden, A. H. Lynch, and D. A. Walker. 2000a. Summer differences among Arctic ecosystems in regional climate forcing. Journal of Climate 13:2002-2010.  43  Chapin, F. S., A. D. McGuire, J. Randerson, R. Pielke, D. Baldocchi, S. E. Hobbie, N. Roulet, W. Eugster, E. Kasischke, E. B. Rastetter, S. A. Zimov, and S. W. Running. 2000b. Arctic and boreal ecosystems of western North America as components of the climate system. Global Change Biology 6:211-223. Chapin, F. S., and G. R. Shaver. 1996. Physiological and growth responses of arctic plants to a field experiment simulating climatic change. Ecology 77:822-840. Chapin, F. S., G. R. Shaver, A. E. Giblin, K. J. Nadelhoffer, and J. A. Laundre. 1995. Responses of Arctic Tundra to Experimental and Observed Changes in Climate. Ecology 76:694-711. Chapin, F. S., M. Sturm, M. C. Serreze, J. P. McFadden, J. R. Key, A. H. Lloyd, A. D. McGuire, T. S. Rupp, A. H. Lynch, J. P. Schimel, J. Beringer, W. L. Chapman, H. E. Epstein, E. S. Euskirchen, L. D. Hinzman, G. Jia, C. L. Ping, K. D. Tape, C. D. C. Thompson, D. A. Walker, and J. M. Welker. 2005. Role of land-surface changes in Arctic summer warming. Science 310:657-660. Clarke, K. R. 1993. Nonparametric Multivariate Analyses of Changes in Community Structure. Australian Journal of Ecology 18:117-143. Clarke, K. R., and R. N. Gorley. 2001. Primer v5: Users Manual / Tutorial. Primer-E Ltd., Plymouth, MA. Cullen, L. E., G. H. Stewart, R. P. Duncan, and J. G. Palmer. 2001. Disturbance and climate warming influences on New Zealand Nothofagus tree-line population dynamics. Journal of Ecology 89:1061-1071. de Groot, W. J., and R. Wein. 2004. Effects of fire severity and season of burn on Betula glandulosa growth dynamics. International Journal of Wildland Fire 13:287-295. Elliott, D. L. 1979. Current Regenerative Capacity of the Northern Canadian Trees, Keewatin, Nwt, Canada - Some Preliminary-Observations. Arctic and Alpine Research 11:243-251. Epstein, H. E., J. Beringer, W. A. Gould, A. H. Lloyd, C. D. Thompson, F. S. Chapin, G. J. Michaelson, C. L. Ping, T. S. Rupp, and D. A. Walker. 2004. The nature of spatial transitions in the Arctic. Journal of Biogeography 31:1917-1933. Fetcher, N., T. F. Beatty, B. Mullinax, and D. S. Winkler. 1984. Changes in Arctic Tussock Tundra 13 Years after Fire. Ecology 65:1332-1333. Flannigan, M. D., and C. E. Vanwagner. 1991. Climate Change and Wildfire in Canada. Canadian Journal of Forest Research-Revue Canadienne De Recherche Forestiere 21:66-72. Foote, M. J. 1983. Classification, description, and dynamics of plant communities after fire in the taiga of interior Alaska. Pacific Northwest Forest and Range Experiment Station, Research Paper PNW-307. U.S.D.A. Forest Service, Portland, Oregon. Garcia, D., R. Zamora, J. M. Gomez, P. Jordano, and J. A. Hodar. 2000. Geographical variation in seed production, predation and abortion in Juniperus communis throughout its range in Europe. Journal of Ecology 88:436-446. Gilbert, H., and S. Payette. 1982. Écologie des populations d’aulne vert (Alnus crispa (Ait.) Pursh) á la limite des forêts, Québec Nordique. . Géographie physique et Quaternaire 36:109-124.  44  Grogan, P., and S. Jonasson. 2006. Ecosystem CO2 production during winter in a Swedish subarctic region: the relative importance of climate and vegetation type. Global Change Biology 12:1479-1495. Gucker, C. L. 2005. Ledum palustre. Fire Effects Information System. in. U.S. Department of Agriculture, Forest Service Rocky Mountain Research Station, Fire Sciences Laboratory. Haeussler, S., D. Coates, and J. Mather. 1990. Autecology of common plants in British Columbia : a literature review. Forestry Canada, Victoria, B.C. Hart, S. C., T. H. DeLuca, G. S. Newman, M. D. MacKenzie, and S. I. Boyle. 2005. Postfire vegetative dynamics as drivers of microbial community structure and function in forest soils. Forest Ecology and Management 220:166-184. Higuera, P. E. 2006. Late Glacial and Holocene Fire History in the Southcentral Brooks Range, Alaska: Direct and Indirect Impacts of Climatic Change on Fire Regimes. . Ph.D. Dissertation. University of Washington, Seattle, WA. Hobbie, S. E., and F. S. Chapin. 1998. An experimental test of the limits to tree establishment in Arctic tundra. . Journal of Ecology 86:449-161. Hogg, E. H., J. P. Brandt, and B. Kochtubajda. 2002. Growth and dieback of Aspen forests in northwestern Alberta, Canada, in relation to climate and insects. Canadian Journal of Forest Research-Revue Canadienne De Recherche Forestiere 32:823-832. Innes, J. L. 1998. The impact of climatic extremes on forests: an introduction. Pages 1-18 in N. Beniston and J. L. Innes, editors. The impacts of climate variability on forests. Springer, Berlin. Jia, G. S. J., H. E. Epstein, and D. A. Walker. 2003. Greening of arctic Alaska, 19812001. Geophysical Research Letters 30:-. Johannessen, O. M., L. Bengtsson, M. W. Miles, S. I. Kuzmina, V. A. Semenov, G. V. Alekseev, A. P. Nagurnyi, V. F. Zakharov, L. P. Bobylev, L. H. Pettersson, K. Hasselmann, and A. P. Cattle. 2004. Arctic climate change: observed and modelled temperature and sea-ice variability. Tellus Series a-Dynamic Meteorology and Oceanography 56:328-341. Johnstone, J. F., and F. S. Chapin. 2003. Non-equilibrium succession dynamics indicate continued northern migration of lodgepole pine. Global Change Biology 9:14011409. Johnstone, J. F., and S. V. Kokelj. In Press. Environmental conditions and vegetation recovery at abandoned-drilling mud sumps in the Mackenzie Delta region, NWT, Canada. Arctic. Kasischke, E. S., and M. R. Turetsky. 2006. Recent changes in the fire regime across the North American boreal region - Spatial and temporal patterns of burning across Canada and Alaska. Geophysical Research Letters 33:-. Kochtubajda, B., M. D. Flannigan, J. R. Gyakum, R. E. Stewart, K. A. Logan, and T. V. Nguyen. 2006. Lightning and fires in the Northwest Territories and responses to future climate change. Arctic 59:211-221. Kokelj, S. V., and C. R. Burn. 2004. Tilt of spruce trees near ice wedges, Mackenzie Delta, northwest territories, Canada. Arctic Antarctic and Alpine Research 36:615-623.  45  Kokelj, S. V., and C. R. Burn. 2005. Near-surface ground ice in sediments of the Mackenzie Delta, Northwest Territories, Canada. Permafrost and Periglacial Processes 16:291-303. Landhäusser, S. M., and R. W. Wein. 1993. Postfire Vegetation Recovery and Tree Establishment at the Arctic Treeline - Climate-Change Vegetation-Response Hypotheses. Journal of Ecology 81:665-672. Lantz, T. C., and S. V. Kokelj. 2008. Increasing rates of retrogressive thaw slump activity in the Mackenzie Delta region, N.W.T., Canada. Geophysical Research Letters 35:L06502, doi:06510.01029/02007GL032433. Legendre, P., L. Legendre, L. Legendre, and L. Legendre. 1998. Numerical ecology, 2nd English edition. Elsevier, Amsterdam ; New York. Littell, R. C. 2006. SAS for mixed models, 2nd edition. SAS Institute, Inc., Cary, N.C. Mackay, J. R. 1963. The Mackenzie Delta Area, N.W.T. Department of Mines and Technical Surveys, Ottawa, ON. Mackay, J. R. 1995. Active Layer Changes (1968 to 1993) Following the Forest-Tundra Fire near Inuvik, Nwt, Canada. Arctic and Alpine Research 27:323-336. Matthews, R. F. 1992. Alnus virisis ssp. crispa. Fire Effects Information System. in. U.S. Department of Agriculture, Forest Service. Rocky Mountain Research Station, Fire Sciences Laboratory. McCoy, V. M., and C. R. Burn. 2005. Potential alteration by climate change of the forestfire regime in the Boreal forest of central Yukon Territory. Arctic 58:276-285. McLeod, T. K. 2001. The Ecology of Picea glauca (Moench) Voss at its range limits in Northwestern Canada. University of British Columbia, Vancouver. Meunier, C., L. Sirois, and Y. Begin. 2007. Climate and Picea mariana seed maturation relationships: a multi-scale perspective. Ecological Monographs 77:361-376. Morneau, C., and S. Payette. 1989. Postfire Lichen Spruce Woodland Recovery at the Limit of the Boreal Forest in Northern Quebec. Canadian Journal of BotanyRevue Canadienne De Botanique 67:2770-2782. Morrell, C. H. 1998. Likelihood ratio testing of variance components in the linear mixedeffects model using restricted maximum likelihood. Biometrics 54:1560-1568. Murphy, P., J. Mudd, and B. Stocks. 2000. Historical fire records in the North American boreal forest. Pages 275–288 in E. Kasischke and B. Stocks, editors. Fire, climate change and carbon cycling in the North American boreal forests. . SpringerVerlag, New York. Myneni, R. B., C. D. Keeling, C. J. Tucker, G. Asrar, and R. R. Nemani. 1997. Increased plant growth in the northern high latitudes from 1981 to 1991. Nature 386:698702. Parsons, A. N., J. M. Welker, P. A. Wookey, M. C. Press, T. V. Callaghan, and J. A. Lee. 1994. Growth-Responses of 4 Sub-Arctic Dwarf Shrubs to Simulated Environmental-Change. Journal of Ecology 82:307-318. Porsild, A. E., and W. J. Cody. 1980. Vascular plants of Continental Northwest Territories, Canada. National Museum of Natural Sciences, National Museums of Canada, Ottawa. Racine, C., R. Jandt, C. Meyers, and J. Dennis. 2004. Tundra fire and vegetation change along a hillslope on the Seward Peninsula, Alaska, USA. Arctic Antarctic and Alpine Research 36:1-10. 46  Rhoades, C., H. Oskarsson, D. Binkley, and R. Stottlemyer. 2001. Alder (Alnus crispa) effects on soils vary with ecosystem type along the Agashashok River, northwest Alaska. Ecoscience 8:89-95. Ritchie, J. C. 1984. Past and present vegetation of the far northwest of Canada. University of Toronto Press, Toronto, ON. Rupp, T. S., F. S. Chapin, and A. M. Starfield. 2000. Response of subarctic vegetation to transient climatic change on the Seward Peninsula in north-west Alaska. Global Change Biology 6:541-555. SAS. 2004. SAS Version 9.1.3. Help and Documentation.. SAS Institute, Cary, NC. Schimel, J. P., C. Bilbrough, and J. A. Welker. 2004. Increased snow depth affects microbial activity and nitrogen mineralization in two Arctic tundra communities. Soil Biology & Biochemistry 36:217-227. Serreze, M. C., J. E. Walsh, F. S. Chapin, T. Osterkamp, M. Dyurgerov, V. Romanovsky, W. C. Oechel, J. Morison, T. Zhang, and R. G. Barry. 2000. Observational evidence of recent change in the northern high-latitude environment. Climatic Change 46:159-207. Silapaswan, C. S., D. L. Verbyla, and A. D. McGuire. 2001. Land cover change on the Seward Peninsula: The use of remote sensing to evaluate the potential influences of climate warming on historical vegetation dynamics. Canadian Journal of Remote Sensing 27:542-554. Sirois, L. 1992. The transition between boreal forest and tundra. Pages 197-215 in H. H. Shugart, R. Leemans, and G. B. Bonan, editors. A systems analysis of the global boreal forest. Cambridge University Press, Cambridge. Sirois, L. 2000. Spatiotemporal variation in black spruce cone and seed crops along a boreal forest-tree line transect. Canadian Journal of Forest Research-Revue Canadienne De Recherche Forestiere 30:900-909. Smithwick, E. A. H., M. G. Turner, M. C. Mack, and F. S. Chapin. 2005. Postfire soil N cycling in northern conifer forests affected by severe, stand-replacing wildfires. Ecosystems 8:163-181. Soil Landscapes of Canada Working Group. 2007. Soil Landscapes of Canada v3.1.1 (digital map and database at 1:1 million scale). in. Agriculture and Agri-Food Canada. Stow, D. A., A. Hope, D. McGuire, D. Verbyla, J. Gamon, F. Huemmrich, S. Houston, C. Racine, M. Sturm, K. Tape, L. Hinzman, K. Yoshikawa, C. Tweedie, B. Noyle, C. Silapaswan, D. Douglas, B. Griffith, G. Jia, H. Epstein, D. Walker, S. Daeschner, A. Petersen, L. M. Zhou, and R. Myneni. 2004. Remote sensing of vegetation and land-cover change in Arctic Tundra Ecosystems. Remote Sensing of Environment 89:281-308. Sturm, M., C. Racine, and K. Tape. 2001. Climate change - Increasing shrub abundance in the Arctic. Nature 411:546-547. Sturm, M., J. Schimel, G. Michaelson, J. M. Welker, S. F. Oberbauer, G. E. Liston, J. Fahnestock, and V. E. Romanovsky. 2005. Winter biological processes could help convert arctic tundra to shrubland. Bioscience 55:17-26. Tape, K., M. Sturm, and C. Racine. 2006. The evidence for shrub expansion in Northern Alaska and the Pan-Arctic. Global Change Biology 12:686-702.  47  Tape, K. D. 2004. Shrub expansion in Arctic Alaska: 50 years of change documented using aerial photography. University of Alaska Fairbanks, Fairbanks, AK. Thompson, C., J. Beringer, F. S. Chapin, and A. D. McGuire. 2004. Structural complexity and land-surface energy exchange along a gradient from arctic tundra to boreal forest. Journal of Vegetation Science 15:397-406. Thonicke, K., S. Venevsky, S. Sitch, and W. Cramer. 2001. The role of fire disturbance for global vegetation dynamics: coupling fire into a Dynamic Global Vegetation Model. Global Ecology and Biogeography 10:661-677. Thorpe, N., S. Eyegetok, N. Hakongak, and K. Elders. 2002. Nowadays it is not the same: Inuit Quajimajatuqangit, climate caribou In the Kitikmeot region of Nunavut, Canada. [eds]. in I. Krupnik and D. Jolly, editors. The earth is faster now: Indigenous observations of Arctic environmental change, pp. 198-239. Arctic Research Consortium of the United States and the Smithsonian Institution, Fairbanks, AK and Washington, D.C. Timoney, K. P., G. H. Laroi, S. C. Zoltai, and A. L. Robinson. 1992. The High SubArctic Forest-Tundra of Northwestern Canada - Position, Width, and Vegetation Gradients in Relation to Climate. Arctic 45:1-9. Tirmenstein, D. 1990. Vaccinium vitis-idaea. Fire Effects Information System. in. U.S. Department of Agriculture, Forest Service Rocky Mountain Research Station, Fire Sciences Laboratory. Turner, M. G., W. H. Romme, and D. B. Tinker. 2003. Surprises and lessons from the 1988 Yellowstone fires. Frontiers in Ecology and the Environment 1:351-358. van Wijk, M. T., K. E. Clemmensen, G. R. Shaver, M. Williams, T. V. Callaghan, F. S. Chapin, J. H. C. Cornelissen, L. Gough, S. E. Hobbie, S. Jonasson, J. A. Lee, A. Michelsen, M. C. Press, S. J. Richardson, and H. Rueth. 2004. Long-term ecosystem level experiments at Toolik Lake, Alaska, and at Abisko, Northern Sweden: generalizations and differences in ecosystem and plant type responses to global change. Global Change Biology 10:105-123. Vavrek, M. C., N. Fetcher, J. B. McGraw, G. R. Shaver, F. S. Chapin, and B. Bovard. 1999. Recovery of productivity and species diversity in Tussock tundra following disturbance. Arctic Antarctic and Alpine Research 31:254-258. Walker, M. D., C. H. Wahren, R. D. Hollister, G. H. R. Henry, L. E. Ahlquist, J. M. Alatalo, M. S. Bret-Harte, M. P. Calef, T. V. Callaghan, A. B. Carroll, H. E. Epstein, I. S. Jonsdottir, J. A. Klein, B. Magnusson, U. Molau, S. F. Oberbauer, S. P. Rewa, C. H. Robinson, G. R. Shaver, K. N. Suding, C. C. Thompson, A. Tolvanen, O. Totland, P. L. Turner, C. E. Tweedie, P. J. Webber, and P. A. Wookey. 2006. Plant community responses to experimental warming across the tundra biome. Proceedings of the National Academy of Sciences of the United States of America 103:1342-1346. Walsh, N. E., T. R. McCabe, J. M. Welker, and A. N. Parsons. 1997. Experimental manipulations of snow-depth: effects on nutrient content of caribou forage. Global Change Biology 3(Suppl. 1) 158-164. Weber, M. G., and M. D. Flannigan. 1997. Canadian boreal forest ecosystem structure and function in a chnaging climate: impact on fire regimes. Environmental Reviews 5:145-166.  48  Wein, R. W. 1975. Vegetation recovery in arctic tundra and forest tundra after fire. ALUR 74-75-62, Indian and Northern Affairs Canada, Ottawa, ON. Wein, R. W., and L. C. Bliss. 1973. Changes in Arctic Eriophorum Tussock Communities Following Fire. Ecology 54:845-852. Weis, I. M., and L. A. Hermanutz. 1993. Pollination Dynamics of Arctic Dwarf Birch (Betula glandulosa Betulaceae) and Its Role in the Loss of Seed Production. American Journal of Botany 80:1021-1027. Yamaguchi, D. K. 1991. A Simple Method for Cross-Dating Increment Cores from Living Trees. Canadian Journal of Forest Research-Revue Canadienne De Recherche Forestiere 21:414-416. Yoshikawa, K., W. R. Bolton, V. E. Romanovsky, M. Fukuda, and L. D. Hinzman. 2003. Impacts of wildfire on the permafrost in the boreal forests of Interior Alaska. Journal of Geophysical Research 107:8148. doi:10.1029/2001JD000438. Zasada, J. C., R. A. Norum, R. M. Vanveldhuizen, and C. E. Teutsch. 1983. Artificial Regeneration of Trees and Tall Shrubs in Experimentally Burned Upland Black Spruce Feather Moss Stands in Alaska. Canadian Journal of Forest ResearchRevue Canadienne De Recherche Forestiere 13:903-913. Zasada, J. C., T. L. Sharik, and M. Nygren. 1992. The reprodcutive process in boreal forest trees. in H. H. Shugart, R. Leemans, and G. B. Bonan, editors. A systems of the global boreal forest. Cambridge University Press Cambridge.  49  Chapter 3. Long-term changes in microenvironment, plant community composition, and shrub growth in retrogressive thaw slumps. 1  INTRODUCTION There is broad agreement that anthropogenic increases in greenhouse gases are altering the global climate (Oreskes 2004, Parry and Intergovernmental Panel on Climate Change. Working Group II. 2007). Climate envelope models predict that warming will result in large shifts in the distribution of many species (Iverson and Prasad 1998, 2001, Bakkenes et al. 2002). In northern regions temperatures changes have been higher than average global increases (ACIA 2004, Johannessen et al. 2004) and there is growing evidence that tundra vegetation is changing (Walther et al. 2002, Walther 2003, Stow et al. 2004, Tape et al. 2006). Equilibrium vegetation models provide a good first approximation of the nature of likely vegetation responses to climate change (Pearson and Dawson 2003), but they do not account for biotic interactions and disturbance, which may mediate species responses by altering dispersal and establishment (Davis et al. 1998, Cramer et al. 2001). Since disturbances can make new substrates available for colonization and reduce competition (Zasada et al. 1983, Zasada et al. 1992, Landhäusser and Wein 1993, White and Jentsch 2001), they are likely to have an important influence on both species and community level responses to climate change (Landhäusser and Wein 1993, Forbes et al. 2001). In forested ecosystems both empirical and modeling studies indicate that changes to disturbance regimes are likely to exert a stronger influence on vegetation dynamics than changes in climate alone (Brubaker 1986, Landhäusser and Wein 1993, Bergeron et al. 1998, Innes 1998, Rupp et al. 2000, Johnstone and Chapin 2003, Turner et al. 2003). Although disturbances in the Low Arctic are generally smaller than the large wildfires of the boreal (Walker and Walker 1991, Forbes et al. 2001), both climate change and 1  A version of this chapter will be submitted for publication. Lantz, T.C., S.V. Kokelj, S.G. Gergel, and G.H.R Henry. Long-term changes in microenvironment, plant community composition, and shrub growth in retrogressive thaw slumps.  50  industrial development are likely to increase the area affected (Forbes et al. 2001, Jorgenson et al. 2001, Holroyd and Retzer 2005, Jorgenson and Osterkamp 2005, Anisimov and Reneva 2006, Lantz and Kokelj 2008). Consequently, understanding the interaction between temperature and disturbance in the Low Arctic is critical to attempts to predict the response of Low Arctic vegetation to global climate change. One species which is likely to show increased growth and recruitment in response to both temperature and disturbance is green alder (Alnus viridis subsp. fruticosa (Ruprecht) Nyman) (Gilbert and Payette 1982, Haeussler et al. 1990, Matthews 1992, Beaudry et al. 2000). Green alder is an early successional species that often dominates mineral soils and newly exposed substrates along streams, floodplains, coastlines, lake edges, and retreating glaciers (Haeussler et al. 1990, Matthews 1992, Beaudry et al. 2000). It also been identified as the species making the greatest contribution to recent increases in shrub cover observed in Alaska (Tape 2004, Tape et al. 2006). Across the Low Arctic the degradation of ice rich permafrost on hill slopes can create areas of exposed mineral soil known as retrogressive thaw slumps (Mackay 1963, Burn and Lewkowicz 1990). These features have also been called bimodal flows, tundra mudflows, and ground ice slumps (Lambert 1972, Lewkowicz 1986, 1987) are referred to in this paper simply as thaw slumps. Thaw slumps result when ice-rich permafrost is exposed and melts (Figure 3.1). As the ice in these soils thaws during the summer, materials slump, slide, and flow down slope exposing an ice rich headwall which can degrade upslope until several hectares of terrain are affected, (Mackay 1963, Lantz and Kokelj 2008). When the headwall stabilizes, plant communities may develop on the exposed scar surfaces. Slumps are common across a temperature gradient in the Mackenzie Delta region (Ritchie 1984, Burn 1997) and they present a unique opportunity to explore the effects of temperature and disturbance on species and community level responses. Although they can also be associated with rivers, coastlines and active layer detachment slides, thaw slumps in the Mackenzie delta uplands are almost exclusively associated with slopes adjacent to tundra lakes. Approximately 8% of the lakes greater than 1 ha in the Mackenzie Delta uplands are affected by thaw slumping (Lantz and Kokelj 2008). To 51  date research in the Low Arctic on re-vegetation of slumps has been largely descriptive and has focused on the initial stages of succession (Kerfoot 1969, Lambert 1972, 1976). Although some work has been conducted in the boreal (Burn and Friele 1989, Burn 2000, Bartleman et al. 2001), little is known about the long-term effects of vegetation changes on abiotic conditions at these sites. In this paper we explore the relative influence of temperature and disturbance on tundra ecosystems and test the hypothesis that thaw slumping alters abiotic conditions, shifting plant community composition and increasing green alder (Alnus viridis subsp. fruticosa) growth and reproduction. To do so, first we describe regional variability in summer air temperature across the Low Arctic and subarctic transition zone. Second we compare abiotic conditions, plant community composition, and the response of green alder among recently active slumps, stable slumps and undisturbed tundra in both the Low Arctic and the northern subarctic transition zone.  Figure 3.1. Photo of an active retrogressive thaw slump showing visible signs of terrain slumping. Orientation of ground based line transects with respect to the slump shown by the arrow. Slump positions where point data were collected are also indicated by the letters: T (TOP), M (Middle), and B (Bottom).  52  METHODS Study Area Our study area in Mackenzie Delta Region of Northern Canada is located between the latitudes of 68°16’N and 69°33’N and the longitudes of 133°00’W and 135°0’W (Figure 3.1). This upland terrain adjacent to the Mackenzie River delta is characterized by rolling hills (30-150 m relief) and thousands of small lakes (Mackay 1963). Quaternary surficial deposits are fine grained tills derived from carbonates and shales of the Mackenzie Basin. The near-surface permafrost is typically ice-rich (Mackay 1963, Rampton 1988, Aylsworth et al. 2000). Microtopography is characterized by earth hummocks and soils are predominantly clayey silts frequently overlain by thick organics (Mackay 1963, Kokelj et al. 2007). Winters are long and cold with monthly mean air temperatures below 0°C from October to April (Ritchie 1984, Burn 1997). Across the study area average summer temperature and total annual precipitation decrease with proximity to the Beaufort Sea (Ritchie 1984, Burn 1997). Slump Sampling To compare the relative effects of regional temperature and retrogressive thaw slumping on abiotic and biotic site conditions we selected 27 thermokarst slumps from near Inuvik to northern Richards Island (Figure 3.2). Although these sites are located along a continuous temperature gradient (Ritchie 1984, Burn 1997) we divided them into two response groups intended to reflect differences in both climate and vegetation structure (Epstein et al. 2004, Thompson et al. 2004). The most southerly sites were within the northern portion of the forest-tundra transition zone where spruce cover occupies between 50% and 0.1% of the landscape (Timoney et al. 1993, Figure 3.2). Sites to the north are within the Low Arctic where spruce cover is less than 0.1% (Timoney et al. 1992, McLeod 2001). Within each latitudinal zone we selected recently active slumps (showing signs of recent activity [n=7]), stable slumps (sites where shrubby vegetation was well-established [n=7]) and undisturbed controls (n=9).  53  Figure 3.2. Map of the study region showing study sites, study sites with temperature loggers, vegetation physiogamy sensu Timoney et al. 1992. Inset map at the bottom right shows the approximate position on the study area in North America. 54  Stable slumps were selected based on the presence of woody vegetation. Ring counts of woody plant material at these sites indicate that: the stable slumps we sampled in the transition zone have been inactive for between 29-73 years (mean 56.8). In the Low Arctic, stable slumps have likely been inactive for between 20-83 years (mean =57). At each site we established a line transect from the headwall of the slump to the lake shore on a bearing that was approximately parallel with the aspect of the slump (Figure 3.1). Below the slump headwall the surface that we sampled ranged from flat to gently sloping (<10º). At each site we also established an undisturbed control on the adjacent tundra. Along these transects we collected data at fixed distances of 5 or 10 meters and at single points located at the top, middle and base of the slump (Figure 3.1). Abiotic Data In March of 2006 we recorded winter snow depth along each transect at 5 meter intervals using a graduated avalanche probe. We also measured the temperature at the base of the snow pack at 10 m intervals using thermistors attached to a wooden dowel. In late August of 2006 active layer depth was measured at the top (upper third), middle (central third), and bottom (basal third) of the slump using a graduated steel probe. We also measured active layer depths at 3 hummock tops along each undisturbed transect. To measure anion (NO3-, PO4-, SO4-) and cation (K+,Ca2+, Mn2+, NH4+, Mg2+) supply rates we installed PRSTM nutrient probes (Western Ag. Innovations Inc. Saskatoon, SK) at the majority of our slump and control sites in June of 2006. At each control transect we installed two anion and two cation probes in mineral soil on hummock tops (n=8/zone). On recently active and stable slump sites we also established two pairs of anion and cation probes in the top, middle, and bottom region of each slump (n =12/zone). Since slumps lacked hummocks, all probes were buried in exposed or near surface mineral soil. Probes were installed between June 21st and 22nd and removed between August 26th and August 27th. Following removal the probes were washed with de-ionized water and each anion and cation subsample was grouped for analysis. The charged PRS membranes attract anions or cations present in the soil solution and provide an index that is useful for the comparison of differences in plant available nutrients (Hangs et al. 2004). These data are presented in milliequivalents / cm2 exchange 55  membrane / burial period (66 days). In cases when the plant available nutrient were below laboratory detection limits we substituted missing data with values intermediate between zero and the minimum detection limits. At each location where we installed probes we also measured percent soil moisture (June and August), organic layer thickness and soil pH. Air temperatures were measured using temperature loggers established at 8 sites distributed along the north-south transect (Figure 3.2). These temperature loggers were installed in radiation shields mounted 1.5 meters above the ground surface. At four sites across the temperature gradient (Figure 3.2) we also measured ground temperatures from September 2005 to August 2006. Ground temperatures were measured using thermistors positioned at 5 cm and 100 cm depths by attaching them to a wooden dowel buried vertically in the ground. At all sites, temperatures were logged every hour (Onset Computing, Pocasset, MA, HOBOTM, TMC6-HD H08-006-04, H08-030-08RS1). Vegetation Sampling At each transect we measured plant community composition by visually estimating percent cover of all vascular and non-vascular plants in nested quadrats. Although slump lengths were variable, by positioning quadrats at 5 or 10 m intervals along each transect, we recorded species abundance in a minimum of 7 subplots per slump. The percent cover of all tall shrubs was estimated using 5 m2 square quadrats. Cover of dwarf shrubs, herbs, forbs, mosses and lichens was estimated using a 0.5 m2 plot randomly nested within the 5 m2 plot. Vascular plant nomenclature used throughout this paper follows Porsild and Cody (1980) and Catling et al. (2005). The growth and life history of green alder was measured on stable slumps and undisturbed tundra. Alder subplots were selected by randomly choosing up to 3 shrub subplots (5 m2) where alder cover was > 0%. To ensure that all plots with alder cover > 0% were included in randomization, we intensively searched each plot for small alders prior to selection. In all 5 m2 plots sampled we excavated and mapped all stems that were rooted within the plot and obtained stem cross sections from above the top of the root collar. Subsequently we dried, sanded and used a dissecting microscope to recorded stem ages by counting growth rings on a minimum of two radii. The age of alder seedlings 56  were estimated by examining stem thin sections using a compound microscope. The high degree of consistency between counts on consecutive radii, and the frequent correspondence of narrow rings within a site (Yamaguchi 1991) suggest that these age estimates are accurate to within 2-3 years. However, since these chronologies have not been statistically cross-dated they represent minimum age estimates. Between August 27 and 30, 2006 we also obtained catkin samples from a number or replicate individuals within each latitudinal zone and disturbance type. We used these data combined with variables recorded in each plot to compare two groups of alder response variables: 1) growth (vertical growth [ramet height / ramet age], radial growth [ramet basal diameter / ramet age], alder stem basal area [cm2 / m2]), and 2) recruitment [alder presence-absence], catkin production [catkins / m2], and percent germination [number of germinants / total number of seeds *100]). Catkins were air dried at room temperature until they released their seeds. Subsequently these seed were used in germination trails where lots of 100 seeds were placed on moist filter paper in Petri dishes. Dishes were kept moist at room temperature under 12 hours of full spectrum light for three weeks. Statistical Analysis To explore the relationship between mean summer temperature and position along the treeline ecotone (latitude) we used linear regression analysis. To test this dataset for first order and higher serial autocorrelation of error terms we calculated the DurbinWatson statistic using the PROC AUTOREG statement in SAS (2004). To test for significant differences and interactions between latitudinal zones, and site type (slumps and controls) we used the PROC MIXED procedure to contrast all alder response variables and all abiotic response variables except continuous ground temperature data. The PROC MIXED procedure uses maximum likelihood to estimate variance components in a general linear model containing both fixed and random effects and is particularly useful for unbalanced designs (SAS 2004). We treated latitudinal zone and site type as fixed effects and individual sites and position within the slump as random factors. For response variables that were measured every 5 or 10 m (snow depth and subnivian temperature) we used fractional slump position (distance to plot / total transect 57  distance) as a continuous random factor. For response variables that were only measured at three points in each disturbance (alder response variables, active layer depth, soil moisture, organic layer thickness, plant available nutrients and pH) we used three class variables (top, middle, and bottom), modeled as a categorical random factor. To assess the importance of random factors in our model we tested their significance by removing terms one at a time and comparing the difference between the log likelihoods of the reduced and complete models using a chi square test (Morrell 1998). In each model we only retained random terms if they significantly improved likelihood. To estimate the error degrees of freedom for all F tests of fixed effects we used the Kenward-Rogers approximation (SAS 2004). To meet the assumptions of normality and equal variance, the following response variables were log transformed: tall shrub cover, vertical growth, basal area, shoot density, stems/clone catkins/m2, and seed viability. To explore differences in community composition on undisturbed tundra with recently active and stable slumps in the Low Arctic and northern transition zones, we used PRIMER (Clarke 1993) to perform an NMDS ordination of a Bray-Curtis distance matrix calculated from percent cover data. We set PRIMER to repeat this analysis 20 times and selected the two dimensional ordination that best represented the multidimensional distance matrix (i.e. exhibited the least stress (Legendre et al. 1998)). To reduce noise and stress we log(1+x) transformed percent cover data. Subsequently, we used ANOSIM to test the null hypothesis that species composition did not differ between the site types that represent the six combinations between latitudinal zone and disturbance type. ANOSIM is roughly analogous to a one way ANOVA and uses ranked Bray-Curtis dissimilarities to test for significant differences in species composition between groups (Legendre et al. 1998). The RANOSIM statistic ranges from zero to one and expresses the similarity between groups. RANOSIM values > 0.75 are indicative of well separated groups, values between 0.5 and 0.75 describe overlapping but distinguishable groups, and values of RANOSIM less than 0.25 are characteristic of groups that can barely be separated (Clarke and Gorley 2001). The significance of the RANOSIM statistic was calculated by performing 9999 randomizations of the original data. To compare dominance of functional groups we also grouped species into: tall shrub (woody species  58  generally > 40 cm), low shrub (woody species generally < 40 cm), herbaceous (all non woody vascular plants), mosses and lichens. RESULTS Regional Temperature Conditions Data from 2005 shows a northward decrease in the average June-August temperature from 10.8°C at Inuvik (67°37N, 133°45W) to 6.8°C at the Beaufort Coast (69°30N, 134°32), representing a decrease in mean temperature of approximately 3°C for every degree of increasing latitude, or 4°C across the study area (Figure 3.3, F1, 6 F= 83.55, r2 =0.93, P< 0.001).  MEAN SUMMER TEMPERATURE (ºC)  13 12 11 10 9  *  8 7  *  6 5 68.0  68.2  68.4  68.6  68.8  69.0  69.2  69.4  69.6  LATITUDE  Figure 3.3. Mean growing season temperature and latitude for 8 sites in the Mackenzie delta uplands, June-August 2005 (r2 =0.933, P< 0.001). Error bars show the 95% confidence interval of the mean. Data indicated with asterisks (69 º 30’N, 133º 34’W and 69º11’N, 134º42’W) provided by Dr. Chris Burn (Carleton University).  59  Snow Pack, Ground Temperature and Active Layer Depth The environmental conditions in both recently active and stable slumps were markedly different than the undisturbed tundra in the Low Arctic zone and the subarctic transition zone. Snow depth, subnivian temperature, active layer depth, plant available nutrients, and tall shrub cover all showed significant differences when blocked by disturbance type and were greater on both recently active and stable disturbances compared to controls (Figures 3.4-3.6, Table 3.1). Snow depth, subnivian temperature, and active layer depth also showed significant differences when blocked by latitudinal zone (Table 3.1) and were greater on undisturbed controls in the transition zone than on controls in the Low Arctic (Figure 3.4). In the Low Arctic late winter snow pack on both recently active and stable slumps was significantly greater (P<0.001) than snow pack at controls sites in the same region, but was similar to snow depths at undisturbed sites in the transition zone (Figure 3.4a). Unlike recently active disturbances in the Low Arctic, recently disturbed sites in the transition zone had snow levels similar to controls in the same zone. However, snow pack at stable slumps in the transition zone was significantly greater (P<0.001) than controls in this zone (Figure 3.4a). In the Low Arctic zone the average temperature at the base of snow pack on stable and recently active slumps was approximately 10ºC warmer than on controls (Figure 3.4b, P<0.001). Subnivian temperature in Low Arctic slumps was similar to controls in the transition zone, which were also significantly warmer than Low Arctic controls (Figure 3.4a, P<0.001). Slumps in the transition zone were also warmer than controls (Figure 3.4a, P<0.01). Active layer depth on recently active and stable slumps was significantly (P<0.01) deeper than on undisturbed terrain in both the Low Arctic and transition zones (Figure 3.4c). Mean active layer depth in undisturbed terrain was greater in the transition zone than the Low Arctic (Figure 3.4b, P<0.05). Continuous temperature data from a smaller number of sites (n=4) shows that completion of active layer freezeback was delayed by between 51 and 139 days on recently active slumps and 28 to 120 days on stable slumps, with freezeback occurring between November 8, 2005 and April 8, 2006 (Figure 3.5).  60  100  c  A b,c Depth of Snow Pack  80  b  b  b  60  40  a  20  0  B Subnivian Temperature  -5  -10  b  b,c  c  c  b  b  b  -15  -20  a -25 140  C 120  b  b,c  Active Layer Depth  100  b  80  60  b c  a  40  a  20  0 16  Organic Layer Thickness  14 12  D  a,c a  a a  a a  a  10  c  8  a c  6 4  b  2  b  0  Control  Active  Stable  Low Arctic  Control  Active  Stable  Transition Zone  Figure 3.4. Abiotic variables measured on active slumps (Active), stable slumps (Stable) and undisturbed controls (Control) in the low arctic and the subarctic transition zone: (A) snow pack depth (cm), (B) active layer depth (cm), (C) subnivian temperature (ºC), and (D) organic layer thickness. Bars represent means for each site type and error bars are 95% confidence intervals of the mean (untransformed). Bars sharing the same letter are not significantly different (P≤0.05, Mixed Model ANOVA and Tukey Adjusted LSD). 61  Start of Freezeback (Day Number)  500 Control Active Slump Stable Slump 450  03/26/06  400  350  12/16/05  300  09/07/05  250 68°36' N South  68°57' N  69°07' N  69°20' N North  Figure 3.5. Starting date of active layer freezeback (at 100 cm depth) at 4 sites in the study area (from north to south: Jimmy Lake, Parsons Lake, Lucas Point, and Denis Lake). Grouped bars show freezeback dates at controls, active and stable slumps at a given slumps at a given site. Dates on the y-axis are in day number starting January 1, 2005. Corresponding Calendar Dates are shown at right. Nutrient Availability and Soils The soils at most recently active slumps had little to no organic layer following disturbance (Figure 3.4d). Although extremely variable, undisturbed controls were characterized by a significantly thicker soil organic layer on hummock tops. On stable slumps, leaf litter and moss accumulation contributes to thicker organic layers than at recently active slumps (P<0.01, Figure 3.4d). Although mean organic thickness was lower on stable slumps compared with controls in both the Low Arctic and the transition zone, this difference was only significant in the Low Arctic (P<0.05, Figure 3.4d). Measurements of soil moisture in June and August showed no significant differences by disturbance type or latitudinal zone (Table 3.1). Sulfate and calcium availability were significantly elevated (P<0.001) on recently active and stable disturbances regardless of position along the latitudinal gradient (Figures 3.6a-3.6b). 62  Mean sulfate concentrations were lower on stable compared to recently active disturbances, but the difference was not significant (Figure 3.6a). Conversely, plant available nitrate concentrations were similar on controls and recently active slumps, but significantly higher on stable slumps (Figure 3.6c, P<0.05). Soil pH was also elevated on recently active and stable slumps within both zones (Figure 3.6d). Although plant available K+, was significantly greater when blocked by disturbance type (Table 3.1), pairwise differences were not significant. Plant available concentrations of other soil nutrients (PO4-, Mn2+, NH4+, and Mg2+) did not differ significantly between site types (Table 3.1). Plant community composition Undisturbed tundra in both the transition zone and the Low Arctic were characterized by an abundance of dwarf shrubs including: Labrador tea (Ledum decumbens (Ait.) Lodd), cranberry (Vaccinium vitis-idaea L.), and cloudberry (Rubus chamaemorus (L.). Scrub birch (Betula glandulosa Michx.) was also abundant at these sites. In contrast, recently active slumps in both zones were generally dominated by herbaceous species, including Calamagrostis canadenis (Michx) Beauv., Arctagrostis latifolia (R.Br) Grieseb), mastodon flower (Senecio congestus (R.Br.) DC and mosses (Figure 3.7). Scrub birch also often persisted on these sites where slumping had not completely removed the vegetation. Stable slumps in both the transition zone and Low Arctic were typified by an abundance of tall shrubs including alder, willows (primarily Salix pulchra Cham. and Salix glauca L.) and scrub birch (Figure 3.7).  63  120  A  b  Sulfate Availability  100  b,d c  80  c,e 60  a,e  40  20  a 0 140  B  b  120  b,c  b b  Calcium Availability  100  a  80  a,c  60  40  20  0  6  C  b  Nitrate Availability  5  4  3  a,b 2  1  a  a  a  a,b  0 10  D D b  8  a,c  b  b,c  a  pH  6  c  bb  4  a 2  a  a,c b  0  Control  Active  Stable  Low Arctic  Control  Active  Stable  Transition Zone  Figure 3.6. Abiotic variables measured on active slumps (Active), stable slumps (Stable) and undisturbed controls (Control) in the Low Arctic and the subarctic transition zone: (A) plant available sulfate (mEq / cm2 / 63 days), (B) plant available calcium (mEq / cm2 / 63 days), (B) plant available calcium (mEq / cm2 / 63 days), (C) plant available nitrate (mEq / cm2 / 63 days), and (D) pH. Bars represent mean for each site type and error bars are 95% confidence intervals of the mean (untransformed). Bars sharing the same letter are not significantly different (P≤0.05, Mixed Model ANOVA and Tukey Adjusted LSD). 64  80 Tall Shrub Herbaceous Dwarf Shrub Moss Lichen  Percent Cover  60  40  20  0  Control  Active  Stable  Low Arctic  Control  Active  Stable  Transition Zone  Figure 3.7. Mean percent cover of plant functional groups by site type. Tall shrubs include shrubs >50cm, the herbaceous group includes all non-woody vascular plants, and the dwarf shrub group contains all woody species <50cm tall. The vegetation on recently active and stable slumps had a significantly different species composition than of the surrounding tundra (RANOSIM values 0.38.-0.98, Table 3.2). Of the 69 species or species groups we encountered at these sites, 16 (28%) were found exclusively on active or stable slumps and 12 (17%) were found only at control sites. Species that were unique to recently active slumps included: mastodon flower, Tilesius wormwood (Artemisia tilesii Ledeb), and fireweed (Epilobium angustifolium L.). Stable slumps had completely re-vegetated 29-73 years following slump stabilization, with significantly different species composition than controls (RANOSIM values 0.53.-0.98, Table 3.2). In addition to dominance by green alder and willow, these sites also had high cover of deciduous shrubs that were rare (Rosa acicularis (Lindl.), or completely absent (Ribes triste (Pall), Sheperdia canadensis (L.) Nutt., Spirea beauverdiana Schneid) in controls. Although less pronounced, there were also significant differences in species composition between recently active and stable slumps evidenced by RANOSIM values between 0.25 and 65  0.48 (Table 3.2). Conversely, plant communities on the same site type (i.e., recently active slumps, stable slumps or controls), on either side of the Low Arctic boundary were virtually indistinguishable (RANOSIM ≤ 0.15, Table 3.2). The magnitude of differences in plant community composition among site types is also shown visually in the NMDS ordination and photographs in Figures 3.8-3.9. Green alder autecology Green alder exhibited strong differences among both active and stable slumps and controls. The proportion of plots with alder present, alder vertical growth, radial growth, basal area, catkin production and seed viability on disturbed sites was significantly greater than on controls in both the Low Arctic and subarctic transition zone (Figures 3.10-3.11, Table 3.3). On stable slumps, green alder showed vertical growth, mean basal area and catkin production was between three and four times greater than controls in both the Low Arctic and transition zone, but only alder presence / absence and seed viability showed significant differences by position along the latitudinal gradient (Figures 3.103.11, Table 3.3). There were no significant interactions between zone and disturbance for any of the response variables measured (Table 3.3).  66  2  0  -2  -4  Low Arctic - Control Low Arctic - Active Slump Low Arctic - Stable Slump Transition Zone - Control Transition Zone - Active Slump Transition Zone - Stable Slump -2  -1  0  1  2  1  2  2  0  -2 Low Arctic - Control Low Arctic - Active Slump Low Arctic - Stable Slump Transition Zone - Control Transition Zone - Active Slump Transition Zone - Stable Slump  -4  -2  -1  0  Figure 3.8. Non-metric multidimensional scaling ordination of plant community composition based on Bray-Curtis similarity matrix. A) Colored ellipses in the upper panel show the differences in species composition and variability between site types. Ellipses were defined using the mean NMDS scores (x and y) for each site type ± standard deviation. B) Symbols in the lower panel represent individual plots sampled in active slumps, stable slumps, and undisturbed controls in the Low Arctic and the transition zone. 67  Figure 3.9. Photographs showing typical plant community structure on slumps and undisturbed sites in Mackenzie Delta uplands: A) thaw slump with two active sections (at left and right) and a stable section (centre), B) undisturbed tundra adjacent to a thaw slump in the low arctic (69°138’), C) herbaceous dominated community on an active slump in the low arctic (68°138’), D) stable slump covered in tall alders and willows in the low arctic (68°138’).  68  12  A b  b  -1  VERTICAL GROWTH (cmyr )  10  8  6  4  a  a  2  0 0.18  b  B  0.16  -1  RADIAL GROWTH (mmyr )  0.14  b  0.12 0.10 0.08  a  a 0.06 0.04 0.02 0.00 50  C  b b  2  -2  BASAL AREA (cm m )  40  30  20  10  a  a  0  Control  Stable  Low Arctic  Control  Stable  Transition Zone  Figure 3.10. Growth and reproduction of green alder measured on stable slumps (stable) and undisturbed controls (control) in the Low Arctic and the subarctic transition zone: (A) vertical growth (cm/year), (B) radial growth (mm/year), and (C) basal area (cm2 / m2). Bars represent mean for each site type and error bars are 95% confidence intervals of the mean (untransformed). Bars sharing the same letter are not significantly different (P≤0.05, Mixed Model ANOVA and Tukey Adjusted LSD). 69  1.2  PROPORTION OF PLOTS WITH ALDER  A c  1.0  b  0.8  b  0.6  0.4  a 0.2  0.0 1200  B  b  1000  b  -2  CATKINS (m )  800  600  400  200  a  a  0 35  c  C  PERCENT GERMINATION  30  25  20  15  b  10  b  5  a 0  Control  Stable  Low Arctic  Control  Stable  Transition Zone  Figure 3.11. Growth and reproduction of green alder measured on stable slumps (stable) and undisturbed controls (control) in the Low Arctic and the subarctic transition zone: (A) proportion of plots with alder (B) catkins / m2, (C) percent germination. Bars represent mean for each site type and error bars are 95% confidence intervals of the mean (untransformed). Bars sharing the same letter are not significantly different (P≤0.05, Mixed Model ANOVA and Tukey Adjusted LSD). 70  Table 3.1. ANOVA Table for Abiotic Response Variables. Response Variable  Snow Depth  Subnivian Temperature  Active Layer Depth  Organic Thickness Soil Moisture  pH  S04  Ca  Mg  K  P  N03  NH4  Effect Latitudinal Zone Disturbance Zone * Disturbance Latitudinal Zone Disturbance Zone * Disturbance Latitudinal Zone Disturbance Zone * Disturbance Latitudinal Zone Disturbance Zone * Disturbance Latitudinal Zone Disturbance Zone * Disturbance Latitudinal Zone Disturbance Zone * Disturbance Latitudinal Zone Disturbance Zone * Disturbance Latitudinal Zone Disturbance Zone * Disturbance Latitudinal Zone Disturbance Zone * Disturbance Latitudinal Zone Disturbance Zone * Disturbance Latitudinal Zone Disturbance Zone * Disturbance Latitudinal Zone Disturbance Zone * Disturbance Latitudinal Zone Disturbance Zone * Disturbance  Numerator degrees of freedom 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2  Denominator degrees of freedom 22 78 78 17 35 35 68 66 68 30 30 30 22 21 21 52 52 52 52 52 52 54 54 54 54 54 54 54 54 54 54 54 54 28 29 29 54 54 54  FValue P 11.38 67.15 2.38 29.52 25.97 1.00 6.42 8.66 0.59 9.41 43.71 0.26 0.23 0.40 0.11 8.41 41.79 0.46 4.63 49.93 0.33 1.08 25.77 0.39 2.37 1.58 1.23 0.30 5.30 0.28 2.21 1.96 0.39 1.70 4.51 2.6 0.084 0.60 1.41  0.0027 <0.0001 0.0995 <0.0001 <0.0001 0.38 0.014 0.0005 0.56 0.005 0.0001 0.7719 0.63 0.68 0.90 0.005 <0.0001 0.6337 0.0360 0.0001 0.7215 0.30 <0.0001 0.68 0.13 0.22 0.30 0.59 0.0079 0.76 0.14 0.15 0.68 0.2023 .0199 .0916 0.77 0.55 0.26  71  Table 3.2. Pair wise comparisons of plant community composition between site types using the ANOSIM procedure. Transition Zone and Low Arctic are abbreviated as T.Z. and L.A. Comparison Between Latitudinal Zones  Site type. Disturbance(Zone) Active Slump(L.A.)-Active Slump(T.Z.) Stable Slump(L.A.)- Stable Slump(T.Z.) Control (L.A.)-Control (T.Z.)  RANOSIM 0.124 0.125 0.15  P Value 0.072 0.91 0.009  Between Controls & Active Slumps  Control (T.Z.)-Active Slump (L.A.) Control (T.Z.)-Active Slump (T.Z.) Control (L.A.)-Active Slump (L.A.) Control (L.A.)-Active Slump (T.Z.)  0.956 0.773 0.501 0.384  0.001 0.001 0.001 0.001  Control (T.Z.)- Stable Slump (L.A.) Controls & Stable Slump Control (T.Z.)- Stable Slump (T.Z.) Control (L.A.)- Stable Slump (T.Z.) Control (L.A.)- Stable Slump (L.A.)  0.983 0.931 0.56 0.528  0.001 0.001 0.002 0.002  Active Slump (L.A.)- Stable Slump (T.Z.) Active & Stable Slump Active Slump (L.A.)- Stable Slump (L.A.) Active Slump (T.Z.)- Stable Slump (T.Z.) Stable Slump (L.A.)-Active Slump (T.Z.)  0.478 0.408 0.355 0.251  0.003 0.016 0.005 0.04  72  Table 3.3. Results of ANOVA for Alder Autecology Response Variables.  Growth  Variable Vertical Growth  Radial Growth  Stem Basal Area  Recruitment  Alder Presence  Catkins  Seed Viability  Effect Latitudinal Zone Disturbance Zone* Disturbance Latitudinal Zone Disturbance Zone* Disturbance Latitudinal Zone Disturbance Zone* Disturbance Latitudinal Zone Disturbance Zone* Disturbance Latitudinal Zone Disturbance Zone* Disturbance Latitudinal Zone Disturbance Zone* Disturbance  NumDF 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1  DenDF 33 33 33 40 40 40 32 32 32 35 35 35 95 95 95 50 74 74  FValue 0.01 172.44 0.38 1.38 76.76 0.66 0.01 26.96 0.06 15.24 24.31 0.30 2.02 42.30 2.53 7.31 56.52 0.08  ProbF 0.9285 <0.0001 0.5439 0.2464 <0.0001 0.42413 0.9224 <0.0001 0.8114 0.0004 <0.0001 0.5870 0.1590 <0.0001 0.1152 0.0093 <0.0001 0.7786  73  DISCUSSION Snow Pack, Ground Temperature and Active Layer Depth Slumps were characterized by warmer soils, deeper snow pack and delayed active layer freezeback, compared to undisturbed tundra. The thawing of ground ice in slump headwalls and the deposition of thawed materials in the scar zone (Lewkowicz 1986, Lewkovicz and Graham 1987, Robinson 2000) resulted in active layer depths that are close to double those in undisturbed tundra. The concave morphology of slumps promoted snow accumulation, which inhibited ground heat loss in winter, elevating ground temperatures measured at the surface and at depth. Latent heat associated with deep active layers and thick snow pack reduced rates of ground heat loss in thaw slumps, delaying freezeback by up to 139 days, compared to adjacent tundra. Higher thermal conductivity of the mineral soils (Goodrich 1978) probably also promoted deeper active layer development in areas affected by slumping. Tall shrub cover in stable slumps may have also contributed to the increased capture of drifting snow (Sturm et al. 2001, Pomeroy et al. 2006) and trapping of latent heat in stable slumps. Similar interactions between vegetation, snow and ground temperatures have been reported by Johnstone and Kokelj (In Press) on drilling mud slumps in the Mackenzie delta and by MacKay and Burn (2002) at an experimentally drained lake on Richards Island, NWT. Overall, slumping had a much larger influence on the soil microclimate than did air temperature differences across the latitudinal gradient. Variation in ground temperatures and active layer depth on undisturbed sites in the Delta uplands is driven primarily by climate gradients associated with proximity to the Beaufort Sea coast. In spring and early summer the persistence of sea ice cools air near the coast resulting in lower summer air temperatures (Ritchie 1984, Burn 1997), colder permafrost (Mackay 1974) and thinner active layers (Palmer 2007) in the northern part of the study area. Winter air temperatures across the study area are similar (Burn 1997, Dyke and Brooks 2000), but deeper snow pack at southern sites (Ritchie 1984, Palmer 2007, Figure 3.4a), inhibits winter heat loss from deeper active layers and results in higher subnivian temperatures in the transition zone compared to the Low Arctic (Walker et al. 1999).  74  Nutrient Availability and Soils Recent research has shown that warmer winter soil temperatures can significantly elevate nutrient mineralization and winter respiration at high latitudes (Schimel et al. 2004, Sturm et al. 2005, Grogan and Jonasson 2006). Despite warmer ground temperatures and thicker active-layers in the transition zone, we did not observe higher plant available nutrients in these soils, compared to those in the Low Arctic. Nutrient availability at controls across the region is probably similar because soils remain unfrozen for similar time periods. Although absolute ground temperatures and permafrost temperatures (Mackay 1974, Palmer 2007) are higher in the transition zone, the date of active layer freezeback in 2004 and 2005 was very similar among controls suggesting that site specific factors such as: vegetation, surface organics, and soil moisture, may be more important in determining the date of freezeback than position along the latitudinal gradient (Palmer 2007). Plant available nutrients including Ca2+ and SO4- and NO3- were significantly elevated on slumps. These increases occurred in association with the release of soluble materials from thawing permafrost in the slump headwall (Kokelj et al. 2005). There is a geochemical contrast between the active layer and underlying near-surface permafrost (Kokelj and Burn 2005) and massive segregated ice of groundwater origin may contain elevated levels of soluble materials (Mackay 1985). Significant increases in Ca2+, Mg2+, Na-, K+, SO4-, and Cl- concentrations have also been reported in slump runoff and adjacent lake water (Kokelj et al. 2005). Plant available Ca2+, and SO4- likely persist in high concentrations in slump soils, because they are among the major ions in the previously unweathered carbonate and shale tills of the Mackenzie basin (Rampton 1988) and leaching may take several decades (Kokelj et al. 2005). Our observation that nitrate concentrations were significantly increased on stable, but not active, slumps suggests that the dominance of nitrogen fixing green alder and soapberry (Sheperdia canadensis (L.) Nutt.) on these sites drives increases in nitrate availability (Rhoades et al. 2001, Rhoades et al. 2007). Increased ground temperatures and a longer period during which soils remain unfrozen may also increase the temporal window for microbial nutrient 75  mineralization in slump soils (Schimel et al. 2004, Sturm et al. 2005, Grogan and Jonasson 2006). Elevated pH on recently active and stable slumps is likely driven by increased concentration of Ca2+ and reduced organic acid deposition from thin organic layers (Bohn et al. 2001). Persistence of altered microenvironmental conditions on stable slumps Altered abiotic conditions in stable thaw slumps suggest that the effects of disturbance may persist for several centuries. Although there are a number of factors that antagonistically affect near-surface conditions in thaw slumps, our data suggest that snow pack is a critical driver. Despite increased organic thickness and dense vegetation on stable slumps, both of which should reduce ground heat flux, ground temperatures and active layer depths remained similar in recently active and stable slumps. In a stabilized slump in the central Yukon where snow pack was similar to undisturbed terrain (Burn 2000), changes in active layer depth, vegetation and soil organic thickness 40 years after stabilization suggest that this slump will return to antecedent conditions after approximately a century (Burn and Friele 1989). Conversely, deeper snow pack on stable slumps in the Mackenzie Delta uplands retard the loss of latent heat in winter and contribute to increased thaw depth, higher subnivian temperatures, and delayed active layer freezeback. Since, the concave morphology of thaw slumps will likely continue to accumulate snow for centuries, abiotic conditions will probably remain distinct from adjacent tundra for the foreseeable future. It is also possible that persistent differences in the thaw slump microenvironment are related to the recurrence of this form of thermokarst. Modelling studies indicate that the thermal disturbance caused by lakeside thaw slumps can cause the talik (layer of unfrozen ground) beneath the lake to migrate laterally. If lakeside permafrost is ice-rich, talik expansion may cause shoreline subsidence and slump re-initiation. The feedbacks between the thermal disturbance of slumping and lateral talik migration suggests that slump activity may be perpetuated as long as the adjacent permafrost remains ice-rich (Kokelj et al. In Prep). The implications of modelling are corroborated by observations of slump polycyclicity throughout the western Arctic (Wolfe et al. 2001, Lantuit and Pollard 2005, Lantuit and Pollard In Press) as well as a recent assessment which shows 76  that 98% of thaw slumps in the Mackenzie Delta uplands are multi-aged (Kokelj et al. In Prep). We have observed a small number of slumps, where vegetation resembling undisturbed tundra indicates long-term inactivity. However, in an area with over 500 thaw slumps, the rarity of such sites implies that long-term stability is uncommon. Biotic Effects Community Composition Microenvironmental differences on thaw slumps that have been stable for up to eighty years suggest that observed differences in species composition may also persist for many years to come. Short term observation of one recently active slump on Gary Island, (Lambert 1976) suggests that slump recovery leading to complete vegetative cover occurs within less than a decade. While our results are consistent with the evidence that basic functional recovery to disturbance in the Low Arctic (sensu Walker and Walker 1991) can be very rapid (Lambert 1972, 1976, Ebersole 1987, Vavrek et al. 1999) our data also show that community composition and microenvironment on slumps remain different from tundra for decades. Differences in plant community composition following disturbance to tundra have been attributed to the loss of soil seed bank containing the mature community (McGraw 1980, Gartner et al. 1983, Ebersole 1987, Vavrek et al. 1999, Johnstone and Kokelj In Press). Given the virtual absence of organic soils that would contain a seed bank, and the abundance of seed sources in undisturbed tundra both within and surrounding slumps, it is more likely that the absence of climax species reflects inferior dispersal and establishment on these nutrient enriched non-acidic surfaces (Billings and Mooney 1968, Chapin and Shaver 1996, Gough 2006). Although locally absent, many of the species unique to active and stable slumps have been described on other disturbances in the region including drained lake beds (Ovenden 1986), bulldozed sites (Hernandez 1973), drilling mud sumps, (Johnstone and Kokelj In Press), and disturbed habitats in the Mackenzie River Delta (Pearce et al. 1988). Thus, the presence of these species provides additional evidence that following severe disturbance succession in the Low Arctic follows a directional replacement model sensu Svoboda and Henry (1987).  77  Our results also suggest that variability in plant succession and community stability in Low Arctic tundra is linked to the magnitude and persistence of disturbance impacts on nutrient availability and thermal regime (Ebersole 1987, Walker et al. 1987, Walker and Walker 1991, Shirazi et al. 1998, Forbes et al. 2001). On slumps, surface soils are removed, nutrient availability increases, the thermal regime is altered, and changes in community composition persist for many years. By contrast, Kemper (2005) reported that winter seismic activity the Mackenzie Delta region that did not expose mineral soil, had few significant long-term effects on either abiotic conditions or plant species composition. Similarly, Varvek et al. (1999) reported the recovery of productivity and diversity in tundra is significantly slower on bulldozed sites where the surface organics are completely removed compared with burned sites in the same area. Persistent differences in community composition on slumps after over 50 years indicate that secondary succession proceeds very slowly on these disturbances. However, it also raises the possibility that, like the ground thermal regime, the plant communities on these sites represent alternative stable states. In a review of studies tracking revegetation on anthropogenic disturbances in the Low and High Arctic, Forbes et al. (2001) noted that 20 to 75 years post-disturbance, sites could still be readily distinguished from undisturbed sites. Our observation of the dominance of deciduous shrubs on stable slumps is also consistent with reports of shifts in dominance of plant functional groups on anthropogenically disturbed sites (Chapin and Shaver 1981, Ebersole 1987, Walker et al. 1987, Emers et al. 1995, Vavrek et al. 1999, Forbes et al. 2001, Johnstone and Kokelj In Press). Increased abundance of deciduous shrubs has also been observed after tundra fires (Landhäusser and Wein 1993, Racine et al. 2004) and in response to experimental warming and nutrient addition (Chapin et al. 1995, Graglia et al. 2001, Walker et al. 2006). Modeling studies (Kokelj et al. In Prep) indicate that the thermal disturbance in stable thaw slumps makes them extremely susceptible to reactivation. Thus, if tundra succession is as slow as some have suggested (Bliss and Wein 1972, Peterson and Billings 1980, Svoboda and Henry 1987, Walker and Walker 1991, Desforges 2000), slump polycyclicity may effectively prevent long-term succession on these sites.  78  Green Alder Our data show that thaw slumps have a much larger impact on green alder than differences in regional temperature. The proportion of plots with alder, alder growth and alder reproduction were all significantly greater on slumps than controls. However, despite the fact that summer temperature in the Low Arctic is approximately 2°C colder than the northern transition zone, we observed few differences in the response of green alder on undisturbed sites across this latitudinal gradient. The uniformity of alder growth variables across the regional temperature gradient indicates that, like other tundra plants, the growth of established individuals is constrained primarily by site specific factors such as nutrient availability and competition (Shaver et al. 1979, Shaver and Chapin 1980, Chapin 1983, Chapin et al. 1995, Gough 2006). This finding is reinforced by our observation that increases in growth and catkin production on nutrient enriched disturbances did not differ between the northern and southern portions of our study area. While differences in alder performance are partly a consequence of the opportunity for colonization presented by exposed surfaces (Forbes et al. 2001), they are also likely related to the sustained increases in plant available nutrients, soil temperatures, increased pH and thaw depth that persist on stable slumps for decades. Slump morphology, and the increased snow pack in slumps also likely provide protection from winter desiccation and snow abrasion (Billings and Bliss 1959, Sturm et al. 2001). Like many deciduous shrubs green alder is likely better adapted than stress tolerant evergreen shrubs to capitalize on the reduced competition, improved microenvironmental conditions, increased nutrient availability and non-acidic substrates on slumps (Billings and Mooney 1968, Kielland 1994, Chapin and Shaver 1996, Gough et al. 2000, McKane et al. 2002). In addition to increases in alder performance of disturbed sites we also observed reductions in the abundance of evergreen shrubs including cranberry (Vaccinium vitis-idaea L.) and northern Labrador tea (Ledum decumbens (Ait.) Lodd) on stable slumps. Decreases in the abundance of evergreens have also been observed following forest-tundra and tundra fires (Landhäusser and Wein 1993, Racine et al. 2004) and anthropogenic disturbances to tundra (Hernandez 1973, Chapin and Shaver 1981, Felix and Raynolds 1989, Emers et al. 1995, Johnstone and Kokelj In Press). Long-term observations of the response of tundra to other disturbances also show that decreases in 79  the dominance of evergreens is often associated with increases in the cover of deciduous shrubs (Landhäusser and Wein 1993, Forbes et al. 2001, Racine et al. 2004, Kemper 2005). Implications Ameliorated environmental conditions on thaw slumps, particularly pH and nutrient availability, indicate that these disturbances may significantly influence tundra vegetation. Although relatively small, slumps are very widespread across the tundra landscape and are growing in size (Robinson 2000, Lantz and Kokelj 2008, Lantuit and Pollard In Press). Manipulative experiments have shown that altering resource availability, particularly nutrients, can cause dramatic shifts in tundra vegetation (Chapin et al. 1995, Jonasson et al. 1999, Shaver et al. 2000, Graglia et al. 2001, Dormann and Woodin 2002). Increases in nutrient availability have also been described in association with other disturbances to permafrost terrain including active layer detachment slides (Kokelj and Lewkowicz 1999) tundra fires (Kokelj and Burn 2003), and seismic activity and other anthropogenic disturbances (Chapin and Shaver 1981, Walker et al. 1987, Chapin et al. 1988, Truett and Kertell 1992). It is possible that many species will move outside of their present geographic ranges by colonizing exposed substrates on disturbances such as slumps (Staniforth and Scott 1991, Wein et al. 1992, Landhäusser and Wein 1993). Thus, as its frequency increases (Holroyd and Retzer 2005, Jorgenson and Osterkamp 2005, Anisimov and Reneva 2006, Lantz and Kokelj 2008), disturbance in the Low Arctic will likely have a growing influence on vegetation dynamics. The magnitude of the effects of retrogressive thaw slumping on alder reproduction and plant community composition is also consistent with other lines of evidence showing that disturbance has a larger impact on ecosystems than temperature alone (Bergeron et al. 1998, Rupp et al. 2000, Cullen et al. 2001, Turner et al. 2003, Stow et al. 2004). Our data show that a single alder plant on a slump in the northern portion of our study region contributes approximately 82 times the quantity of viable seed of an individual growing at an undisturbed site in the same area. In contrast to the effects of disturbance, an increase of 2˚C in mean summer temperature (equivalent to the average temperature difference between the two zones) may have only increased the availability 80  of viable seed 6-fold. Several Arctic grasses also show increased seed viability on slumps in the Mackenzie Delta region (Trimble 2008). Overall, this suggests that disturbed sites may likely act as seed sources for large areas of undisturbed terrain. Forbes et al. (2001) have described Arctic disturbances as “dynamic focal points,” stressing that the persistent alternative stable states that follow disturbance are likely to exert a long-term influence on a range of biotic and abiotic processes in the terrain surrounding disturbances. Tall shrubs like green alder can significantly increase snow depth, active layer depth, and nutrient availability, (Rhoades et al. 2001, Sturm et al. 2001, Epstein et al. 2004, Schimel et al. 2004, Sturm et al. 2005, Grogan and Jonasson 2006). Thus, changes in their population ecology catalyzed by increases in thermokarst activity may also lead to changes in permafrost conditions, wildlife habitat and ecosystem function (Walsh et al. 1997, Forbes et al. 2001, McGuire et al. 2006). Since vegetation exerts strong controls on regional ecosystem processes (Chapin et al. 2000a, Chapin et al. 2000b, Thompson et al. 2004), and modifies important animal habitat (Walsh et al. 1997) understanding the effects of disturbance on short-term and long-term successional trajectories is also a critical component of efforts to understand global climate change in the Arctic.  ACKNOWLEDGEMENTS The authors would like to acknowledge the following individuals for the contribution to the development of this paper. For assistance in the field and laboratory we thank: Selina Agbayani, Sahara Borgarti, Douglas Esagok, Celina Gabriel, Michael Janssen, Robert Jenkins, Alexis Johnson, Stephanie Mills, Peter Morse, Michael Palmer, Pippa Secombe-Hett, Rufus Tingmiak, and Anika Trimble. We would also like to thank for Isla Myers-Smith and Nicholas Coops for helpful comments on drafts of this manuscript. Funding support was received from: Arctic Institute of North America (Grant-inAid), Aurora Research Institute (Research Fellowship), Canon USA and the AAAS (Canon National Parks Science Scholarship), Global Forest Research (Research Grant GF-18-2004-210), Indian and Northern Affairs Canada (Water Resources Division and 81  the Northern Science Training Program), Killiam Trusts (Predoctoral Fellowship), Natural Resources Canada (Polar Continental Shelf Program), Natural Sciences and Engineering Research Council of Canada (PGS-B and Northern Internship to T.C Lantz, and an operating Grant to G.H.R Henry).  82  LITERATURE CITED ACIA. 2004. Arctic Climate Impact Assessment: Impacts of Warming Climate. . Cambridge University Press, Cambridge. Anisimov, O., and S. Reneva. 2006. Permafrost and changing climate: The Russian perspective. Ambio 35:169-175. Aylsworth, J. M., M. M. Burgess, D. T. Desrochers, A. Duk-Rodkin, T. Robertson, and J. A. Traynor. 2000. Surficial geology, subsurface materials, and thaw sensitivity of sediments. Pages 41-48 in L. D. Dyke and G. R. Brooks, editors. The phsyical environment of the Mackenzie Valley, Northwest Territories: A base line for the assessment of environmental change. Geological Survey of Canada Bulletin 547. Geological Survey of Canada, Ottawa, ON. Bakkenes, M., J. R. M. Alkemade, F. Ihle, R. Leemans, and J. B. Latour. 2002. Assessing effects of forecasted climate change on the diversity and distribution of European higher plants for 2050. Global Change Biology 8:390-407. Bartleman, A. P., K. Miyanishi, C. R. Burn, and M. M. Cote. 2001. Development of vegetation communities in a retrogressive thaw slump near Mayo, Yukon Territory: A 10-year assessment. Arctic 54:149-156. Beaudry, L., R. Coupé, C. Delong, and J. Pojar. 2000. Plant Indicator Guide for Northern British Columbia: Boreal, Sub-Boreal, and Subalpine Biogeoclimatic Zones (BWBS, SBS, SBPS, and northern ESSF). B.C. Ministry of Forests Research Branch, Victoria, B.C. Bergeron, Y., O. Engelmark, B. Harvey, H. Morin, and L. Sirois. 1998. Key issues in disturbance dynamics in boreal forests: Introduction. Journal of Vegetation Science 9:464-468. Billings, W. D., and L. C. Bliss. 1959 An Alpine Snowbank Environment and Its Effects on Vegetation, Plant Development, and Productivity. Ecology 40:388-397. Billings, W. D., and H. A. Mooney. 1968. Ecology of Arctic and Alpine Plants. Biological Reviews of the Cambridge Philosophical Society 43:481-&. Bliss, L. C., and R. W. Wein. 1972. Plant Community Responses to Disturbances in Western Canadian Arctic. Canadian Journal of Botany 50:1097-&. Bohn, H. L., B. L. McNeal, and G. A. O'Connor. 2001. Soil chemistry, 3rd edition. Wiley, New York. Brubaker, L. B. 1986. Responses of Tree Populations to Climatic-Change. Vegetatio 67:119-130. Burn, C. R. 1997. Cryostratigraphy, paleogeography, and climate change during the early Holocene warm interval, western Arctic coast, Canada. Canadian Journal of Earth Sciences 34:912-925. Burn, C. R. 2000. The thermal regime of a retrogressive thaw slump near Mayo, Yukon Territory. Canadian Journal of Earth Sciences 37:967-981. Burn, C. R., and P. A. Friele. 1989. Geomorphology, Vegetation Succession, Soil Characteristics and Permafrost in Retrogressive Thaw Slumps near Mayo, Yukon Territory. Arctic 42:31-40. Burn, C. R., and A. G. Lewkowicz. 1990. Retrogressive thaw slumps. Canadian Geographer 34:273-276.  83  Catling, P. M., W. J. Cody, and G. Mitrow. 2005. A compilation of additions to the flora of the continental portions of the Northwest Territories and Nunavut. in BEN. Chapin, F. S. 1983. Direct and Indirect Effects of Temperature on Arctic Plants. Polar Biology 2:47-52. Chapin, F. S., W. Eugster, J. P. McFadden, A. H. Lynch, and D. A. Walker. 2000a. Summer differences among Arctic ecosystems in regional climate forcing. Journal of Climate 13:2002-2010. Chapin, F. S., N. Fetcher, K. Kielland, K. R. Everett, and A. E. Linkins. 1988. Productivity and Nutrient Cycling of Alaskan Tundra - Enhancement by Flowing Soil-Water. Ecology 69:693-702. Chapin, F. S., A. D. McGuire, J. Randerson, R. Pielke, D. Baldocchi, S. E. Hobbie, N. Roulet, W. Eugster, E. Kasischke, E. B. Rastetter, S. A. Zimov, and S. W. Running. 2000b. Arctic and boreal ecosystems of western North America as components of the climate system. Global Change Biology 6:211-223. Chapin, F. S., and G. R. Shaver. 1981. Changes in Soil Properties and Vegetation Following Disturbance of Alaskan Arctic Tundra. Journal of Applied Ecology 18:605-617. Chapin, F. S., and G. R. Shaver. 1996. Physiological and growth responses of arctic plants to a field experiment simulating climatic change. Ecology 77:822-840. Chapin, F. S., G. R. Shaver, A. E. Giblin, K. J. Nadelhoffer, and J. A. Laundre. 1995. Responses of Arctic Tundra to Experimental and Observed Changes in Climate. Ecology 76:694-711. Clarke, K. R. 1993. Nonparametric Multivariate Analyses of Changes in Community Structure. Australian Journal of Ecology 18:117-143. Clarke, K. R., and R. N. Gorley. 2001. Primer v5: Users Manual / Tutorial. Primer-E Ltd., Plymouth, MA. Cramer, W., A. Bondeau, F. I. Woodward, I. C. Prentice, R. A. Betts, V. Brovkin, P. M. Cox, V. Fisher, J. A. Foley, A. D. Friend, C. Kucharik, M. R. Lomas, N. Ramankutty, S. Sitch, B. Smith, A. White, and C. Young-Molling. 2001. Global response of terrestrial ecosystem structure and function to CO2 and climate change: results from six dynamic global vegetation models. Global Change Biology 7:357-373. Cullen, L. E., G. H. Stewart, R. P. Duncan, and J. G. Palmer. 2001. Disturbance and climate warming influences on New Zealand Nothofagus tree-line population dynamics. Journal of Ecology 89:1061-1071. Davis, A. J., L. S. Jenkinson, J. H. Lawton, B. Shorrocks, and S. Wood. 1998. Making mistakes when predicting shifts in species range in response to global warming. Nature 391:783-786. Desforges, M. 2000. Plant Succession after active layer detachment slides, in a High Arctic tundra, Fosheim Peninsula, Ellesmere Island, Canada. University of British Columbia, Vancouver, B.C. Dormann, C. F., and S. J. Woodin. 2002. Climate change in the Arctic: using plant functional types in a meta-analysis of field experiments. Functional Ecology 16:417.  84  Dyke, L. D., and G. R. Brooks. 2000. The phsyical environment of the Mackenzie Valley, Northwest Territories: A base line for the assessment of environmental change. Geological Survey of Canada Bulletin. Ebersole, J. J. 1987. Short-term vegetation recovery at an Alaskan arctic coastal plain site Arctic and Alpine Research 19:442-450. Emers, M., J. C. Jorgenson, and M. K. Raynolds. 1995. Response of Arctic Tundra PlantCommunities to Winter Vehicle Disturbance. Canadian Journal of Botany-Revue Canadienne De Botanique 73:905-917. Epstein, H. E., J. Beringer, W. A. Gould, A. H. Lloyd, C. D. Thompson, F. S. Chapin, G. J. Michaelson, C. L. Ping, T. S. Rupp, and D. A. Walker. 2004. The nature of spatial transitions in the Arctic. Journal of Biogeography 31:1917-1933. Felix, N. A., and M. K. Raynolds. 1989. The Effects of Winter Seismic Trails on Tundra Vegetation in Northeastern Alaska, USA. Arctic and Alpine Research 21:188202. Forbes, B. C., J. J. Ebersole, and B. Strandberg. 2001. Anthropogenic disturbance and patch dynamics in circumpolar arctic ecosystems. Conservation Biology 15:954969. Gartner, B. L., F. S. Chapin, and G. R. Shaver. 1983. Demographic Patterns of Seedling Establishment and Growth of Native Graminoids in an Alaskan Tundra Disturbance. Journal of Applied Ecology 20:965-980. Goodrich, L. E. 1978. Some results of a numerical study of ground thermal regimes. Pages 29-34 in Proceedings, Third International Conference on Permafrost. National Research Council, Ottawa, On, Edmonton, Alberta. Gough, L. 2006. Neighbor effects on germination, survival, and growth in two arctic tundra plant communities. Ecography 29:44-56. Gough, L., G. R. Shaver, J. Carroll, D. L. Royer, and J. A. Laundre. 2000. Vascular plant species richness in Alaskan arctic tundra: the importance of soil pH. Journal of Ecology 88:54-66. Graglia, E., S. Jonasson, A. Michelsen, I. K. Schmidt, M. Havstrom, and L. Gustavsson. 2001. Effects of environmental perturbations on abundance of subarctic plants after three, seven and ten years of treatments. Ecography 24:5-12. Grogan, P., and S. Jonasson. 2006. Ecosystem CO2 production during winter in a Swedish subarctic region: the relative importance of climate and vegetation type. Global Change Biology 12:1479-1495. Haeussler, S., D. Coates, and J. Mather. 1990. Autecology of common plants in British Columbia : a literature review. Forestry Canada, Victoria, B.C. Hangs, R. D., K. J. Greer, and C. A. Sulewski. 2004. The effect of interspecific competition on conifer seedling growth and nitrogen availability measured using ion-exchange membranes. Canadian Journal of Forest Research-Revue Canadienne De Recherche Forestiere 34:754-761. Hernandez, H. 1973. Natural Plant Recolonization of Surficial Disturbances, Tuktoyaktuk Peninsula Region, Northwest Territories. Canadian Journal of Botany-Revue Canadienne De Botanique 51:2177-2196. Holroyd, P., and H. Retzer. 2005. A Peak into the Future: The Potential Landscape Impacts from Gas Development in Northern Canada. The Pembina Institute for Appropriate Development, Calgary, AB. 85  Innes, J. L. 1998. The impact of climatic extremes on forests: an introduction. Pages 1-18 in N. Beniston and J. L. Innes, editors. The impacts of climate variability on forests. Springer, Berlin. Iverson, L. R., and A. M. Prasad. 1998. Predicting abundance of 80 tree species following climate change in the eastern United States. Ecological Monographs 68:465-485. Iverson, L. R., and A. M. Prasad. 2001. Potential changes in tree species richness and forest community types following climate change. Ecosystems 4:186-199. Johannessen, O. M., L. Bengtsson, M. W. Miles, S. I. Kuzmina, V. A. Semenov, G. V. Alekseev, A. P. Nagurnyi, V. F. Zakharov, L. P. Bobylev, L. H. Pettersson, K. Hasselmann, and A. P. Cattle. 2004. Arctic climate change: observed and modelled temperature and sea-ice variability. Tellus Series a-Dynamic Meteorology and Oceanography 56:328-341. Johnstone, J. F., and F. S. Chapin. 2003. Non-equilibrium succession dynamics indicate continued northern migration of lodgepole pine. Global Change Biology 9:14011409. Johnstone, J. F., and S. V. Kokelj. In Press. Environmental conditions and vegetation recovery at abandoned-drilling mud sumps in the Mackenzie Delta region, NWT, Canada. Arctic. Jonasson, S., A. Michelsen, I. K. Schmidt, and E. V. Nielsen. 1999. Responses in microbes and plants to changed temperature, nutrient, and light regimes in the arctic. Ecology 80:1828-1843. Jorgenson, M. T., and T. E. Osterkamp. 2005. Response of boreal ecosystems to varying modes of permafrost degradation. Canadian Journal of Forest Research-Revue Canadienne De Recherche Forestiere 35:2100-2111. Jorgenson, M. T., C. H. Racine, J. C. Walters, and T. E. Osterkamp. 2001. Permafrost degradation and ecological changes associated with a warming climate in central Alaska. Climatic Change 48:551-579. Kemper, J. T. 2005. Short and long-term effects of winter seismic exploration on low arctic plant communities of the Kendall Island Migratory Bird Sanctuary, Northwest Territories. Masters Thesis. University of Alberta, Edmonton, Alberta. Kerfoot, D. E. 1969. The geomorphology and permafrost conditions of Garry Island, N.W.T. Ph.D. Dissertation. University of British Columbia, Vancouver, B.C. . Kielland, K. 1994. Amino-Acid-Absorption by Arctic Plants - Implications for Plant Nutrition and Nitrogen Cycling. Ecology 75:2373-2383. Kokelj, S. V., and C. R. Burn. 2003. Ground ice and soluble cations in near-surface permafrost, Inuvik, Northwest Territories, Canada. Permafrost and Periglacial Processes 14:275-289. Kokelj, S. V., and C. R. Burn. 2005. Geochemistry of the active layer and near-surface permafrost, Mackenzie delta region, Northwest Territories, Canada. Canadian Journal of Earth Sciences 42:37-48. Kokelj, S. V., C. R. Burn, and C. Tarnocai. 2007. The structure and dynamics of earth hummocks in the subarctic forest near Inuvik, Northwest Territories, Canada. Arctic Antarctic and Alpine Research 39:99-109. Kokelj, S. V., R. E. Jenkins, D. Milburn, C. R. Burn, and N. Snow. 2005. The influence of thermokarst disturbance on the water quality of small upland lakes, Mackenzie  86  Delta Region, Northwest Territories, Canada. Permafrost and Periglacial Processes 16:343-353. Kokelj, S. V., T. C. Lantz, J. Kanigan, S. S. Smith, and R. Coutts. In Prep. On the origin and polycyclic behaviour of tundra thaw slumps, Mackenzie Delta region, Northwest Territories, Canada. Kokelj, S. V., and A. G. Lewkowicz. 1999. Salinization of permafrost terrain due to natural geomorphic disturbance, Fosheim Peninsula, Ellesmere Island. Arctic 52:372-385. Lambert, J. D. H. 1972. Plant Succession on Tundra Mudflows - Preliminary Observations. Arctic 25:99-&. Lambert, J. D. H. 1976. Plant Succession on an Active Tundra Mud Slump, Garry-Island, Mackenzie-River-Delta, Northwest-Territories. Canadian Journal of BotanyRevue Canadienne De Botanique 54:1750-1758. Landhäusser, S. M., and R. W. Wein. 1993. Postfire Vegetation Recovery and Tree Establishment at the Arctic Treeline - Climate-Change Vegetation-Response Hypotheses. Journal of Ecology 81:665-672. Lantuit, H., and W. H. Pollard. 2005. Temporal stereophotogrammetric analysis of retrogressive thaw slumps on Herschel Island, Yukon territory. Natural Hazards and Earth System Sciences 5:413-423. Lantuit, H., and W. H. Pollard. In Press. Fifty years of coastal erosion and retrogressive thaw slumpactivity on Herschel Island, southern Beaufort Sea, Yukon Territory, Canada. Geomorphology. Lantz, T. C., and S. V. Kokelj. 2008. Increasing rates of retrogressive thaw slump activity in the Mackenzie Delta region, N.W.T., Canada. Geophysical Research Letters 35:L06502, doi:06510.01029/02007GL032433. Legendre, P., L. Legendre, L. Legendre, and L. Legendre. 1998. Numerical ecology, 2nd English edition. Elsevier, Amsterdam ; New York. Lewkovicz, A. G., and P. C. Graham. 1987. Hydrologic Processes in a Ground Ice Slump, Bank Island, Western Canadian Arctic. in M. Pécsl and H. M. French, editors. Loess and Periglacial Phenomena. Akadémiai Kladó, Budapest. Lewkowicz, A. G. 1986. Rate of Short-Term Ablation of Exposed Ground Ice, Banks Island, Northwest Territories, Canada. Journal of Glaciology 32:511-519. Lewkowicz, A. G. 1987. Headwall Retreat of Ground-Ice Slumps, Banks-Island, Northwest-Territories. Canadian Journal of Earth Sciences 24:1077-1085. Mackay, J. R. 1963. The Mackenzie Delta Area, N.W.T. Department of Mines and Technical Surveys, Ottawa, ON. Mackay, J. R. 1974. Seismic shot holes and ground temperatures, Mackenzie delta area, N.W.T. . Geological Survey of Canada, Paper 74:389-390. Mackay, J. R. 1985. Pingo Ice of the Western Arctic Coast, Canada. Canadian Journal of Earth Sciences 22:1452-1464. Mackay, J. R., and C. R. Burn. 2002. The first 20 years (1978-1979 to 1998-1999) of icewedge growth at the Illisarvik experimental drained lake site, western Arctic coast, Canada. Canadian Journal of Earth Sciences 39:95-111. Matthews, R. F. 1992. Alnus virisis ssp. crispa. Fire Effects Information System. in. U.S. Department of Agriculture, Forest Service. Rocky Mountain Research Station, Fire Sciences Laboratory. 87  McGraw, J. B. 1980. Seed bank size and distribution of seeds in cottongrass tussock tundra, Eagle Creek, Alaska. Canadian Journal of Botany-Revue Canadienne De Botanique 58:1607-1611. McGuire, A. D., F. S. Chapin, J. E. Walsh, and C. Wirth. 2006. Integrated regional changes in arctic climate feedbacks: Implications for the global climate system. Annual Review of Environment and Resources 31:61-91. McKane, R. B., L. C. Johnson, G. R. Shaver, K. J. Nadelhoffer, E. B. Rastetter, B. Fry, A. E. Giblin, K. Kielland, B. L. Kwiatkowski, J. A. Laundre, and G. Murray. 2002. Resource-based niches provide a basis for plant species diversity and dominance in arctic tundra. Nature 415:68-71. McLeod, T. K. 2001. The Ecology of Picea glauca (Moench) Voss at its range limits in Northwestern Canada. University of British Columbia, Vancouver. Morrell, C. H. 1998. Likelihood ratio testing of variance components in the linear mixedeffects model using restricted maximum likelihood. Biometrics 54:1560-1568. Oreskes, N. 2004. Beyond the ivory tower - The scientific consensus on climate change. Science 306:1686-1686. Ovenden, L. 1986. Vegetation Colonizing the Bed of a Recently Drained Thermokarst Lake (Illisarvik), Northwest-Territories. Canadian Journal of Botany-Revue Canadienne De Botanique 64:2688-2692. Palmer, M. J. 2007. Vegetation, snow cover and air and near-surface ground temperatures across treeline in the uplands east of the Mackenzie Delta, Northwest Territories. Carleton University, Ottawa, ON. Parry, M. L., and Intergovernmental Panel on Climate Change. Working Group II. 2007. Climate Change 2007 : impacts, adaptation and vulnerability : contribution of Working Group II to the fourth assessment report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, U.K. ; New York. Pearce, C. M., D. Mclennan, and L. D. Cordes. 1988. The Evolution and Maintenance of White Spruce Woodlands on the Mackenzie Delta, Nwt, Canada. Holarctic Ecology 11:248-258. Pearson, R. G., and T. P. Dawson. 2003. Predicting the impacts of climate change on the distribution of species: are bioclimate envelope models useful? Global Ecology and Biogeography 12:361-371. Peterson, K. M., and W. D. Billings. 1980. Tundra Vegetational Patterns and Succession in Relation to Micro-Topography near Atkasook, Alaska. Arctic and Alpine Research 12:473-482. Pomeroy, J. W., D. S. Bewley, R. L. H. Essery, N. R. Hedstrom, T. Link, R. J. Granger, J. E. Sicart, C. R. Ellis, and J. R. Janowicz. 2006. Shrub tundra snowmelt. Hydrological Processes 20:923-941. Porsild, A. E., and W. J. Cody. 1980. Vascular plants of Continental Northwest Territories, Canada. National Museum of Natural Sciences, National Museums of Canada, Ottawa. Racine, C., R. Jandt, C. Meyers, and J. Dennis. 2004. Tundra fire and vegetation change along a hillslope on the Seward Peninsula, Alaska, USA. Arctic Antarctic and Alpine Research 36:1-10. Rampton, V. 1988. Quaternary geology of the Tuktoyaktuk Coastlands, Northwest Territories. Geological Survey of Canada, Memoir 423. 88  Rhoades, C., D. Binkley, H. Oskarsson, and R. Stottlemyer. 2007. Soil nitrogen accretion along a floodplain terrace chronosequence in Northwest Alaska: Influence of the nitrogen-fixing shrub Shepherdia canadensis. Ecoscience 15:In Press. Rhoades, C., H. Oskarsson, D. Binkley, and R. Stottlemyer. 2001. Alder (Alnus crispa) effects on soils vary with ecosystem type along the Agashashok River, northwest Alaska. Ecoscience 8:89-95. Ritchie, J. C. 1984. Past and present vegetation of the far northwest of Canada. University of Toronto Press, Toronto, ON. Robinson, S. D. 2000. Thaw slump derived thermokarst near Hot Weather Creek, Ellesmere Island, Nunavut. Pages 335-345 in M. Garneau and B. T. Alt, editors. Environmental Response to Climate Change in the Canadian High Arctic. Geological Survey of Canada, Bulletin 529, Ottawa, ON. Rupp, T. S., F. S. Chapin, and A. M. Starfield. 2000. Response of subarctic vegetation to transient climatic change on the Seward Peninsula in north-west Alaska. Global Change Biology 6:541-555. SAS. 2004. SAS Version 9.1.3. Help and Documentation.. SAS Institute, Cary, NC. Schimel, J. P., C. Bilbrough, and J. A. Welker. 2004. Increased snow depth affects microbial activity and nitrogen mineralization in two Arctic tundra communities. Soil Biology & Biochemistry 36:217-227. Shaver, G. R., J. Canadell, F. S. Chapin, J. Gurevitch, J. Harte, G. Henry, P. Ineson, S. Jonasson, J. Melillo, L. Pitelka, and L. Rustad. 2000. Global warming and terrestrial ecosystems: A conceptual framework for analysis. Bioscience 50:871882. Shaver, G. R., and F. S. Chapin. 1980. Response to Fertilization by Various Plant-Growth Forms in an Alaskan Tundra - Nutrient Accumulation and Growth. Ecology 61:662-675. Shaver, G. R., F. S. Chapin, and W. D. Billings. 1979. Ecotypic Differentiation in Carex aquatilis on Ice-Wedge Polygons in the Alaskan Coastal Tundra. Journal of Ecology 67:1025-1046. Shirazi, M. A., P. K. Haggerty, C. W. Hendricks, and M. Reporter. 1998. The role of thermal regime in tundra plant community restoration. Restoration Ecology 6:111-117. Staniforth, R. J., and P. A. Scott. 1991. Dynamics of Weed Populations in a Northern Sub-Arctic Community. Canadian Journal of Botany-Revue Canadienne De Botanique 69:814-821. Stow, D. A., A. Hope, D. McGuire, D. Verbyla, J. Gamon, F. Huemmrich, S. Houston, C. Racine, M. Sturm, K. Tape, L. Hinzman, K. Yoshikawa, C. Tweedie, B. Noyle, C. Silapaswan, D. Douglas, B. Griffith, G. Jia, H. Epstein, D. Walker, S. Daeschner, A. Petersen, L. M. Zhou, and R. Myneni. 2004. Remote sensing of vegetation and land-cover change in Arctic Tundra Ecosystems. Remote Sensing of Environment 89:281-308. Sturm, M., J. P. McFadden, G. E. Liston, F. S. Chapin, C. H. Racine, and J. Holmgren. 2001. Snow-shrub interactions in Arctic tundra: A hypothesis with climatic implications. Journal of Climate 14:336-344.  89  Sturm, M., J. Schimel, G. Michaelson, J. M. Welker, S. F. Oberbauer, G. E. Liston, J. Fahnestock, and V. E. Romanovsky. 2005. Winter biological processes could help convert arctic tundra to shrubland. Bioscience 55:17-26. Svoboda, J., and G. H. R. Henry. 1987. Succession in Marginal Arctic Environments. Arctic and Alpine Research 19:373-384. Tape, K., M. Sturm, and C. Racine. 2006. The evidence for shrub expansion in Northern Alaska and the Pan-Arctic. Global Change Biology 12:686-702. Tape, K. D. 2004. Shrub expansion in Arctic Alaska: 50 years of change documented using aerial photography. University of Alaska Fairbanks, Fairbanks, AK. Thompson, C., J. Beringer, F. S. Chapin, and A. D. McGuire. 2004. Structural complexity and land-surface energy exchange along a gradient from arctic tundra to boreal forest. Journal of Vegetation Science 15:397-406. Timoney, K. P., G. H. Laroi, S. C. Zoltai, and A. L. Robinson. 1992. The High SubArctic Forest-Tundra of Northwestern Canada - Position, Width, and Vegetation Gradients in Relation to Climate. Arctic 45:1-9. Timoney, K. P., G. H. Laroi, S. C. Zoltai, and A. L. Robinson. 1993. Vegetation Communities and Plant-Distributions and Their Relationships with Parent Materials in the Forest-Tundra of Northwestern Canada. Ecography 16:174-188. Trimble, A. 2008. Effects of thaw slumping on seed weight and germination in arctic grasses. B.Sc. Thesis. Thompson Rivers University, Kamloops, British Columbia. Truett, J. C., and K. Kertell. 1992. Tundra disturbance and ecosystem production: implications for impact assessment. Environmental Management 16:485-494. Turner, M. G., W. H. Romme, and D. B. Tinker. 2003. Surprises and lessons from the 1988 Yellowstone fires. Frontiers in Ecology and the Environment 1:351-358. Vavrek, M. C., N. Fetcher, J. B. McGraw, G. R. Shaver, F. S. Chapin, and B. Bovard. 1999. Recovery of productivity and species diversity in Tussock tundra following disturbance. Arctic Antarctic and Alpine Research 31:254-258. Walker, D. A., D. Cate, J. Brown, and C. Racine. 1987. Disturbance and recovery of arctic Alaskan tundra terrain: a review of recent investigations. CRREL Report 87-11. US Army Corps of Engineers. Cold Regions Research and Engineering Laboratory, Hanover, NH. Walker, D. A., and M. D. Walker. 1991. History and Pattern of Disturbance in Alaskan Arctic Terrestrial Ecosystems - a Hierarchical Approach to Analyzing Landscape Change. Journal of Applied Ecology 28:244-276. Walker, M. D., C. H. Wahren, R. D. Hollister, G. H. R. Henry, L. E. Ahlquist, J. M. Alatalo, M. S. Bret-Harte, M. P. Calef, T. V. Callaghan, A. B. Carroll, H. E. Epstein, I. S. Jonsdottir, J. A. Klein, B. Magnusson, U. Molau, S. F. Oberbauer, S. P. Rewa, C. H. Robinson, G. R. Shaver, K. N. Suding, C. C. Thompson, A. Tolvanen, O. Totland, P. L. Turner, C. E. Tweedie, P. J. Webber, and P. A. Wookey. 2006. Plant community responses to experimental warming across the tundra biome. Proceedings of the National Academy of Sciences of the United States of America 103:1342-1346. Walker, M. D., D. A. Walker, J. M. Welker, A. M. Arft, T. Bardsley, P. D. Brooks, J. T. Fahnestock, M. H. Jones, M. Losleben, A. N. Parsons, T. R. Seastedt, and P. L. Turner. 1999. Long-term experimental manipulation of winter snow regime and  90  summer temperature in arctic and alpine tundra. Hydrological Processes 13:23152330. Walsh, N. E., T. R. McCabe, J. M. Welker, and A. N. Parsons. 1997. Experimental manipulations of snow-depth: effects on nutrient content of caribou forage. Global Change Biology 3(Suppl. 1) 158-164. Walther, G. R. 2003. Plants in a warmer world. Perspectives in Plant Ecology Evolution and Systematics 6:169-185. Walther, G. R., E. Post, P. Convey, A. Menzel, C. Parmesan, T. J. C. Beebee, J. M. Fromentin, O. Hoegh-Guldberg, and F. Bairlein. 2002. Ecological responses to recent climate change. Nature 416:389-395. Wein, R. W., G. Wein, S. Bahret, and W. J. Cody. 1992. Invading vascular plant species in the vicinity of Wood Buffalo National Park, Canada. Canadian Field Naturalist 106: 216-224. White, P. S., and A. Jentsch. 2001. The search for generality in studies of disturbance and ecosytem dynamics. Progress in Botany 62:400-450. Wolfe, S. A., E. Kotler, and S. R. Dallimore. 2001. Surficial characteristics and the distribution of thaw landforms (1970 to 1999), Shingle Point to Kay Point, Yukon Territory. Geological Survey of Canada, Open File 4088. Geological Survey of Canada, Ottawa, ON. Yamaguchi, D. K. 1991. A Simple Method for Cross-Dating Increment Cores from Living Trees. Canadian Journal of Forest Research-Revue Canadienne De Recherche Forestiere 21:414-416. Zasada, J. C., R. A. Norum, R. M. Vanveldhuizen, and C. E. Teutsch. 1983. Artificial Regeneration of Trees and Tall Shrubs in Experimentally Burned Upland Black Spruce Feather Moss Stands in Alaska. Canadian Journal of Forest ResearchRevue Canadienne De Recherche Forestiere 13:903-913. Zasada, J. C., T. L. Sharik, and M. Nygren. 1992. The reprodcutive process in boreal forest trees. in H. H. Shugart, R. Leemans, and G. B. Bonan, editors. A systems of the global boreal forest. Cambridge University Press Cambridge.  91  Chapter 4. Increasing rates of retrogressive thaw slump activity in the Mackenzie Delta region, N.W.T. 1 INTRODUCTION Recent temperature increases in the Arctic and subarctic have been significantly greater than global averages and the North American western Arctic is one of the most rapidly warming regions on the planet (Serreze et al. 2000, Johannessen et al. 2004). Permafrost temperatures are rising in response to 20th century climate warming in Alaska and northwestern Canada and the frequency and magnitude of terrain disturbances associated with thawing permafrost, including the degradation of ice wedges and lake shrinkage, is increasing (Osterkamp and Romanovsky 1999, Yoshikawa and Hinzman 2003, Smith et al. 2005, Jorgenson et al. 2006). Thawing of ice-rich permafrost on sloping terrain can lead to the development of retrogressive thaw slumps (Burn and Lewkowicz 1990). These conspicuous disturbances are common along coastlines and lakeshores in the western Arctic initiate when ice-rich soils are exposed and thaw (Mackay 1963, Kokelj et al. 2005, Lantuit and Pollard 2005) (Figure 4.1). As ground ice melts, materials turn into a mud slurry and flow to the base of the exposure. If terrain is sufficiently ice-rich and air temperatures are warm, the slump headwall may retreat upslope by several meters in a single summer. Thaw slumps are polycyclic in nature and individual disturbances are often comprised of old, recently stabilized and active scar areas and can affect several hectares of terrain (Figure 4.1) (Lewkowicz 1987). In addition to the geomorphic significance of thaw slumps (Lewkowicz 1987), these disturbances also affect the chemistry of soils, surface runoff and lake water (Kokelj et al. 2002, Kokelj et al. 2005), release organic carbon sequestered in frozen ground (Lantuit and Pollard 2005), and disturb tundra vegetation and create a mosaic of successional stages (Bartleman et al. 2001). The effect of thaw slumps on 1  A version of this chapter has been published as: Lantz, T. C., and S. V. Kokelj. 2008. Increasing rates of retrogressive thaw slump activity in the Mackenzie Delta region, N.W.T., Canada. Geophysical Research Letters 35:L06502, doi:06510.01029/02007GL032433.  92  terrain stability also makes their activity and distribution relevant in planning linear infrastructure such as oil and gas pipelines.  Figure 4.1. Retrogressive thaw slump adjacent to a tundra lake, Richards Island, Mackenzie Delta region. The arrow at left indicates two field researchers. The areal extent of the entire disturbance is approximately 1ha. Photo: Peter Morse Accelerated climate warming and the abundance of ice-rich permafrost in the western Arctic raises the possibility that thaw slump activity is increasing. In this paper our objectives are: 1) to examine recent temperature trends (1926-2006) in the Mackenzie Delta region, and 2) to assess changes in the rates of slump growth and slump headwall retreat. METHODS To examine changes in the annual air temperature in the Mackenzie Delta region we performed regression analyses of historical (1926-2006) temperatures recorded at Inuvik and Aklavik (mean annual, mean summer (June-Sept.), and the number of days where the maximum temperature > 20ºC) (SAS 2004, Environment Canada 2005). Daily 93  temperature data was not recorded in 1936, 1939, 1943, 1944, so these years were omitted from the analysis of the number of days where the maximum temperature exceeded 20ºC. To test for first and higher order autocorrelation in the error terms of all time series we calculated the Durbin-Watson statistic using the PROC AUTOREG statement in SAS (SAS 2004). To calculate the density of thaw slumps in the tundra uplands of the Mackenzie Delta region we identified and recorded the location of all thaw slumps within a 3739 km2 study corridor using aerial photographs (Figure 4.2). We also recorded the surficial material associated with each disturbance (Aylsworth et al. 2000). Rates of thaw slump activity were estimated by randomly choosing 25 colour aerial photographs (1:30 000) taken in 2004 (Figure 4.2). We also obtained air photos from 1950 (1:40 000) and 1973 (1:54 000), which correspond to the 25 plot areas in the 2004 images. Each plot used to determine rates of slump activity was defined as the extent of the 2004 aerial photograph and covered approximately 49 km2. 2004 images were scanned at 1814 dpi (14 µm) from negatives using a high-resolution photogrammetric quality scanner. Greyscale images from 1950 and 1973 were scanned from at 1200 dpi (21 µm) prints using an Epson 1640XL flatbed scanner. Effective pixel sizes of images in each time period were 0.4 m (2004), 1.3 m (1973), and 0.9 m (1950). All photos were georeferenced and displayed onscreen in stereo using the DVP digital photography software (DVP-GS, Québec, Canada). Mean standard error associated with the absolute orientation of all photos was 3.0 m. As two of these plots fell within the area burned by an intense 1968 wildfire, which degraded near-surface permafrost (Heginbottom 1972, Mackay 1995), we removed them from the analysis. The 23 photos retained cover approximately 1100 km2.  94  Figure 4.2. Map of the retrogressive thaw slumps in the upland tundra study region east of the Mackenzie River Delta. Areas bounded by a single line represent study plots where disturbances were mapped on aerial photographs taken in 1950, 1973 and 2004.  95  Slumps visible within the 23 study plots were digitized for all three time periods while being viewed in stereo using DVP (DVP-GS, Québec, Canada). To ensure that our measurements were not biased by differences in effective pixel sizes of photos from different years, we used a minimum mapping area of 10 m2. The annual rates of slump growth from 1950-1973 and 1973-2004 were obtained by dividing the change in area by the number of years in the period between photos. Thaw slumps that were inactive over the entire period of study were removed from the analyses. Slumps may grow laterally and upslope, but their development may also be concurrent with lake expansion. Our mapping did not account for disturbed terrain lost to lake expansion, which occurred in approximately ten percent of studied cases. The lake shorelines at the remainder of the disturbances were comparatively stable. Due to terrain loss in association with shoreline retreat, our estimates of slump growth are likely conservative. Thus, in order to provide a second metric of slump activity we measured the changes in slump headwall position along transects established through highly active slumps (1950-2004 growth rates > 70 m2/year, n=50) using a minimum mapping distance of 10 m. The annual rates of maximum headwall retreat (1950-1973, 1973-2004) were estimated by dividing the change in headwall position by the number of years in the period between photos. To determine if these rates of slump activity differed between the two time periods, we performed a within-subject t-test (SAS 2004). To meet the assumption of normality, the rates of slump growth and headwall retreat were logtransformed.  RESULTS Air temperatures fluctuated considerably from year to year, but mean annual, mean summer temperature, and the number of days with temperature over 20°C per year show a significant increase from 1926 to 2006 (Figure 4.3). The linear regression of mean annual temperatures from 1926 to 2006 describes a 1.9ºC increase in average annual temperature (F1, 79 = 22.74, P<0.001 r2=0.224) (Figure 4.3a). Similarly, the analysis of summer conditions shows that mean June to September air temperatures have risen by 2.2°C since 1926 (F1, 79 = 22.0, P<0.001 r2=0.218) (Figure 4.3b). Reflecting the 96  increasing mean summer temperatures, the annual total number of days with the maximum temperatures above 20ºC have also increased by 14.3 days since 1928 (F1, 73 = 17.32, P<0.001 r2=0.192) (Figure 4.3c). These time series did not exhibit significant first or higher order autocorrelation of error terms. Summer temperatures are cooler at the coast than at inland locations (Burn 1997), but strong correlations in air temperatures between Inuvik, Shingle Point (68 º57’N, 133º13’W) and Tuktoyaktuk (69º26’N, 133º00’W) (r > 0.95) indicate that the warming trend observed in the long-term composite record for the central Delta has been a regional phenomena. A total of 541 areas affected by thaw slumps were mapped along lakeshores in our 3739 km2 study region (Figure 4.2). Morainal deposits comprising approximately 55% of the landscape are associated with 70% of the thaw slumps, while the remaining disturbances were distributed approximately proportional to the relative areas occupied by glaciofluvial (22%), lacustrine (17%) and colluvial (1%) deposits. All slumps mapped were associated with tundra lakes and ponds, impacting approximately 8% of the 2880 lakes greater than 1 ha in area. From 1950 to 1973, the areal extent of all thaw slumps mapped increased by about 15% and by 2004 the total area had grown by approximately 36% from 1973. The mean rate of slump growth from 1973 to 2004 was about 1.4 times the rate estimated for the period from 1950 to 1973 (Figure 4.4a) and the difference in the rates from the two time periods was significantly greater than 0 (t110=5.19, P<0.001). The mean rate of slump headwall retreat during the period from 1973 to 2004 was approximately double that estimated for the period from 1950 to 1973 and the difference was significantly greater than 0 (Figure 4.4b, t49=3.86, P<0.001). In 1950, 1973, and 2004 mean slump sizes were 1.02, 1.15, and 1.34 ha, respectively.  97  Figure 4.3. Air temperature time series for the central Mackenzie Delta region from 1926 to 2006. Series plotted include: (A) mean annual, (B) mean summer (JuneSeptember), and (C) number of days where the maximum temperature exceeded 20ºC (1928-2006). 98  DISCUSSION Long-term activity of thaw slumps is determined by the frequency of slump reactivation and the rates and extent of headwall retreat. The growth of an active slump is related to the ice content of the thawing terrain, slump aspect, and morphology (Lewkowicz 1986, 1987). Within this context of site specific conditions, slump growth will occur in response to the ablation of ground ice during the thaw season (Lewkowicz 1986). Our observation that a random sample of active slumps exhibited significantly higher growth rates from 1973-2004 than from 1950-1973 suggests that a regional driver of slump growth has subsumed site specific controls. Process studies have shown that a model including net radiation, temperature, vapour pressure and wind speed best describe ground ice ablation in slumps, although air temperature alone can also explain a significant proportion of the short-term variability of ground ice ablation (r2= 0.39) (Lewkowicz 1986). At longer time scales, models that only employ air temperature successfully explain rates of ablation (r2=0.98) and headwall retreat (r2=0.68-0.97) (Kerfoot 1969, Robinson 2000). Over the period that the rate of slump growth has increased, the mean summer air temperatures at Inuvik have risen by 1.3°C and the annual total of days with air temperature exceeding 20ºC have increased by 9.9 days. Summer warming and increased ground ice ablation is the most plausible explanation for the regional increase in rates of slump headwall retreat and overall slump growth rates. Increases in the number of thaw slumps west of the Mackenzie Delta have also been observed over a similar time period (Wolfe et al. 2001, Lantuit and Pollard In Press).  99  Figure 4.4. Mean rates of slump growth in the Mackenzie Delta region. (A) Average annual rates of slump growth estimated from the change in areal extent of disturbance from 1950 to 1973 (n=110) and from 1973 to 2004 (n=110) for all active slumps mapped on the 23 s study plots, and (B) average annual rates of headwall retreat from 1950 to 1973 (n=50) and 1973 to 2004 (n=50). Error bars represent ± SE of the mean.  100  The rates of slump activity that we describe here are considerably lower than those, which have been reported in other areas (Kerfoot 1969, Wolfe et al. 2001, Lantuit and Pollard 2005, Lantuit and Pollard In Press) largely due to the decadal time periods for which our growth rates have been estimated. It is likely that growth of even the most active slumps has been punctuated by periods of stability (Robinson 2000, Wolfe et al. 2001, Lantuit and Pollard In Press). Given the polycyclic nature of slumps, a regional increase in slump area suggests that frequency of reactivation may also be increasing. Thermal erosion is well-established as a trigger mechanism for thaw slumps at the coastline or along rivers (Lewkowicz 1987, Lantuit and Pollard In Press), but initiating factors in upland settings around small lakes, including active layer deepening, warming permafrost, wave action, and gullying by surface runoff, remain poorly understood. Thaw slumps are widespread across the tundra landscape and the aerial extent of disturbance is growing. Since slumping results in the degradation of near-surface permafrost (Burn 2000), this process is also likely to release carbon stored in soils (Lantuit and Pollard 2005, Zimov et al. 2006). Areas affected by thaw slumping are characterized by ion-rich mineral substrates (Kokelj et al. 2002) and an ameliorated microclimate with respect to the undisturbed terrain (Burn 2000). Since barriers to establishment constrain the effects of climate warming on vegetation, (Hurtt et al. 1998, Walther et al. 2002) disturbances such as slumps represent localized areas where vegetation may show rapid responses to recent temperature increases (Forbes et al. 2001). The development of unique plant communities on recent and stabilized slumps, in conjunction with our frequent observation of large and small mammals and songbirds, suggest that slumps are ecologically important disturbances within the tundra environment. Thaw slumps have discernable effects on lake-water chemistry even where disturbances occupy only small percentage of the catchment area. Degrading permafrost releases soluble materials, which are transported by surface runoff from ion-rich slump soils into adjacent aquatic systems, elevating ionic concentrations in lake water (Kokelj et al. 2005). A positive association between solute concentrations and the proportion of catchment area influenced by thaw slumping (Kokelj and Burn 2005) suggests that  101  permafrost disturbance will be an important factor influencing the chemistry of thousands of lakes in a warming western Arctic. Overall we draw the following conclusions: 1. In the central Mackenzie Delta region since 1926, mean annual air temperatures, mean summer air temperatures, and the number of days with maximum air temperatures greater than 20ºC have increased by 1.9 ºC, 2.2ºC and 14.3 days, respectively. 2. Rates of slump growth from 1973 to 2004 were 1.4 times greater than rates from 1950-1973 and rates of headwall retreat increased approximately twofold. Slumps enlarge in summer due to ablation of ground ice, which is driven by net radiation and air temperature. The regional acceleration in slump activity between 1950-1973 and 1973-2004 is likely occurring in response to warming air temperatures. 3. In ice-rich terrain, the rates and areal extent of thaw slumping can be expected to increase with future warming.  102  ACKNOWLEDGEMENTS This work was supported by the Arctic Institute of North America, the Aurora Research Institute, the Canon National Park Science Scholars Program, Global Forest Research, Natural Resources Canada, the Natural Sciences and Engineering Research Council of Canada, the Northern Science Training Program, the Polar Continental Shelf Project and the Water Resources Division and the Mackenzie Valley Air Photo Project of Indian and Northern Affairs Canada. The authors would also like to thank Chris Burn, Sarah Gergel, and an anonymous reviewer for providing comments which improved the manuscript.  103  LITERATURE CITED Aylsworth, J. M., M. M. Burgess, D. T. Desrochers, A. Duk-Rodkin, T. Robertson, and J. A. Traynor. 2000. Surficial geology, subsurface materials, and thaw sensitivity of sediments. Pages 41-48 in L. D. Dyke and G. R. Brooks, editors. The phsyical environment of the Mackenzie Valley, Northwest Territories: A base line for the assessment of environmental change. Geological Survey of Canada Bulletin 547. Geological Survey of Canada, Ottawa, ON. Bartleman, A. P., K. Miyanishi, C. R. Burn, and M. M. Cote. 2001. Development of vegetation communities in a retrogressive thaw slump near Mayo, Yukon Territory: A 10-year assessment. Arctic 54:149-156. Burn, C. R. 1997. Cryostratigraphy, paleogeography, and climate change during the early Holocene warm interval, western Arctic coast, Canada. Canadian Journal of Earth Sciences 34:912-925. Burn, C. R. 2000. The thermal regime of a retrogressive thaw slump near Mayo, Yukon Territory. Canadian Journal of Earth Sciences 37:967-981. Burn, C. R., and A. G. Lewkowicz. 1990. Retrogressive thaw slumps. Canadian Geographer 34:273-276. Environment Canada. 2005. Adjusted Historical Canadian Climate Data. in. Environment Canada. Available Online: http://www.cccma.bc.ec.gc.ca/hccd/. Forbes, B. C., J. J. Ebersole, and B. Strandberg. 2001. Anthropogenic disturbance and patch dynamics in circumpolar arctic ecosystems. Conservation Biology 15:954969. Heginbottom, J. A. 1972. Some effects of a forest fire on the permafrost active layer at Inuvik, N.W.T. Pages 149-156 in D. Kerfoot, editor. 22nd International Geographical Congress. Brock University Press. Hurtt, G. C., P. R. Moorcroft, S. W. Pacala, and S. A. Levin. 1998. Terrestrial models and global change: challenges for the future. Global Change Biology 4:581-590. Johannessen, O. M., L. Bengtsson, M. W. Miles, S. I. Kuzmina, V. A. Semenov, G. V. Alekseev, A. P. Nagurnyi, V. F. Zakharov, L. P. Bobylev, L. H. Pettersson, K. Hasselmann, and A. P. Cattle. 2004. Arctic climate change: observed and modelled temperature and sea-ice variability. Tellus Series a-Dynamic Meteorology and Oceanography 56:328-341. Jorgenson, M. T., Y. L. Shur, and E. R. Pullman. 2006. Abrupt increase in permafrost degradation in Arctic Alaska. Geophysical Research Letters 33:-. Kerfoot, D. E. 1969. The geomorphology and permafrost conditions of Garry Island, N.W.T. Ph.D. Dissertation. University of British Columbia, Vancouver, B.C. . Kokelj, S. V., and C. R. Burn. 2005. Geochemistry of the active layer and near-surface permafrost, Mackenzie delta region, Northwest Territories, Canada. Canadian Journal of Earth Sciences 42:37-48. Kokelj, S. V., R. E. Jenkins, D. Milburn, C. R. Burn, and 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.  104  Kokelj, S. V., C. A. S. Smith, and C. R. Burn. 2002. Physical and chemical characteristics of the active layer and permafrost, Herschel Island, western Arctic Coast, Canada. Permafrost and Periglacial Processes 13:171-185. Lantuit, H., and W. H. Pollard. 2005. Temporal stereophotogrammetric analysis of retrogressive thaw slumps on Herschel Island, Yukon territory. Natural Hazards and Earth System Sciences 5:413-423. Lantuit, H., and W. H. Pollard. In Press. Fifty years of coastal erosion and retrogressive thaw slumpactivity on Herschel Island, southern Beaufort Sea, Yukon Territory, Canada. Geomorphology. Lewkowicz, A. G. 1986. Rate of Short-Term Ablation of Exposed Ground Ice, Banks Island, Northwest Territories, Canada. Journal of Glaciology 32:511-519. Lewkowicz, A. G. 1987. Headwall Retreat of Ground-Ice Slumps, Banks-Island, Northwest-Territories. Canadian Journal of Earth Sciences 24:1077-1085. Mackay, J. R. 1963. The Mackenzie Delta Area, N.W.T. Department of Mines and Technical Surveys, Ottawa, ON. Mackay, J. R. 1995. Active Layer Changes (1968 to 1993) Following the Forest-Tundra Fire near Inuvik, Nwt, Canada. Arctic and Alpine Research 27:323-336. Osterkamp, T. E., and V. E. Romanovsky. 1999. Evidence for warming and thawing of discontinuous permafrost in Alaska. Permafrost and Periglacial Processes 10:1737. Robinson, S. D. 2000. Thaw slump derived thermokarst near Hot Weather Creek, Ellesmere Island, Nunavut. Pages 335-345 in M. Garneau and B. T. Alt, editors. Environmental Response to Climate Change in the Canadian High Arctic. Geological Survey of Canada, Bulletin 529, Ottawa, ON. SAS. 2004. SAS Version 9.1.3. Help and Documentation.. SAS Institute, Cary, NC. Serreze, M. C., J. E. Walsh, F. S. Chapin, T. Osterkamp, M. Dyurgerov, V. Romanovsky, W. C. Oechel, J. Morison, T. Zhang, and R. G. Barry. 2000. Observational evidence of recent change in the northern high-latitude environment. Climatic Change 46:159-207. Smith, S. L., M. M. Burgess, D. Riseborough, and F. M. Nixon. 2005. Recent trends from Canadian permafrost thermal monitoring network sites. Permafrost and Periglacial Processes 16:19-30. Walther, G. R., E. Post, P. Convey, A. Menzel, C. Parmesan, T. J. C. Beebee, J. M. Fromentin, O. Hoegh-Guldberg, and F. Bairlein. 2002. Ecological responses to recent climate change. Nature 416:389-395. Wolfe, S. A., E. Kotler, and S. R. Dallimore. 2001. Surficial characteristics and the distribution of thaw landforms (1970 to 1999), Shingle Point to Kay Point, Yukon Territory. Geological Survey of Canada, Open File 4088. Geological Survey of Canada, Ottawa, ON. Yoshikawa, K., and L. D. Hinzman. 2003. Shrinking thermokarst ponds and groundwater dynamics in discontinuous permafrost near Council, Alaska. Permafrost and Periglacial Processes 14:151-160. Zimov, S. A., S. P. Davydov, G. M. Zimova, A. I. Davydova, E. A. G. Schuur, K. Dutta, and F. S. Chapin. 2006. Permafrost carbon: Stock and decomposability of a globally significant carbon pool. Geophysical Research Letters 33:-.  105  Chapter 5. Variability in the vegetation of the shrub tundra ecotone in the Mackenzie Delta Region, NWT: Implications for Global Change. 1 INTRODUCTION The Arctic is often described as one of the major components in the earth’s cooling system. It plays a critical role in the global climate system by reflecting incoming solar radiation and by radiating energy gains transferred from the tropics (Chapin et al. 2005a, McGuire et al. 2006). While large portions of the Arctic have high albedo because they are covered in snow and ice for much of the year, the structure of the vegetation is far from uniform. Moving northward, temperature decreases are generally accompanied by a shift through the following sequence of vegetation types: 1) forest tundra, 2) upright shrub, 3) erect dwarf shrub, 4) prostate dwarf shrubs, and 5) cushionforbs and exposed substrates (Bliss and Matveyeva 1992, Walker 2000, Epstein et al. 2004a). Transitions between these ecosystems are associated with large differences in ecosystem properties (e.g. albedo, net primary productivity, biomass, heterotrophic respiration, carbon storage, permafrost conditions) (Chapin et al. 2000, McGuire et al. 2002, Thompson et al. 2004, Euskirchen et al. 2007). Thus, as temperatures increase at high latitudes (Serreze et al. 2000, Stafford et al. 2000, ACIA 2004, Johannessen et al. 2004) resulting changes to the structure of Arctic vegetation (Kaplan et al. 2003, Hassol 2004, Notaro et al. 2007) are likely to feedback to the global climate system (Chapin et al. 2005b, McGuire et al. 2006). Arctic vegetation also exhibits heterogeneity at fine scales, which can influence both regional and microenvironmental feedbacks. For example, shifts from upright shrub tundra to dwarf shrub tundra often occur at a scale finer than the grain of typical climate models (Bliss and Matveyeva 1992, Walker et al. 1994, Quattrochi and Goodchild 1997, Epstein et al. 2004b). Yet these distinct tundra communities also exhibit important differences in albedo, sensible heat, and snow pack depth and duration (Epstein et al. 2004a, Chapin et al. 2005b, Sturm et al. 2005). The transition between shrub tundra and 1  A version of this chapter will be submitted for publication. Lantz, T.C., S.V. Kokelj, S.E. Gergel, and N.C. Coops. Variability in the vegetation of the shrub tundra ecotone in the Mackenzie Delta Region, NWT: Implications for Global Change.  106  dwarf shrub tundra (i.e., the shrub tundra ecotone) is also the Arctic ecotone expected to respond most rapidly to climate warming (Epstein et al. 2004a). Furthermore, changes in vegetation composition and patch size across this ecotone may drive positive feedbacks to climate by altering the physical properties of these ecosystems (Epstein et al. 2004a, Sturm et al. 2005, Pomeroy et al. 2006). Thus, accurate identification of the shrub tundra ecotone, and the fine-scale variability within it, is critical. In this paper, we explore two objectives related to understanding the shrub tundra transition zone and the fine scale variability in vegetation. Objective 1: Describe latitudinal changes in the proportion and patch sizes of shrub tundra and dwarf shrub tundra. In the Mackenzie delta uplands north of Inuvik, NWT there is a general shift from tundra communities dominated by shrubs >50 cm tall, to those characterized by the abundance of dwarf shrubs and sedges <50 cm (Mackay 1963, Corns 1974, Forest Management Institute 1975). Recent evidence suggests that shrub tundra is encroaching into areas of dwarf shrub and tussock tundra across the entire circumpolar region (Silapaswan et al. 2001, Sturm et al. 2001b, Stow et al. 2004, Tape 2004, Tape et al. 2006), but in the Mackenzie Delta Region, information on this transition is lacking. Since changes in ecosystem properties across this ecotone may only be associated with vegetation patches of certain sizes (Pomeroy et al. 1995, Sturm et al. 2001a, Grogan and Jonasson 2006), understanding variability in patch size of shrub tundra is also critical. Here we use object-based classification of airphotos to map and then quantify changes in shrub abundance and patch size in the Mackenzie Delta uplands. Objective 2: Determine the relative importance of biophysical drivers of the shrub tundra ecotone. The transition from dwarf shrub tundra to shrub tundra is correlated with broad scale differences in temperature (Walker 2000, Epstein et al. 2004b) implying that climate has an important influence on the position of this ecotone. Experimental evidence also shows that Arctic shrubs respond to warming and increased nutrient availability on decadal time scales (Parsons et al. 1994, Chapin et al. 1995, Bret-Harte et 107  al. 2001, Bret-Harte et al. 2002, Walker et al. 2006). However, nutrient availability, topography, and soil pH also strongly influence the distribution Arctic plant communities at fine and broad scales (Walker et al. 1994, Ostendorf and Reynolds 1998, Gough et al. 2000, Walker et al. 2002). Thus, it is unclear to what degree these factors will constrain the response of shrub tundra and dwarf shrub tundra to a warming climate. The landscape of the Mackenzie delta uplands is structured by a strong summer temperature gradient driven by proximity to the Beaufort Sea (Ritchie 1984, Burn 1997), and has relatively homogenous parent materials, but subtle variability in local topography and soils (Mackay 1963, Aylsworth et al. 2000). Thus it is an excellent place to explore how regional temperature, soils and microtopography interact to govern the shrub tundra ecotone. Here we use correlation analysis and hierarchical variance portioning to evaluate the relative importance of biophysical variables including temperature, topography and soils in structuring the shrub tundra ecotone.  METHODS Study Area Our study area in Northern Canada covers approximately 11 000 km2 located between the latitudes of 68°30’ and 69˚50N and the longitudes of 134°54’W and 132°40’W (Figure 5.1). This area east of the Mackenzie River delta consists of Pleistocene upland terrain with low rolling topography (30-150 m) and thousands of small lakes (Mackay 1963, Aylsworth et al. 2000). Soils in this hummocky permafrost terrain are ice rich clayey silts overlain by organic layers of variable depth (Mackay 1963, Kokelj and Burn 2004). From October through April mean air temperatures in the region are less 0°C (Ritchie 1984, Burn 1997), and during the short summer growing season proximity to the Beaufort Sea drives a strong north-south temperature gradient (Ritchie 1984, Burn 1997).  108  Figure 5.1. Map of the study region showing the study area, settlements, water (blue), temperature loggers, and airphoto study plots. Inset map at the bottom right shows the approximate position on the study area in North America.  109  Airphoto Selection and Image Manipulation To quantify the abundance of shrub tundra and dwarf shrub tundra in the uplands of the Mackenzie Delta region, we randomly selected 18 plots along a latitudinal gradient extending from 68˚26’ to 69˚34’ (Figure 5.1). Each of these plots consisted of a colour aerial photograph (1:30 000) captured in 2004, covering an area of approximately 49 km2. To ensure that these areas had not been recently impacted by either fires or seismic exploration, we rejected photos of areas with seismic line densities exceeding 5 km/km2, or within 1 km of known fires. Negatives were scanned at 1814 dpi (1 pixel = 0.41 m) on a high-resolution photogrammetric quality scanner. Scanned airphotos were orthorectified using a 30 m DEM (Geological Survey of Canada 2002) and Landsat 7 panchromatic orthoimagery (15 m pixel). Orthorectification was performed using a nearest neighbor algorithm (PCI Geomatics 2001) resulting in relative RMS error generally less than 3 pixels. Orthorectified images were clipped to a 36 km2 area surrounding the principle point to remove the image borders. We tested clipped images for systematic brightness bias by sampling pixel values along transects extending from the principle point to the outside edges of the image. To control for differences in brightness due to vegetation patch structure we used semivariograms of the green band to determine that beyond a lag distance of 180 m fine scale spatial autocorrelation was less than 0.1. Subsequently, we sub-sampled images from the principle point to the photo edge at intervals of 180 m and tested for brightness changes with distance from the principle point using regression analysis (SAS 2004). Rationale for Object-Based Classification To classify and describe patch structure in the vegetation of the Mackenzie Delta uplands we used an object-based approach. This method enables multi-scale description of patch structure and has been used successfully to map fine scale pattern using air photos in shrub dominated ecosystems (Laliberte et al. 2004). An object-based approach to image classification differs from conventional pixel-based methods by assigning class membership to groups of pixels (objects) rather than individual pixels. It is essentially a two step process that involves segmenting image pixels into image objects followed by 110  the classification of these objects. In the Definiens software package, segmentation is bottom-up region merging algorithm that optimizes object creation by minimizing their internal heterogeneity. The region merging process stops when the heterogeneity of an object exceeds a threshold defined by a unitless scale parameter. User modification of this threshold results in the creation of larger (higher heterogeneity), or smaller objects (lower heterogeneity) (Benz et al. 2004). By iteratively modifying the heterogeneity threshold (by changing the scale parameter) users segment the image into objects that reflect the structure of the landscape (Blaschke and Hay 2001). Users can also change the relative weighting of spectral vs. shape characteristics, and compactness vs. smoothness criteria in the segmentation algorithm (Benz et al. 2004). Subsequent classification of image objects is accomplished by using defined membership rules or a nearest neighbor classification based on the selection of training data (Laliberte et al. 2004). By creating a nested hierarchy of image objects derived from a multi-scale segmentation, an object-based approach also makes use of contextual information not otherwise contained in pixels (Blaschke and Blaschke 2003). Object-Based Segmentation and Classification Procedure Object-based classification (Definiens 2006) was used to divide each image into four classes reflecting the structure of the dominant vegetation. Classes included: 1) shrub tundra (vegetation dominated by Alnus viridis, tall Salix spp. and Betula glandulosa), 2) dwarf shrub tundra (vegetation dominated by ericaceous dwarf shrubs and sedges), 3) water (lakes, rivers, ponds, and the Beaufort Sea), and 4) bare ground. Wet sedge dominated communities are infrequent across the study area (Corns 1974), and thus were grouped with dwarf shrub tundra. Bare ground occupied less 0.05% of the total area mapped and thus it was not feasible to include this cover type in accuracy assessments.  111  Figure 5.2. Diagram showing the sequence of operations in the object-based classification each upland tundra plot. Panel 1. Aerial photo of upland tundra plot at coarse (1A) and fine (1B) scales. Panel 2. Segmentation of the image at a coarse scale (2A) is followed by segmentation at a fine scale (2B). Panel 3. Coarse scale objects are classified into land and water (3A) and fine scale object are classified as small water bodies, shrub tundra and dwarf shrub tundra (4B), constrained by their membership in the coarse scale classification (4A). 112  Segmentation To minimize confusion between water and dark coloured shrub tundra in this lake rich region, we segmented each image at two scales. First we segmented images into large heterogeneous objects whose borders corresponded to the boundary between water bodies and land (Figure 5.2). These coarse-scale objects were created by performing segmentation on the red, green and blue bands, as well two texture measures. Textural co-occurrence measures (contrast and entropy) were calculated using a greyscale band (ENVI 2006). Segmentation was performed using a scale parameter of 600, a colour to shape ratio of 0.9-0.1 and a compactness to smoothness ratio of 0.5 to 0.5. Secondly we performed a fine-scale segmentation to distinguish between shrub tundra and dwarf shrub tundra, using a scale parameter that yielded small homogenous objects that approximated the size and shape of isolated patches of shrub tundra or dwarf shrub tundra (Figure 5.2). These fine-scale objects were by created by performing segmentation of the red, green and blue bands using a scale parameter of 25, a colour to shape to ratio of 0.9-0.1, and compactness to smoothness to 0.5 to 0.5. Classification Classification of these groups of objects was also performed in two steps. First we used a standard nearest neighbour (SNN) classifier to define coarse-scale objects as either land or water. Subsequently, we assigned fine-scale objects within the broader “land” to shrub tundra, dwarf shrub tundra, smaller water body classes bodies, or bare ground using a SNN classifier. Fine-scale objects contained within the broader “water” class were automatically classified as water. To perform each SNN classifications we used a combination of vegetation plots, high resolution ground truth images and experience in the Delta region to select training areas. Training images were collected in the summer of 2006 using a Canon PowerShot S80 digital camera mounted on a helicopter. Photographs were captured at an altitude of ~450 m, had pixel sizes typically less than 0.25 m (Figure 5.3), and were georeferenced in ARCGIS using the 1:30 000 scale orthophotographs of each plot. We used Definiens to calculate the total area of each cover type and the proportion of shrub tundra in each photo. We also used Definiens to merge all contiguous objects of the same cover type and subsequently to 113  determine the total number of patches and mean patch size for shrub tundra and dwarf shrub tundra classes. Accuracy Assessment To examine the accuracy of our object-based based classifications we constructed confusion matrices using two methods. Ground truth data were derived from independent manual classifications of 5 randomly selected images collected in the same manner as our training images. First, to conduct a standard pixel-based accuracy assessment, we compared 1500 random points (300 / ground truth image) from each cover type in our classifications with classified ground truth photos. These comparisons also provided estimates of errors of commission and omission of shrub tundra and dwarf shrub tundra that we used to evaluate our estimates of the proportion shrub tundra and dwarf shrub tundra. Since we were also interested in quantifying changes in patch sizes we also conducted a polygon-based accuracy assessment (at the object level) to evaluate our estimates of patch sizes. To do this we compared 1500 randomly selected objects from each cover type with the ground-truth classification. We calculated overall accuracy, per class user’s and producer’s accuracies and the kappa statistic which better represents accuracies for cover types of different relative abundances.  114  Figure 5.3. Example image from helicopter surveys used to train image classifiers and in accuracy assessments. Image shows shrub tundra (shades of green and yellow), dwarf shrub tundra (lavender grey), and two small lakes (blue).  115  Regional Biophysical Variability To characterize variability in regional temperature, microtopography and soil development that may be correlated with the dominance of shrub tundra or dwarf shrub tundra, we compiled information from several data sources. To document the variability in temperature in the study area thermistors in radiation shields mounted 1.5 meters above the ground (Onset Computer Corporation, Pocasset MA HOBOTM, H08-030-08, RS1) recorded the temperature every hour at five locations in the study area (Figure 5.1). Overall we established 8 temperature loggers in the study area, but only five of them were within the study plots sampled here. Consequently, we also used regression of latitude vs. mean summer temperature (F1, 8 =293, r2 =0.9734, P< 0.001) to estimate the average temperature for all study plots (Table 5.1). As an index of potential terrain moisture we calculated a topographic wetness index (TWI) for each plot using the hydrology tools in ARCGIS and a 30 m resolution digital elevation model of the land surface (Wilson and Gallant 2000). Subsequently we classified each TWI image as: 1) well drained (TWI: 0-30), and poorly drained (TWI >60) and calculated the proportion of these potential terrain moisture classes. Using the Soil Landscapes of Canada we calculated the proportion of the land surface occupied by the dominant soil types. These included brunosolic cryosols (permafrost soils with well developed B horizons) and orthic cryosols (permafrost soils with poorly developed B horizons) (Soil Landscapes of Canada Working Group 2007). Statistical Analyses Objective 1: Describe latitudinal changes in the proportion and patch sizes of shrub tundra and dwarf shrub tundra. To describe changes in the proportion and patch sizes of shrub tundra and dwarf shrub tundra with latitude we compared linear and non linear models (R Development Core Team 2006). Model performance was evaluated by comparing Akaike's information criterion (AIC), and AIC weights (Anderson et al. 2000). We examined residual plots to ensure that models met the assumptions of equal variance and normality.  116  117  Latitude  68°37.6'N 68°42.7' N 68°52.7' N 68°55.1' N 69°00.3' N 69°02.4' N 69°05.6' N 69°08.3' N 69°16.2' N 69°18.9' N 69°21.5' N 69°21.4' N 69°26.8' N 69°29.4' N 69°31.9' N 69°34.4' N 69°37.3' N 69°40.3' N  Plot  1 2 3 4 5 6 7 8 9 10 11 13 12 14 15 16 17 18  133°07.1'W 133°47.1'W 133°55.6'W 133°56.2'W 133°27.4'W 134°13.0'W 134°14.7'W 134°11.6'W 133°12.0'W 134°32.9'W 134°18.5'W 134°22.6'W 133°54.6'W 134°03.9'W 134°02.9'W 134°15.4'W 134°14.8'W 134°19.3'W  Longitude 9.3 (9.2) 9.0 8.6 8.4 (8.4) 8.2 8.1 7.9 7.8 (7.8) 7.4 (7.6) 7.2 7.1 7.1 6.9 6.8 (6.77) 6.6 6.5 6.3 6.2  Temperature (°C) 0.31 0.40 0.56 0.11 0.39 0.32 0.04 0.28 0.13 0.35 0.07 0.39 0.28 0.06 0.46 0.20 0.02 0.34  Well Drained Terrain 0.12 0.07 0.04 0.45 0.09 0.13 0.69 0.14 0.25 0.17 0.62 0.08 0.12 0.60 0.06 0.22 0.77 0.18  Poorly Drained Terrain  Brunosolic Cryosols 0 0 0 0 1 0 0 0 0 0 0.55 0.20 0.36 1 1 0.14 0.55 0.24  Orthic Cryosols 1 1 1 1 0 1 1 1 1 1 0.45 0.80 0 0 0 0.86 0.45 0.76  Table 5.1. Biophysical characteristics of each sample plot. Temperatures shown were estimated using regression of latitude with data from all temperature loggers in the study area (F1, 8 =293, r2 =0.9734, P< 0.001). Measured temperatures for sample plots that had loggers are shown in parentheses. All other values are proportions of the land surface in each plot.  Objective 2: Determine the relative importance of biophyscical drivers of the shrub tundra ecotone. To explore the relative importance of the biophysical variables shown in Table 5.1 we conducted several analyses. First we examined the strength of relationships between the proportion of shrub tundra, latitude, temperature, proportion of well and poorly drained terrain, and the proportion of orthic cryosols and brunosolic cryosols by calculating the Spearman rank correlation coefficient for all variable combinations (SAS 2004). Correlations between temperature and other biophysical variables were based on measured temperatures. To estimate the relative importance of correlated biophysical variables in models predicting the proportion of shrub tundra, we used hierarchical variance partitioning. In this approach the independent and joint effects of a single variable are estimated by a hierarchical comparison of all possible models (2k, k= number of explanatory variables). For example, to estimate the independent effect of mean summer temperature: xtemp, the fit (R2) of a model containing xtemp is compared to the fit of the models excluding xtemp at each level in the hierarchy of model complexity. Subsequently, the independent contribution of xtemp is estimated by averaging the improvement in fit for models containing xtemp across all levels in the hierarchy. The joint contribution of xtemp and all other variables equals the difference between the independent contribution of xtemp and the variance explained by a model including xtemp as the only explanatory variable (Quinn and Keough 2002, Millington et al. 2007). We used hierarchical variance portioning in R, (hier.part, R Version 2.6.2, R Development Core Team 2006) to estimate the total independent and joint contribution of variables significantly correlated with the proportion of shrub tundra (i.e. mean temperature, the proportion of orthic cryosols, and the proportion of brunosolic cryosols). Since only 5 of 18 sample plots had temperature loggers, we conducted 2 separate analyses. In Analysis 1 we used estimated temperatures and biophysical data from all plots in the study area. In Analysis 2 we only used biophysical data from the five plots with temperature loggers. In both analyses we assessed the significance of each independent contribution using 100 randomizations of the original data (hier.part, R Version 2.6.2, R Development Core Team 2006).  118  RESULTS Image Brightness and Accuracy Assessment Images did not exhibit any consistent systematic brightness bias. Regressions of brightness with distance varied from weakly positive to weakly negative, did not explain more that than 5% of the variance, and were not significant (Figure 5.4). Pixel-and object-based estimates of overall classification accuracy were 83.8% and 69.9% percent, respectively. The estimate of the kappa statistic using a pixel-based method was 0.787 and 0.664 using a polygon-based method (Table 5.2). Both methods of estimating classification accuracy indicate that water bodies were extremely well classified (user and producers accuracies 93.3%-97.6%), but that shrub tundra and dwarf shrub tundra were more likely to be misclassified and confused with each other (user and producers accuracy 66.6%-83.8%). Specifically, user accuracies for shrub tundra were 76.9% and 75.3%, while producer’s accuracies were 81% and 77.1% for pixel and object-based estimation methods, respectively. For dwarf shrub tundra users accuracies were 83.8 % and 69.2 %, and producers accuracies were 79.1% and 66.6% for pixel and object-based estimation methods, respectively (Table 5.2). Errors of commission and omission for a given class and accuracy assessment method were of similar magnitude (Table 5.2). Objective 1: Describe latitudinal changes in the proportion and patch sizes of shrub tundra and dwarf shrub tundra. Moving southward across the study area, the proportional area and mean patch size of shrub tundra increased (Figure 5.5). This change was driven both by an increase in mean shrub tundra patch size and the number of shrub tundra patches. This southward increase in the dominance of shrub tundra was paralled by a northward increase in the abundance and patch size of dwarf shrub tundra (Figure 5.6). This change was driven primarily by an increase in dwarf tundra patch size. All models showed strong evidence of non-linear relationships and non-linear models had improved fit and lower AIC’s (Table 5.3). All models showed the greatest changes in relative abundance north of 68.768.9°N. 119  200  180  Brightness Value  160  140  120  100  80  p=0.9201 r2<0.001  60 180 160  Brightness Value  140 120 100 80 60 40 20 200  p=0.4438 r2=0.027  180  Brightness Value  160 140 120 100 80 60 40  p=0.6988 r2>0.01 0  1000  2000  3000  4000  Distance from Principal Point (m)  Figure 5.4. To evaluate possible bias in brightness values within airphotos, the relationship between pixel brightness vs. distance from principle point was examined in three sample plots in the study area. Points plotted are sampled at successive distances of 180 m from the principle point of each photo to avoid autocorrelation associated with vegetation patch structure.  120  SHRUB TUNDRA 0.7  Proportion of Study Plot  A 0.6  0.5  0.4  0.3  0.2  0.1 800  p<0.001 r2=0.81  B  Mean Patch Size (m2)  700 600 500 400 300 200 100 0  p<0.001 r2=0.73 68.6  68.8  69.0  69.2  69.4  69.6  Latitude Figure 5.5. Least squares regressions of: A) proportion of shrub-tundra vs. latitude (y=73.3495x - 0.5336x2 - 2519.9439, F2, 15 = 38.25, P<0.001 Adjusted R2=0.8142) and B) mean shrub tundra patch size vs. latitude (y=63013x - 458.8x2 - 2163009, F2, 2 15 = 24.97, P<0.001 Adjusted R =0.7382).  121  DWARF SHRUB TUNDRA Proportion of Study Plot  1.0  A  0.8  0.6  0.4  p<0.001 r2=0.81  0.2  6000  B  Mean Patch Size (m2)  5000  4000  3000  2000  1000  p<0.001 r2=0.64  0 68.6  68.8  69.0  69.2  69.4  69.6  Latitude Figure 5.6. Least squares regressions of: A) proportion of dwarf shrub tundra vs. latitude (y=-73.3495x + 0.5336x2 + 2520.9439, F2, 15 = 38.25, P<0.001 Adjusted R2=0.8142) and B) mean dwarf shrub patch size vs. latitude (y=-759116x + 5512x2 + 26137774, F2, 15 = 15.97, P<0.001 Adjusted R2=0.6379).  122  Objective 2: Determine the relative importance of biophysical drivers of the shrub tundra ecotone. Mean summer temperature decreased with increasing latitude at the five locations where we had temperature loggers within the study area (Table 5.1). The strong correlation between latitude and temperature (Table 5.4) is consistent with linear regression analysis of summer temperatures at a larger number of sites, which shows that on average there is decrease of 3°C in mean temperature for every degree of increasing latitude between 67.6°N and 69.5°N (Chapter 3). The proportion of orthic cryosols also decreased moving northward (Tables 5.1, 5.4). Conversely, the proportion of brunosolic cryosols was positively correlated with latitude. Although they varied across the study area, the proportion of well and poorly drained terrain was not significantly correlated with latitude (Tables 5.1, 5.4). Biophysical variables that were significantly correlated with latitude were also significantly correlated with the proportion of shrub tundra (Tables 5.1, 5.4). Hierarchical variance partitioning of models including biophysical variables significantly correlated with shrub tundra show that summer temperature made the largest single contribution to the variance explained (76.7%). The proportion of orthic cryosols also made a independent contribution (14.6%) to the variance explained (Table 5.5). In the analysis based only on plots with temperature loggers summer temperature also explained the majority of variance in the proportion of shrub tundra. In both analyses summer temperature was the only significant explanatory variable.  123  124  Ground Truth Data Water 1451 31 4 97.6% 2.4% 85.8% 0.787 Ground Truth Data Water 850 44 10 94.0% 6.0% 78.12% 0.664  PIXEL-BASED Classified Data Water Shrub Tundra Dwarf Shrub Tundra Producers’ Accuracies Commission Error Overall Accuracy Kappa Coefficient  OBJECT-BASED Classified Data Water Shrub Tundra Dwarf Shrub Tundra Producers’ Accuracies Commission Error Overall Accuracy Kappa Coefficient Shrub Tundra 27 1163 319 77.1% 22.9%  Shrub Tundra 31 1154 239 81.0% 19.0%  Dwarf Shrub Tundra 34 337 740 66.6% 33.4%  Dwarf Shrub Tundra 18 315 1257 79.1% 20.9%  User’s Accuracies 93.3% 75.3% 69.2%  User’s Accuracies 96.7% 76.9% 83.8%  Omission Error 6.7% 24.7% 30.8%  Omission Error 3.3% 23.1% 16.2%  Table 5.2. Classification accuracy assessments. Table shows raw tallies, producers and users accuracies, errors of commission and omission, overall accuracy and the kappa coefficient.  125  Dwarf Shrub Tundra Patch Size  Shrub Tundra Patch Size  Proportion Dwarf Shrub Tundra  Dependant Variable Proportion Shrub Tundra  Model Latitude Latitude2 Latitude + Latitude2 Latitude Latitude2 Latitude + Latitude2 Latitude Latitude2 Latitude + Latitude2 Latitude Latitude2 Latitude + Latitude2  AIC ΔAIC Adjusted R2 AIC Weight -85.82 4.97 0.740 0.071 -85.89 4.90 0.741 0.074 -90.78 0 0.811 0.855 -85.48 5.12 0.740 0.067 -85.55 5.04 0.741 0.069 -90.60 0 0.813 0.864 165.90 2.08 0.690 0.305 165.85 2.03 0.691 0.313 163.82 0 0.737 0.864 249.79 3.87 0.530 0.125 249.74 3.83 0.532 0.127 245.92 0 0.639 0.864  Table 5.3. Comparison of linear and non-linear models using Adjusted R2, AIC, and AIC weights. Best models (shown in bold) are plotted in Figures 5.4 and 5.5.  -0.565 0.603 -0.222 0.152 0.900 -0.891 1.000  0.644 -0.646 0.356 -0.275 -1.000 1.000  Shrub Tundra Latitude  Summer Well Drained Poorly Drained Orthic Turbic Temperature Terrain Terrain Cryosols -0.707 -0.076 0.061 -0.982 0.707 0.055 -0.008 1.000 -0.700 1.000 -0.940 0.700 1.000 1.000  Brunoslic Cryosols 1.000  Table 5.4. Spearman rank correlation coefficients for comparisons among the proportion of shrub tundra and biophysical variables. Significant (α= 0.05) rank correlations are shown in bold.  Brunoslic Cryosols Orthic Turbic Cryosols Poorly Drained Terrain Well Drained Terrain Summer Temperature Latitude Shrub Tundra  126  127  Joint Contribution 0.142 0.129 0.092 Joint Contribution 0.433 0.453 0.453  Analysis 1: All plot and estimated temperatures Independent Dependent Variable Explanatory variables Contribution Proportion Shrub Tundra Temperature 0.611 Orthic Cyrosols 0.116 Brunosolic Cryosols 0.069  Analysis 2: Plots with measured temperatures Independent Dependent Variable Explanatory variables Contribution Proportion Shrub Tundra Temperature 0.479 Orthic Cyrosols 0.236 Brunosolic Cryosols 0.236  Total Contribution 0.912 0.689 0.689  Total Contribution 0.753 0.245 0.161  Percent Independent 50.32 24.84 24.84  Percent Independent 76.76 14.57 8.67  Table 5.5. Results of the hierarchical variance partitioning analysis. Explanatory variables are sorted by the strength of their independent contribution. Significant independent contributions are bolded.  DISCUSSION Latitudinal Changes in Shrub Abundance: Implications for Global Temperatures Our classifications of the vegetation in the Mackenzie Delta uplands document a gradual transition between shrub and dwarf shrub dominated ecosystems, which begins north of 68.7°N. This large ecotone is often mapped as a homogeneous cover type (Gould et al. 2002, Walker et al. 2002, Gould et al. 2003), yet variability in the relative abundance and patch size of shrub tundra and dwarf shrub tundra is clearly evident. The effect of shrub tundra on the duration of the snow free season and albedo (Epstein et al. 2004a, Chapin et al. 2005b) make accurate maps of this transition particularly important for global climate models and underscores the importance of matching classification grain with the intended map use. Although our study area represents a small portion of the tundra biome, over-generalization of vegetation structure and map inaccuracies in this transition have a potentially large influence on global models. For example, it is estimated that shrub expansion in tundra ecosystems may increase atmospheric heating by between 6.4 and 8.9 W/m2, approximately twice that expected from a doubling in atmospheric CO2 (Chapin et al. 2005b, Sturm et al. 2005). Based on the results presented here, we conclude that the portion of the study area north of 69.36°N (where shrub tundra < 30%) should be mapped as dwarf shrub tundra rather than to homogeneous shrub tundra (Walker et al. 2002). Thus, maps depicting this area as homogeneous shrub tundra would overestimate regional heating by between 8.7 and 12.1 TW, if used in models of global heat balance. Classification accuracy also has a potentially large influence on predictions of heat balance from biophysical models. Here we mapped vegetation of the Mackenzie Delta uplands with an overall accuracy of 86% with accuracies ranging from 77% to 84% for shrub classes. Overall this represents a significant improvement over the accuracy of previous classifications of the region, which had per class accuracies often less than 50% (IEG 2002). However, errors of omission and commission between 16% and 23% for vegetation classes are likely to greatly reduce predictive accuracy of biophysical models used to predict variables such as: active layer thickness, methane flux and carbon dioxide flux (Nelson et al. 1997, Reeburgh et al. 1998, Oechel et al. 2000). Our classifications 128  were based on true colour airphotos, and it is likely that object-based classifications using high resolution multispectral imagery would improve accuracy. Additional contextual layers including object elevation within a high resolution DEM, and object texture could also be used to further refine object-based classifications of these ecosystems (Dorren et al. 2003, Bock et al. 2005). Latitudinal Changes in Shrub Patch Size: Implications for Regional Models Feedbacks between shrub cover and the tundra microenvironment make the position of this ecotone equally important to regional biophysical models (~10 000 km2). For example, increased snow pack around upright shrubs has a strong effect on ground temperatures (Sturm et al. 2001a, Mackay and Burn 2002, Pomeroy et al. 2006, Palmer 2007, Johnstone and Kokelj In Press), making the current and future position of the shrub tundra ecotone important in infrastructure planning, such as oil and gas pipelines. However, feedbacks between shrub cover and microenvironment may only operate over a narrow range of patch sizes (Pomeroy et al. 1995, Sturm et al. 2001b, Grogan and Jonasson 2006) making some parts of this transition more sensitive than others. For example, it is likely that shrub proliferation in terrain currently dominated by large continuous patches of shrub tundra (e.g. south of 69.2°N) may have little added impact on snow pack and ground temperatures. However, shrub proliferation in areas where shrub patches are small and dispersed (e.g. north of 69.2°N) may drive large localized changes in microenvironment affecting terrain stability. Thus, decreases in mean shrub tundra patch size with latitude suggest that not all portions of the transition are likely to be equally sensitive to changes in ground temperature initiated by shrub proliferation. Research to identify which patch sizes of shrub tundra are associated with sharp abiotic changes, in conjunction with improved maps of fine scale patch structure, could help identify areas sensitive to changes in ground temperature and active layer depth. The accuracy of these maps is also likely to be critical to realistic predictions of thaw sensitivity. Our object-based accuracy assessment yielded lower overall and per class accuracies than a pixel-based approach, suggesting that our estimates of patch size are less reliable than our estimates of overall cover. It is possible that higher errors were 129  related to disparity between the scales of manual and semi-automated polygon delineation (Wulder et al. 2006). Our manual classifications tended to have a broader scale of generalization than the semi-automated classifications produced using Definiens. Consequently, very small objects classified in Definiens were frequently not delineated in manual classifications, likely increasing errors of commission and omission of small objects. Despite uncertainty associated with the source of these errors, our results clearly show that relative patch size changes with latitude and emphasizes the need for additional research. Given potential mismatches between classification and vector based reference data, additional work to assess the change in absolute patch size should be conducted using plot-based field data. Patch Size and Positive Shrub-Snow-Nutrient Feedbacks Sturm et al. (2001a, 2005) have proposed that abiotic changes resulting from shrub proliferation will drive a strong positive feedback loop that facilitates continued shrub expansion in the Low Arctic. Changes in snow pack and ground temperatures associated with shrub tundra have also been associated with increased nutrient availability (Schimel et al. 2004, Sturm et al. 2005). Since deciduous shrubs show rapid increases in response to experimental nutrient addition (Chapin et al. 1995, Dormann and Woodin 2002, Walker et al. 2006), it is likely that elevated nutrient availability caused by deeper snow pack will promote further shrub growth and expansion (Sturm et al. 2001a, Sturm et al. 2005). Estimates suggest that changes in shrub cover in the Low Arctic could extend the duration of soil microbial activity by up to two months, thereby significantly increasing nutrient mineralization (Sturm et al. 2005). Increased dominance by nitrogen fixing shrubs such as green alder (Alnus viridis subsp. fruticosa) is also likely to increase soil nitrogen levels (Rhoades et al. 2001). Since a shrub nutrient availability feedback mechanism is mediated by increased snow pack (Sturm et al. 2001a, Sturm et al. 2005), it is likely that there is a portion of the shrub-tundra ecotone where increases in shrub cover may facilitate continued expansion and another where shrub proliferation may inhibit additional change. For example, in portions of our study area completely dominated by shrubs, a lack of redistribution of snow by wind creates relatively uniform snow pack depth (Pomeroy et al. 1995, Palmer 130  2007) and may not increase ground temperature around individual shrubs relative to patches without shrub cover. Additionally, at certain densities ground shading by a thick shrub canopy would likely reduce ground heat flux (Epstein et al. 2004a, Thompson et al. 2004), driving permafrost aggradation. Thinning of the active layer would reduce microbial access to carbon and nitrogen sources during the winter, decreasing nutrient availability and favouring dwarf evergreen shrubs (Kielland 1994, McKane et al. 2002). Conversely, in areas with low shrub tundra cover, an increase in the number individual shrub patches could create abiotic conditions favouring short term shrub expansion. However, as individual shrub patches coalesce into large patches the direction of the shrub-microenvironment feedbacks described by Sturm et al. (2001a, 2005) are likely to shift from positive to negative. Relative importance of biophysical drivers of the shrub tundra ecotone. Regional differences in summer temperature are the primary determinant of the position of shrub tundra ecotone. In the Mackenzie Delta uplands, the latitude at which the proportion of shrub tundra declines below 50% corresponds approximately to the mean 10°C July isotherm (Pelletier 2008). A similar correlation between with the tussock tundra-shrub tundra boundary has been also been described in Alaska (Muller et al. 1999, Walker 2000). This is also the approximate latitude at which average green alder seed viability declines below 5% (Chapter 2). This suggests that cold summer temperatures limit the abundance of shrub tundra at higher latitudes, likely by reducing the probability of recruitment (Chapter 2, Hobbie and Chapin 1998). This interpretation is also consistent with the results of the correlation and hierarchical variance partitioning analyses presented here. Our results also suggest that soil type may be of secondary importance. Soil properties, particularly pH, are frequently emphasized as an important driver of both fine and broad scale differences in tundra vegetation (Timoney et al. 1993, Walker et al. 1994, Gough et al. 2000, Walker 2000). Warm temperatures and the abundance of orthic cryosols (permafrost soils with poorly developed B horizons) are both correlated with increased shrub tundra. The contribution of the proportion of orthic cryosols to models of shrub tundra suggest that increasing dominance of soils with poorly developed mineral 131  horizons is also linked to the distribution of shrub tundra. It is also possible these differences in soil type correspond to a broad scale pH gradient. Unfortunately, beyond broad scale mapping, little descriptive research has been conducted on variability of soils in the Mackenzie Delta uplands. Additional field investigations and more detailed soil mapping are required to test this hypothesis. Fine scale topographic moisture gradients have also been linked to the distribution of Arctic plant communities (Walker et al. 1994, Walker 2000). However, in the Mackenzie Delta Region, the lack of correlation between potential terrain moisture and the proportion of shrub tundra suggests that the low rolling topography in our study area is not a primary determinant of broad scale vegetation structure. In Alaska, finer scale topographic variability has been used to successfully predict vegetation structure, but model performance declines at broader scales (Ostendorf and Reynolds 1998). Although fine-scale variability in Low Arctic plant communities has clearly been shown to be related to topography (Walker et al. 1994) , based on the results presented here we conclude that fine scale influence of topographic moisture gradients is constrained by regional differences in summer temperature. Future Shrub Proliferation Vegetation transitions in the Arctic predicted to show rapid responses to changing climate are those which exhibit gradual (rather than abrupt) changes in abundance across ecotones, and are more strongly correlated with climatic, rather than soil, conditions (Epstein et al. 2004a). Shrub abundance in the uplands of the Mackenzie Delta changes gradually across a distance of approximately 80 km. As we have shown, changes are also more strongly correlated with regional temperature differences than broad scale differences in topography and soil type. This strong correspondence between summer temperatures and shrub abundance implies that increasing summer temperatures in the region (Lantz and Kokelj 2008) will increase shrub abundance. In Arctic Alaska increases in the abundance of upright shrubs evident on historical photographs have been attributed to direct and indirect effects of regional temperature increases (Tape et al. 2006). Plot level alterations of temperature and nutrient availability also provide evidence that temperatures increases will drive changes in shrub dominance (Parsons et 132  al. 1994, Chapin et al. 1995, Bret-Harte et al. 2001, Bret-Harte et al. 2002, Dormann and Woodin 2002, Walker et al. 2006). Age distributions of green alder populations in the northern forest-tundra transition zone also show evidence of recent recruitment (Chapter 2). Since changes in shrub abundance will have important feedbacks to regional microenvironmental processes and the global climate system, accurate mapping of fine scale differences in vegetation types is essential both to parameterize biophysical models and to use as a baseline to measure future changes. Conclusions To date, there has been insufficient research describing and mapping the transition between shrub tundra and dwarf shrub tundra in the Mackenzie Delta uplands. However, it is clear that the relative abundance of shrub tundra across this ecotone has implications for global climate and regional biophysical models. Furthermore, changes in shrub patch size across this ecotone are also likely to influence abiotic feedbacks arising from shrub proliferation. Consequently, better understanding and improved mapping of the spatial heterogeneity in this transition is critical.  ACKNOWLEDGEMENTS The authors would like to acknowledge the following individuals for the contribution to the development of this paper. For assistance in the field and laboratory we thank: Sahara Borgarti, Douglas Esagok, Robert Jenkins, Stephen Schwarz, Matt Tomlison, and Rory Tooke. We would also like to thank Greg Henry for helpful comments on drafts of this manuscript. Funding support was received from: Aurora Research Institute (Research Fellowship), Canon USA and the AAAS (Canon National Parks Science Scholarship), Global Forest Research (Research Grant GF-18-2004-212), Indian and Northern Affairs Canada (Water Resources Division, the Northern Science Training Program, and the Mackenzie Valley Airphoto Project), Killiam Trusts (Predoctoral Fellowship), Natural  133  Resources Canada (Polar Continental Shelf Program), Natural Sciences and Engineering Research Council of Canada (PGS-B and Northern Internship to T.C Lantz).  134  LITERATURE CITED ACIA. 2004. Arctic Climate Impact Assessment: Impacts of Warming Climate. . Cambridge University Press, Cambridge. Anderson, D. R., K. P. Burnham, and W. L. Thompson. 2000. Null hypothesis testing: Problems, prevalence, and an alternative. Journal of Wildlife Management 64:912-923. Aylsworth, J. M., M. M. Burgess, D. T. Desrochers, A. Duk-Rodkin, T. Robertson, and J. A. Traynor. 2000. Surficial geology, subsurface materials, and thaw sensitivity of sediments. Pages 41-48 in L. D. Dyke and G. R. Brooks, editors. The phsyical environment of the Mackenzie Valley, Northwest Territories: A base line for the assessment of environmental change. Geological Survey of Canada Bulletin 547. Geological Survey of Canada, Ottawa, ON. Benz, U. C., P. Hofmann, G. Willhauck, I. Lingenfelder, and M. Heynen. 2004. Multiresolution, object-oriented fuzzy analysis of remote sensing data for GIS-ready information. Isprs Journal of Photogrammetry and Remote Sensing 58:239-258. Blaschke, T., and T. Blaschke. 2003. Object-based contextual image classification built on image segmentation Object-based contextual image classification built on image segmentation. Pages 113-119 in Advances in Techniques for Analysis of Remotely Sensed Data, 2003 IEEE Workshop on. Blaschke, T., and G. J. Hay. 2001. Object-oriented image analysis and scale-space: theory and methods for modeling and evaluating multiscale landscape structure International Archives of Photogrammetry and Remote Sensing 34:22- 29. Bliss, L. C., and N. V. Matveyeva. 1992. Circumpolar Arctic Vegetation. Pages 59-89 in F. S. Chapin, editor. Arctic ecosystems in a changing climate : an ecophysiological perspective. Academic Press, San Diego. Bock, M., P. Xofis, J. Mitchley, G. Rossner, and M. Wissen. 2005. Object-oriented methods for habitat mapping at multiple scales - Case studies from Northern Germany and Wye Downs, UK. Journal for Nature Conservation 13:75-89. Bret-Harte, M. S., G. R. Shaver, and F. S. Chapin. 2002. Primary and secondary stem growth in arctic shrubs: implications for community response to environmental change. Journal of Ecology 90:251-267. Bret-Harte, M. S., G. R. Shaver, J. P. Zoerner, J. F. Johnstone, J. L. Wagner, A. S. Chavez, R. F. Gunkelman, S. C. Lippert, and J. A. Laundre. 2001. Developmental plasticity allows Betula nana to dominate tundra subjected to an altered environment. Ecology 82:18-32. Burn, C. R. 1997. Cryostratigraphy, paleogeography, and climate change during the early Holocene warm interval, western Arctic coast, Canada. Canadian Journal of Earth Sciences 34:912-925. Chapin, F. S., M. Berman, T. V. Callaghan, P. Convey, A.-S. Crépin, K. Danell, H. W. Ducklow, B. Forbes, G. Kofinas, A. D. McGuire, M. Nuttall, R. Virginia, O. Young, S. E. Zimov, T. Christensen, A. Godduhn, E. J. Murphy, D. Wall, C. Zockler, and B. Fitzharris. 2005a. Polar systems. Pages 717-743 in R. Hassan, R. Scholes, and N. Ash, editors. Millennium Ecosystem Assessment: Ecosystems 135  and human well-being: current state and trends, Volume 1. Island Press, Washington, D.C. Chapin, F. S., A. D. McGuire, J. Randerson, R. Pielke, D. Baldocchi, S. E. Hobbie, N. Roulet, W. Eugster, E. Kasischke, E. B. Rastetter, S. A. Zimov, and S. W. Running. 2000. Arctic and boreal ecosystems of western North America as components of the climate system. Global Change Biology 6:211-223. Chapin, F. S., G. R. Shaver, A. E. Giblin, K. J. Nadelhoffer, and J. A. Laundre. 1995. Responses of Arctic Tundra to Experimental and Observed Changes in Climate. Ecology 76:694-711. Chapin, F. S., M. Sturm, M. C. Serreze, J. P. McFadden, J. R. Key, A. H. Lloyd, A. D. McGuire, T. S. Rupp, A. H. Lynch, J. P. Schimel, J. Beringer, W. L. Chapman, H. E. Epstein, E. S. Euskirchen, L. D. Hinzman, G. Jia, C. L. Ping, K. D. Tape, C. D. C. Thompson, D. A. Walker, and J. M. Welker. 2005b. Role of land-surface changes in Arctic summer warming. Science 310:657-660. Corns, I. G. W. 1974. Arctic Plant Communities East of Mackenzie-Delta. Canadian Journal of Botany-Revue Canadienne De Botanique 52:1731-&. Definiens. 2006. Definiens Professional 5 User Guide. Definiens AG, Munich. Dormann, C. F., and S. J. Woodin. 2002. Climate change in the Arctic: using plant functional types in a meta-analysis of field experiments. Functional Ecology 16:417. Dorren, L. K. A., B. Maier, and A. C. Seijmonsbergen. 2003. Improved Landsat-based forest mapping in steep mountainous terrain using object-based classification. Forest Ecology and Management 183:31-46. ENVI. 2006. ENVI User's Guide. Version 4.3. Research Systems, Inc. Epstein, H. E., J. Beringer, W. A. Gould, A. H. Lloyd, C. D. Thompson, F. S. Chapin, G. J. Michaelson, C. L. Ping, T. S. Rupp, and D. A. Walker. 2004a. The nature of spatial transitions in the Arctic. Journal of Biogeography 31:1917-1933. Epstein, H. E., M. P. Calef, M. D. Walker, F. S. Chapin, and A. M. Starfield. 2004b. Detecting changes in arctic tundra plant communities in response to warming over decadal time scales. Global Change Biology 10:1325-1334. Euskirchen, E. S., A. D. Mcguire, and F. S. Chapin. 2007. Energy feedbacks of northern high-latitude ecosystems to the climate system due to reduced snow cover during 20th century warming. Global Change Biology 13:2425-2438. Forest Management Institute. 1975. Vegetation types of the Lower Mackenzie and Yukon corridor. Canadian Forest Service, Ottawa. Geological Survey of Canada. 2002. DEM Mackenzie River Valley corridor (spatial resolution 30m) based on 1:50,000 scale basemaps. in. Natural Resources Canada, Geological Survey of Canada. Gough, L., G. R. Shaver, J. Carroll, D. L. Royer, and J. A. Laundre. 2000. Vascular plant species richness in Alaskan arctic tundra: the importance of soil pH. Journal of Ecology 88:54-66. Gould, W. A., S. Edlund, S. Zoltai, M. Raynolds, D. A. Walker, and H. Maier. 2002. Canadian Arctic vegetation mapping. International Journal of Remote Sensing 23:4597-4609.  136  Gould, W. A., M. Raynolds, and D. A. Walker. 2003. Vegetation, plant biomass, and net primary productivity patterns in the Canadian Arctic. Journal of Geophysical Research-Atmospheres 108:-. Grogan, P., and S. Jonasson. 2006. Ecosystem CO2 production during winter in a Swedish subarctic region: the relative importance of climate and vegetation type. Global Change Biology 12:1479-1495. Hassol, S. J. 2004. Arctic Climate Impact Assessment: Impacts of Warming Climate. Cambridge University Press, Cambridge. Hobbie, S. E., and F. S. Chapin. 1998. An experimental test of the limits to tree establishment in Arctic tundra. . Journal of Ecology 86:449-161. IEG. 2002. Vegetation classification and wildlife habitat suitability modeling in the Mackenzie Delta Region. Project #5003-01. Inuvialuit Environmental and Geotechnical Inc., Calgary, AB. Johannessen, O. M., L. Bengtsson, M. W. Miles, S. I. Kuzmina, V. A. Semenov, G. V. Alekseev, A. P. Nagurnyi, V. F. Zakharov, L. P. Bobylev, L. H. Pettersson, K. Hasselmann, and A. P. Cattle. 2004. Arctic climate change: observed and modelled temperature and sea-ice variability. Tellus Series a-Dynamic Meteorology and Oceanography 56:328-341. Johnstone, J. F., and S. V. Kokelj. In Press. Environmental conditions and vegetation recovery at abandoned-drilling mud sumps in the Mackenzie Delta region, NWT, Canada. Arctic. Kaplan, J. O., N. H. Bigelow, I. C. Prentice, S. P. Harrison, P. J. Bartlein, T. R. Christensen, W. Cramer, N. V. Matveyeva, A. D. McGuire, D. F. Murray, V. Y. Razzhivin, B. Smith, D. A. Walker, P. M. Anderson, A. A. Andreev, L. B. Brubaker, M. E. Edwards, and A. V. Lozhkin. 2003. Climate change and Arctic ecosystems: 2. Modeling, paleodata-model comparisons, and future projections. Journal of Geophysical Research-Atmospheres 108:-. Kielland, K. 1994. Amino-Acid-Absorption by Arctic Plants - Implications for Plant Nutrition and Nitrogen Cycling. Ecology 75:2373-2383. Kokelj, S. V., and C. R. Burn. 2004. Tilt of spruce trees near ice wedges, Mackenzie Delta, northwest territories, Canada. Arctic Antarctic and Alpine Research 36:615-623. Laliberte, A. S., A. Rango, K. M. Havstad, J. F. Paris, R. F. Beck, R. McNeely, and A. L. Gonzalez. 2004. Object-oriented image analysis for mapping shrub encroachment from 1937 to 2003 in southern New Mexico. Remote Sensing of Environment 93:198-210. Lantz, T. C., and S. V. Kokelj. 2008. Increasing rates of retrogressive thaw slump activity in the Mackenzie Delta region, N.W.T., Canada. Geophysical Research Letters 35:L06502, doi:06510.01029/02007GL032433. Mackay, J. R. 1963. The Mackenzie Delta Area, N.W.T. Department of Mines and Technical Surveys, Ottawa, ON. Mackay, J. R., and C. R. Burn. 2002. The first 20 years (1978-1979 to 1998-1999) of active-layer development, Illisarvik experimental drained lake site, western Arctic coast, Canada. Canadian Journal of Earth Sciences 39:1657-1674.  137  McGuire, A. D., F. S. Chapin, J. E. Walsh, and C. Wirth. 2006. Integrated regional changes in arctic climate feedbacks: Implications for the global climate system. Annual Review of Environment and Resources 31:61-91. McGuire, A. D., C. Wirth, M. Apps, J. Beringer, J. Clein, H. Epstein, D. W. Kicklighter, J. Bhatti, F. S. Chapin, B. de Groot, D. Efremov, W. Eugster, M. Fukuda, T. Gower, L. Hinzman, B. Huntley, G. J. Jia, E. Kasischke, J. Melillo, V. Romanovsky, A. Shvidenko, E. Vaganov, and D. Walker. 2002. Environmental variation, vegetation distribution, carbon dynamics and water/energy exchange at high latitudes. Journal of Vegetation Science 13:301-314. McKane, R. B., L. C. Johnson, G. R. Shaver, K. J. Nadelhoffer, E. B. Rastetter, B. Fry, A. E. Giblin, K. Kielland, B. L. Kwiatkowski, J. A. Laundre, and G. Murray. 2002. Resource-based niches provide a basis for plant species diversity and dominance in arctic tundra. Nature 415:68-71. Millington, J. D. A., G. L. W. Perry, and R. Romero-Calcerrada. 2007. Regression techniques for examining land use/cover change: A case study of a mediterranean landscape. Ecosystems 10:562-578. Muller, S. V., A. E. Racoviteanu, and D. A. Walker. 1999. Landsat MSS-derived landcover map of northern Alaska: extrapolation methods and a comparison with photo-interpreted and AVHRR-derived maps. International Journal of Remote Sensing 20:2921-2946. Nelson, F. E., N. I. Shiklomanov, G. R. Mueller, K. M. Hinkel, D. A. Walker, and J. G. Bockheim. 1997. Estimating active-layer thickness over a large region: Kuparuk River Basin, Alaska, USA. Arctic and Alpine Research 29:367-378. Notaro, M., S. Vavrus, and Z. Y. Liu. 2007. Global vegetation and climate change due to future increases in CO2 as projected by a fully coupled model with dynamic vegetation. Journal of Climate 20:70-90. Oechel, W. C., G. L. Vourlitis, J. Verfaillie, T. Crawford, S. Brooks, E. Dumas, A. Hope, D. Stow, B. Boynton, V. Nosov, and R. Zulueta. 2000. A scaling approach for quantifying the net CO2 flux of the Kuparuk River Basin, Alaska. Global Change Biology 6:160-173. Ostendorf, B., and J. F. Reynolds. 1998. A model of arctic tundra vegetation derived from topographic gradients. Landscape Ecology 13:187-201. Palmer, M. J. 2007. Vegetation, snow cover and air and near-surface ground temperatures across treeline in the uplands east of the Mackenzie Delta, Northwest Territories. Carleton University, Ottawa, ON. Parsons, A. N., J. M. Welker, P. A. Wookey, M. C. Press, T. V. Callaghan, and J. A. Lee. 1994. Growth-Responses of 4 Sub-Arctic Dwarf Shrubs to Simulated Environmental-Change. Journal of Ecology 82:307-318. PCI Geomatics. 2001. OrthoEngine Reference Manual. Version 8.2. PCI Geomatics, Richmond Hill, ON. Pelletier, B. R. 2008. Environmental Atlas of the Beaufort Coastlands. Geological Survey of Canada. http://gsc.nrcan.gc.ca/beaufort/index_e.php. Pomeroy, J. W., D. S. Bewley, R. L. H. Essery, N. R. Hedstrom, T. Link, R. J. Granger, J. E. Sicart, C. R. Ellis, and J. R. Janowicz. 2006. Shrub tundra snowmelt. Hydrological Processes 20:923-941.  138  Pomeroy, J. W., P. Marsh, H. G. Jones, and T. D. Davies. 1995. Spatial distribution of snow chemical load at the tundra-taiga transition. Pages 191-203 in K. A. Tonnessen, M. W. Williams, and M. Tranter, editors. Biogeochemistry of Seasonally Snow-covered Catchments. International Association of Hydrological Sciences No. 228. IAHS Press, Wallingford, UK. Quattrochi, D. A., and M. F. Goodchild. 1997. Scale in remote sensing and GIS. Lewis Publishers, Boca Raton, Fla. Quinn, G. P., and M. J. Keough. 2002. Experimental design and data analysis for biologists. Cambridge University Press, Cambridge, UK ; New York. R Development Core Team. 2006. R: a language and environment for statistical computing, reference index, version 2.6.2. R Foundation for Statistical Computing. http://www.R-project.org, Vienna, Austria. Reeburgh, W. S., J. Y. King, S. K. Regli, G. W. Kling, N. A. Auerbach, and D. A. Walker. 1998. A CH4 emission estimate for the Kuparuk River basin, Alaska. Journal of Geophysical Research-Atmospheres 103:29005-29013. Rhoades, C., H. Oskarsson, D. Binkley, and R. Stottlemyer. 2001. Alder (Alnus crispa) effects on soils vary with ecosystem type along the Agashashok River, northwest Alaska. Ecoscience 8:89-95. Ritchie, J. C. 1984. Past and present vegetation of the far northwest of Canada. University of Toronto Press, Toronto, ON. SAS. 2004. SAS Version 9.1.3. Help and Documentation.. SAS Institute, Cary, NC. Schimel, J. P., C. Bilbrough, and J. A. Welker. 2004. Increased snow depth affects microbial activity and nitrogen mineralization in two Arctic tundra communities. Soil Biology & Biochemistry 36:217-227. Serreze, M. C., J. E. Walsh, F. S. Chapin, T. Osterkamp, M. Dyurgerov, V. Romanovsky, W. C. Oechel, J. Morison, T. Zhang, and R. G. Barry. 2000. Observational evidence of recent change in the northern high-latitude environment. Climatic Change 46:159-207. Silapaswan, C. S., D. L. Verbyla, and A. D. McGuire. 2001. Land cover change on the Seward Peninsula: The use of remote sensing to evaluate the potential influences of climate warming on historical vegetation dynamics. Canadian Journal of Remote Sensing 27:542-554. Soil Landscapes of Canada Working Group. 2007. Soil Landscapes of Canada v3.1.1 (digital map and database at 1:1 million scale). in. Agriculture and Agri-Food Canada. Stafford, J. M., G. Wendler, and J. Curtis. 2000. Temperature and precipitation of Alaska: 50 year trend analysis. Theoretical and Applied Climatology 67:33-44. Stow, D. A., A. Hope, D. McGuire, D. Verbyla, J. Gamon, F. Huemmrich, S. Houston, C. Racine, M. Sturm, K. Tape, L. Hinzman, K. Yoshikawa, C. Tweedie, B. Noyle, C. Silapaswan, D. Douglas, B. Griffith, G. Jia, H. Epstein, D. Walker, S. Daeschner, A. Petersen, L. M. Zhou, and R. Myneni. 2004. Remote sensing of vegetation and land-cover change in Arctic Tundra Ecosystems. Remote Sensing of Environment 89:281-308. Sturm, M., J. P. McFadden, G. E. Liston, F. S. Chapin, C. H. Racine, and J. Holmgren. 2001a. Snow-shrub interactions in Arctic tundra: A hypothesis with climatic implications. Journal of Climate 14:336-344. 139  Sturm, M., C. Racine, and K. Tape. 2001b. Climate change - Increasing shrub abundance in the Arctic. Nature 411:546-547. Sturm, M., J. Schimel, G. Michaelson, J. M. Welker, S. F. Oberbauer, G. E. Liston, J. Fahnestock, and V. E. Romanovsky. 2005. Winter biological processes could help convert arctic tundra to shrubland. Bioscience 55:17-26. Tape, K., M. Sturm, and C. Racine. 2006. The evidence for shrub expansion in Northern Alaska and the Pan-Arctic. Global Change Biology 12:686-702. Tape, K. D. 2004. Shrub expansion in Arctic Alaska: 50 years of change documented using aerial photography. University of Alaska Fairbanks, Fairbanks, AK. Thompson, C., J. Beringer, F. S. Chapin, and A. D. McGuire. 2004. Structural complexity and land-surface energy exchange along a gradient from arctic tundra to boreal forest. Journal of Vegetation Science 15:397-406. Timoney, K. P., G. H. Laroi, S. C. Zoltai, and A. L. Robinson. 1993. Vegetation Communities and Plant-Distributions and Their Relationships with Parent Materials in the Forest-Tundra of Northwestern Canada. Ecography 16:174-188. Walker, D. A. 2000. Hierarchical subdivision of Arctic tundra based on vegetation response to climate, parent material and topography. Global Change Biology 6:19-34. Walker, D. A., W. A. Gould, H. A. Maier, and M. K. Raynolds. 2002. The Circumpolar Arctic Vegetation Map: AVHRR-derived base maps, environmental controls, and integrated mapping procedures. International Journal of Remote Sensing 23:45514570. Walker, M. D., C. H. Wahren, R. D. Hollister, G. H. R. Henry, L. E. Ahlquist, J. M. Alatalo, M. S. Bret-Harte, M. P. Calef, T. V. Callaghan, A. B. Carroll, H. E. Epstein, I. S. Jonsdottir, J. A. Klein, B. Magnusson, U. Molau, S. F. Oberbauer, S. P. Rewa, C. H. Robinson, G. R. Shaver, K. N. Suding, C. C. Thompson, A. Tolvanen, O. Totland, P. L. Turner, C. E. Tweedie, P. J. Webber, and P. A. Wookey. 2006. Plant community responses to experimental warming across the tundra biome. Proceedings of the National Academy of Sciences of the United States of America 103:1342-1346. Walker, M. D., D. A. Walker, and N. A. Auerbach. 1994. Plant-Communities of a Tussock Tundra Landscape in the Brooks Range Foothills, Alaska. Journal of Vegetation Science 5:843-866. Wilson, J. P., and J. C. Gallant. 2000. Terrain analysis : principles and applications. Wiley, New York. Wulder, M. A., J. C. White, J. E. Luther, G. Strickland, T. K. Remmel, and S. W. Mitchell. 2006. Use of vector polygons for the accuracy assessment of pixelbased land cover maps. Canadian Journal of Remote Sensing 32:268-279.  140  Chapter 6. Conclusions Over the last several decades, Arctic plant ecologists have focused primarily on understanding the effects of climate warming on tundra vegetation (Arft et al. 1999, Shaver et al. 2000, Walker et al. 2006, Figure 6.1A). This emphasis has been motivated by predictions that temperature changes will be more rapid at high latitudes than over the rest of the globe (Serreze et al. 2000, Hassol 2004) and the strong influence of northern vegetation on the global climate system (Chapin et al. 2000b, Thompson et al. 2004, McGuire et al. 2006). Two decades of plot-level experimental research in these ecosystems has shown that Arctic vegetation is extremely sensitive to the direct and indirect effects of climate change (Chapin et al. 1995, Shaver et al. 2000). Recent observations have also shown that temperature increases in the Arctic have, indeed, been more than double global changes (Johannessen et al. 2004, Parry and Intergovernmental Panel on Climate Change. Working Group II. 2007). As Arctic ecosystems warm, landscape level changes are becoming increasingly evident (Lloyd and Fastie 2002, Stow et al. 2004, Tape et al. 2006). In addition to affecting vegetation, a warming climate is also likely to alter the frequency and intensity of natural disturbance at high latitudes (Jorgenson et al. 2001, Kasischke and Turetsky 2006). Growing industrial activity will also increase the footprint of anthropogenic disturbance in the Arctic (Walker 1996, Forbes et al. 2001, Holroyd and Retzer 2005l, Figure 6.1B). Disturbances can increase growth, productivity, and reproduction relative to undisturbed sites (Hernandez 1973, Truett and Kertell 1992, Vavrek et al. 1999, de Groot and Wein 2004, Racine et al. 2004) and the functional recovery of Low Arctic ecosystems from disturbance can be very rapid. However, since complete recovery from disturbance and a return to antecedent conditions may occur on centennial or millennial time scales (Svoboda and Henry 1987, Walker et al. 1987, Walker et al. 1994, Figure 6.1C) long-term successional trajectories are poorly understood. To fully understand the changes that are taking place in the Arctic, longterm monitoring is required. Individually, climate change and increasing disturbance are both likely to significantly alter northern ecosystems (Figure 6.1). However, the interactions between 141  these two factors are not well understood. Since changes in the structure of northern vegetation will affect ecosystem function, human livelihoods, as well as the global climate system (Hassol 2004, Chapin et al. 2005, McGuire et al. 2006), disentangling the effects and interactions of temperature and disturbance is a critical research need. Plotlevel experiments in the Low Arctic highlight the importance of indirect effects such as resource availability on vegetation, but responses have also varied through time and among ecosystems (Chapin et al. 1995, Chapin and Shaver 1996, Arft et al. 1999, Shaver et al. 2000, Walker et al. 2006). Thus, as the impacts of temperature and disturbance are increasingly witnessed at the landscape scale, research that links plot-level trends and predictions with regional variability also becomes essential (Shaver et al. 2000, Arctic LTER 2004, Dunne et al. 2004, Hollister et al. 2005).  Figure 6.1. Generalized effects and feedbacks of climate and disturbance on tundra ecosystems. (A) Direct and indirect effects of temperature on vegetation and abiotic conditions will feedback to regional climate. (B) Increasing temperatures will alter the frequency of natural disturbance. (C) Disturbance will alter tundra ecosystems, which may lead to additional changes in the frequency of natural disturbance. There are a number of long-term monitoring programs in the North American Arctic (Henry and Molau 1997, Arctic LTER 2007, ArcticNet 2008, ITEX 2008) and several broad scale transects studies (McGuire et al. 2002, Walker et al. 2003), but there are a number of gaps throughout the circumarctic that demand additional case studies. My dissertation research contributes to our understanding of the importance of disturbances in Low Arctic, by nesting plot-level observations within a regional context, and evaluating the impact of disturbance after comparatively long time periods (50-80 142  years). My major findings (summarized next) highlight the long-term effects (>100 years) of climate and disturbance, as well as the influence of disturbance patch dynamics on the tundra landscape, as areas of key uncertainty requiring further investigation. SUMMARY My results provide additional evidence that both a warming climate and increased disturbance will have important impacts on northern vegetation. The relative roles of temperature and fire on forest-tundra and tundra vegetation were assessed by sampling green alder (Alnus viridis subsp. fruticosa) abundance, growth, and reproduction and plant community composition on burned and unburned sites across a natural temperature gradient. Increases in green alder seed viability and abundance with higher regional temperatures, as well as evidence of recent alder recruitment, suggest that warmer temperatures have, and will likely continue, to facilitate alder expansion by increasing the availability of viable seed (Chapter 2). The magnitude of fire effects on alder growth, abundance and reproduction indicate that increased fire frequency may have a larger effect on alder spread than temperature (Chapter 2). This conclusion is supported by the results presented in Chapter 3, where I examined the relative importance of regional temperature and disturbance in the Low Arctic by comparing microenvironmental variables, plant community composition, and the response of green alder on retrogressive thaw slumps and undisturbed terrain. Increases in productivity, catkin production, seed viability, height and density of green alder were evident at thaw slumps, as compared to controls. The magnitude of these differences indicates that disturbances exert a stronger influence on green alder than differences in temperature (Chapter 3). Differences in plant community composition on disturbed sites (up to 80 years old) show that the cumulative effects of disturbance are non-trivial (Chapters 2, 3). Thaw slumps had increased nutrient availability, soil pH, depth of snow pack, ground temperatures, and thaw depth, compared with undisturbed sites. These abiotic differences are important drivers of plant community structure, which also differed on thaw slumps versus undisturbed terrain (Chapter 3). Long-term differences in plant community composition were also seen on burned sites at southern limit of the Low 143  Arctic (Chapter 2). At the plot level, disturbance resulted in greater changes in alder population ecology, microclimate, and community composition than did differences across a regional temperature gradient (Chapters 2, 3). The abundance of deciduous shrubs and the presence of species with boreal affinities on old disturbances raises the possibility that disturbances may act as ‘stepping stones’ for the northward movement of species, thereby facilitating greater change at the landscape-level (Chapters 2, 3). Conversely, age distributions of alder stems on undisturbed sites in the northern transition zone indicate that recent temperature increases have already impacted alder abundance by increasing seed viability and recruitment (Chapter 2). Thus, it is unclear if the patchlevel effects of disturbance will override the broad scale effects of temperature on recruitment across the tundra landscape. To examine the importance of temperature at the landscape scale, I mapped finescale changes in the abundance and patch size of shrub tundra and dwarf shrub tundra in undisturbed terrain using aerial photos and object-based classification techniques. I explored the relative influence of biophysical variables including regional temperature, soil type and microtopography in determining the position of this transition. In the Mackenzie Delta uplands, vegetation changes from shrub tundra to dwarf shrub tundra over a distance of approximately 80 km (Chapter 5). Mean temperature makes the largest independent contribution to models of shrub tundra abundance, and declines below 50% shrub tundra cover are coincident with the 10°C mean July isotherm. This suggests that at broad scales temperature is the primary determinant of the location of this transition and that continued warming in the Arctic will likely cause shifts the in position of this ecotone (Chapter 5). Clearly temperature and disturbance both have an important influence on the vegetation of the Low Arctic. However, assessing their relative importance is complicated by differences in the spatial extent of their influence and simultaneous changes in both temperature and the frequency of disturbance. To examine recent changes in temperature and disturbance I analyzed historical climate records and mapped rates of thaw slumping since 1950 using aerial photographs. Retrogressive thaw slump activity from 1973-2004 was significantly greater than the preceding period (1950-1973). These changes have likely occurred in response to a significant warming of annual and 144  summer air temperatures over the same time period (Chapter 4). These results are also consistent with other evidence that both temperature and the frequency of natural disturbance are increasing across the north (Jorgenson et al. 2001, Kasischke and Turetsky 2006). As the frequency of disturbance increases with future warming, the effects of disturbance will likely magnify the direct effects of temperature on Low Arctic ecosystems.  UNCERTAINTY AND RESEARCH NEEDS Spatial Dynamics in a Warming World: Linking Plots and Landscapes By removing existing plant communities, exposing mineral substrates, and increasing nutrient availability and ground temperatures, severe disturbances (Figure 6.1C) clearly have much larger plot-level effects on Arctic vegetation than do differences in temperature (Chapters 2, 3, Figure 6.1A). However, at the landscape scale our understanding of the relative importance of disturbance (Figure 6.1C) and temperature (Figure 6.1A) remains imprecise. Will the large, but localized, impacts of disturbance affect the surrounding tundra, potentially overriding the impact of temperature at the landscape scale (Chapter 5)? The effect of disturbance on seed production and viability suggest that changes in the frequency of disturbance could facilitate landscape level change by increasing the density of seed sources. By acting as sources of viable seed and increasing the likelihood of recruitment, disturbances may provide ‘stepping stones’ for the proliferation of novel species and assemblages, which although better adapted to current climate, are otherwise inhibited in undisturbed tundra. Disturbances in the Low Arctic are small, but widespread (Figure 6.2), and they are becoming more frequent. Consequently, changing disturbance regimes could facilitate rapid shifts in vegetation by linking the northern boreal forest with the Low Arctic. Since the size and spatial arrangement of these ‘stepping stones’ is likely to influence the spread of species into disturbed and undisturbed terrain, there is a need model patch dynamics in a spatially explicitly manner.  145  As natural and anthropogenic disturbances increase in the Low Arctic, it will also be important to determine if their influence is directly proportional to their spatial extent. Observed changes in fire frequency and thaw slump activity highlight the possibility that disturbance may have already influenced patterns of recruitment in undisturbed tundra by increasing the abundance of viable seed. Similarly, temperature changes described in Chapter 4 also make it unlikely that the effects observed on burns and thaw slumps can be attributed solely to disturbance. Sensitivity analysis of a model of shrub tundra spread that includes disturbance and temperature effects on seed viability and recruitment would provide a framework to examine the relative influence of disturbance on the tundra landscape. Research using historical airphotos to map changes in shrub abundance in disturbed and undisturbed tundra will also help to disentangle these factors and is underway. This work will be combined with efforts to document fine-scale microenvironmental differences on disturbed and undisturbed sites where shrub tundra has proliferated. Although it is difficult to replicate the scale of even small natural disturbances, manipulations to mimic disturbance, or experimental warming of homogeneous areas of disturbed terrain could also make an important contribution to efforts to tease apart the influence of temperature and disturbance. Since, increases in disturbance may mediate both plot and landscape level changes in vegetation, future research should be carried out at multiple spatial scales. Observational, experimental, and modeling research conducted in the Kuparuk River Basin, AK within a nested hierarchy of spatial scales (Nelson et al. 1997, Reeburgh et al. 1998, Oechel et al. 2000, Arctic Geobotanical Atlas 2007) provides a good model for future efforts to integrate research conducted across various spatial and temporal scales in the Mackenzie Delta Region. Long-term Cumulative Impacts Although increases in disturbance affect a considerably smaller area than changes in regional temperature, disturbance effects are unlikely to be trivial. The persistence of biotic and abiotic differences following disturbance, for close to a century, implies that impact of numerous small disturbances will accumulate through time. Despite the size of the Mackenzie Delta uplands, it is difficult to locate areas that do not bear visible 146  evidence of disturbance (Figure 6.2). The growing frequency of disturbance across the Low Arctic make understanding the duration of their effects essential in evaluating cumulative ecological impacts.  Figure 6.2. Map showing the distribution of the major disturbances in the study area. Thaw slumps shown were mapped within the area bounded by the dashed line were mapped by Lantz and Kokelj (2008). Lakes drained since 1950 are based on mapping by Marsh (2005). Drilling mud sumps were mapped by Kokelj and 147  GeoNorth Limited (2002). Fires are based on Wein (1975) and data on historical seismic was obtained from WWF (2002).  My research explored the impacts of disturbance after comparatively long time periods. However, it should be stressed that the temporal scales considered here may only represent a small proportion of the time that may be required for complete ecosystem recovery following Arctic disturbance (Cargill and Chapin 1987, Svoboda and Henry 1987, Walker and Walker 1991). Large differences in species composition and environment on disturbed sites after long time periods raise the possibility that postdisturbance plant communities developing in a modern climate are producing novel successional trajectories (Forbes et al. 2001, Williams and Jackson 2007). Alternatively, it is also possible that secondary succession in the Low Arctic simply occurs on very long time scales (Peterson and Billings 1980, Cargill and Chapin 1987, Svoboda and Henry 1987). Palynological re-construction of Quaternary plant communities in the western Arctic shows that past changes in vegetation have occurred very rapidly (Ritchie 1984, Anderson and Brubaker 1993) and as Shaver (2000) has noted “we should expect some surprises”. The possibility that the greater flammability of shrub tundra will increase the frequency of tundra fires (Higuera et al. 2008), makes understanding long-term secondary succession (>100 years) particularly critical. Since different successional trajectories (e.g. dense shrub tundra, spruce woodland, grassland-steppe) (Landhäusser and Wein 1993, Rupp et al. 2000, Racine et al. 2004, Chapter 2) will each have very different feedbacks to climate, as well as impacts on wildlife and human communities, understanding secondary succession in the Low Arctic represents one of the most important challenges faced by Arctic plant ecologists. Thus, additional case studies examining historical disturbances are also necessary. Ultimately, long-term monitoring and re-sampling of historical disturbances and modeling will all be required to improve our ability to predict the future state of northern vegetation.  148  FUTURE CONSEQEUNCES OF A WARMING WORLD In the coming decades temperature increases are likely to continue to facilitate the proliferation of shrub tundra (Chapters 2, 3, 4, Figure 6.1A). Secondary succession on disturbed sites will also probably drive rapid changes in the community composition in forest-tundra and tundra ecosystems (Chapters 2, 3, Figure 6.1C). In both cases the resultant increase in the abundance of deciduous shrubs is likely to amplify regional heating by decreasing albedo (Chapin et al. 2000a, Epstein et al. 2004, Sturm et al. 2005) Increased shrubbiness will also drive positive feedbacks that: alter snow pack, increase active layer depth and nutrient availability, and modify wildlife habitat (Walsh et al. 1997, Rhoades et al. 2001, Schimel et al. 2004, Sturm et al. 2005, Grogan and Jonasson 2006, Johnstone and Kokelj In Press, Figure 6.1A). Over longer time periods there is much greater uncertainty about the consequences of climate change and the altered frequency of disturbance. It is possible that the increased abundance of deciduous shrubs may increase tundra flammability (Higuera et al. 2008, Figure 6.1C). Changes in the area affected by fire and thus, secondary succession, could greatly alter: forest and soil carbon stocks, vegetation albedo, and the extent of discontinuous and continuous permafrost (Figure 6.1C). Changes in Arctic ecosystems driven by increased global temperatures and an increasing frequency of natural and anthropogenic disturbance will also significantly alter the availability and abundance of traditional foods in indigenous communities across Canada. Ultimately, research to understand the future of northern vegetation can also contribute to ongoing efforts in indigenous communities to employ traditional ecological knowledge to develop adaptive strategies to respond to these changes.  149  LITERATURE CITED Arctic Geobotanical Atlas. 2007. Kuparuk River Region/Toolik Lake (Toolik Atlas). Alaska Geobotany Center, http://www.arcticatlas.org/atlas/htla/. Arctic LTER. 2004. The Arctic LTER: Regional Variation In Ecosystem Processes And Landscape Linkages. NSF Proposal for renewal of the Arctic LTER. Available: http://ecosystems.mbl.edu/ARC/otherdocs/proposal/2004_ARC_LTER_Proposal. pdf. Arctic LTER. 2007. The Arctic LTER Project at Toolik Lake, Alaska: NSF site review. Available: http://ecosystems.mbl.edu/ARC/otherdocs/ARC%20site%20review%202007.pdf. ArcticNet. 2008. Arctic Net Website. Accessed April, 2008. www.arcticnet.ulaval.ca. Chapin, F. S., M. Berman, T. V. Callaghan, P. Convey, A.-S. Crépin, K. Danell, H. W. Ducklow, B. Forbes, G. Kofinas, A. D. McGuire, M. Nuttall, R. Virginia, O. Young, S. E. Zimov, T. Christensen, A. Godduhn, E. J. Murphy, D. Wall, C. Zockler, and B. Fitzharris. 2005. Polar systems. Pages 717-743 in R. Hassan, R. Scholes, and N. Ash, editors. Millennium Ecosystem Assessment: Ecosystems and human well-being: current state and trends, Volume 1. Island Press, Washington, D.C. Chapin, F. S., W. Eugster, J. P. McFadden, A. H. Lynch, and D. A. Walker. 2000a. Summer differences among Arctic ecosystems in regional climate forcing. Journal of Climate 13:2002-2010. Chapin, F. S., A. D. McGuire, J. Randerson, R. Pielke, D. Baldocchi, S. E. Hobbie, N. Roulet, W. Eugster, E. Kasischke, E. B. Rastetter, S. A. Zimov, and S. W. Running. 2000b. Arctic and boreal ecosystems of western North America as components of the climate system. Global Change Biology 6:211-223. Chapin, F. S., G. R. Shaver, A. E. Giblin, K. J. Nadelhoffer, and J. A. Laundre. 1995. Responses of Arctic Tundra to Experimental and Observed Changes in Climate. Ecology 76:694-711. de Groot, W. J., and R. Wein. 2004. Effects of fire severity and season of burn on Betula glandulosa growth dynamics. International Journal of Wildland Fire 13:287-295. Dunne, J. A., S. R. Saleska, M. L. Fischer, and J. Harte. 2004. Integrating experimental and gradient methods in ecological climate change research. Ecology 85:904-916. Epstein, H. E., J. Beringer, W. A. Gould, A. H. Lloyd, C. D. Thompson, F. S. Chapin, G. J. Michaelson, C. L. Ping, T. S. Rupp, and D. A. Walker. 2004. The nature of spatial transitions in the Arctic. Journal of Biogeography 31:1917-1933. Forbes, B. C., J. J. Ebersole, and B. Strandberg. 2001. Anthropogenic disturbance and patch dynamics in circumpolar arctic ecosystems. Conservation Biology 15:954969. Grogan, P., and S. Jonasson. 2006. Ecosystem CO2 production during winter in a Swedish subarctic region: the relative importance of climate and vegetation type. Global Change Biology 12:1479-1495. Hassol, S. J. 2004. Arctic Climate Impact Assessment: Impacts of Warming Climate. Cambridge University Press, Cambridge. Henry, G. H. R., and U. Molau. 1997. Tundra plants and climate change: the International Tundra Experiment (ITEX). Global Change Biology 3(Suppl. 1):1-9. 150  Hernandez, H. 1973. Natural Plant Recolonization of Surficial Disturbances, Tuktoyaktuk Peninsula Region, Northwest Territories. Canadian Journal of Botany-Revue Canadienne De Botanique 51:2177-2196. Higuera, P. E., L. B. Brubaker, P. M. Anderson, T. A. Brown, A. T. Kennedy, and F. S. Hu. 2008. Frequent Fires in Ancient Shrub Tundra: Implications of Paleo-records for Arctic Environmental Change. PLoS ONE 3: e0001744. . Hollister, R. D., P. J. Webber, and C. Bay. 2005. Plant response to temperature in Northern Alaska: Implications for predicting vegetation change. Ecology 86:1562-1570. Holroyd, P., and H. Retzer. 2005. A Peak into the Future: The Potential Landscape Impacts from Gas Development in Northern Canada. The Pembina Institute for Appropriate Development, Calgary, AB. ITEX. 2008. International Tundra Experiment Website. Accessed April, 2008. www.geog.ubc.ca/itex. Johannessen, O. M., L. Bengtsson, M. W. Miles, S. I. Kuzmina, V. A. Semenov, G. V. Alekseev, A. P. Nagurnyi, V. F. Zakharov, L. P. Bobylev, L. H. Pettersson, K. Hasselmann, and A. P. Cattle. 2004. Arctic climate change: observed and modelled temperature and sea-ice variability. Tellus Series a-Dynamic Meteorology and Oceanography 56:328-341. Johnstone, J. F., and S. V. Kokelj. In Press. Environmental conditions and vegetation recovery at abandoned-drilling mud sumps in the Mackenzie Delta region, NWT, Canada. Arctic. Jorgenson, M. T., C. H. Racine, J. C. Walters, and T. E. Osterkamp. 2001. Permafrost degradation and ecological changes associated with a warming climate in central Alaska. Climatic Change 48:551-579. Kasischke, E. S., and M. R. Turetsky. 2006. Recent changes in the fire regime across the North American boreal region - Spatial and temporal patterns of burning across Canada and Alaska. Geophysical Research Letters 33:-. Kokelj, S. V., and GeoNorth Limited. 2002. Drilling mud sumps in the Mackenzie Delta Region. Construction, abandonment and past performance. Report submitted to the Department of Indian Affairs and Northern Development, NWT Region. Lantz, T. C., and S. V. Kokelj. 2008. Increasing rates of retrogressive thaw slump activity in the Mackenzie Delta region, N.W.T., Canada. Geophysical Research Letters 35:L06502, doi:06510.01029/02007GL032433. Lloyd, A. H., and C. L. Fastie. 2002. Spatial and temporal variability in the growth and climate response of treeline trees in Alaska. Climatic Change 52:481-509. Marsh, P., M. Russell, and C. Onclin. 2005. The hydrology of lakes in the Western Canadian Arctic: Implications to proposed natural gas development. Pages 131140 in 15th International Northern Research Basins Symposium and Workshop Water Resources Engineering, Lund University, Luleå to Kvikkjokk, Sweden,. McGuire, A. D., F. S. Chapin, J. E. Walsh, and C. Wirth. 2006. Integrated regional changes in arctic climate feedbacks: Implications for the global climate system. Annual Review of Environment and Resources 31:61-91. McGuire, A. D., C. Wirth, M. Apps, J. Beringer, J. Clein, H. Epstein, D. W. Kicklighter, J. Bhatti, F. S. Chapin, B. de Groot, D. Efremov, W. Eugster, M. Fukuda, T. Gower, L. Hinzman, B. Huntley, G. J. Jia, E. Kasischke, J. Melillo, V. 151  Romanovsky, A. Shvidenko, E. Vaganov, and D. Walker. 2002. Environmental variation, vegetation distribution, carbon dynamics and water/energy exchange at high latitudes. Journal of Vegetation Science 13:301-314. Nelson, F. E., N. I. Shiklomanov, G. R. Mueller, K. M. Hinkel, D. A. Walker, and J. G. Bockheim. 1997. Estimating active-layer thickness over a large region: Kuparuk River Basin, Alaska, USA. Arctic and Alpine Research 29:367-378. Oechel, W. C., G. L. Vourlitis, J. Verfaillie, T. Crawford, S. Brooks, E. Dumas, A. Hope, D. Stow, B. Boynton, V. Nosov, and R. Zulueta. 2000. A scaling approach for quantifying the net CO2 flux of the Kuparuk River Basin, Alaska. Global Change Biology 6:160-173. Parry, M. L., and Intergovernmental Panel on Climate Change. Working Group II. 2007. Climate Change 2007 : impacts, adaptation and vulnerability : contribution of Working Group II to the fourth assessment report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, U.K. ; New York. Racine, C., R. Jandt, C. Meyers, and J. Dennis. 2004. Tundra fire and vegetation change along a hillslope on the Seward Peninsula, Alaska, USA. Arctic Antarctic and Alpine Research 36:1-10. Reeburgh, W. S., J. Y. King, S. K. Regli, G. W. Kling, N. A. Auerbach, and D. A. Walker. 1998. A CH4 emission estimate for the Kuparuk River basin, Alaska. Journal of Geophysical Research-Atmospheres 103:29005-29013. Rhoades, C., H. Oskarsson, D. Binkley, and R. Stottlemyer. 2001. Alder (Alnus crispa) effects on soils vary with ecosystem type along the Agashashok River, northwest Alaska. Ecoscience 8:89-95. Schimel, J. P., C. Bilbrough, and J. A. Welker. 2004. Increased snow depth affects microbial activity and nitrogen mineralization in two Arctic tundra communities. Soil Biology & Biochemistry 36:217-227. Shaver, G. R., J. Canadell, F. S. Chapin, J. Gurevitch, J. Harte, G. Henry, P. Ineson, S. Jonasson, J. Melillo, L. Pitelka, and L. Rustad. 2000. Global warming and terrestrial ecosystems: A conceptual framework for analysis. Bioscience 50:871882. Stow, D. A., A. Hope, D. McGuire, D. Verbyla, J. Gamon, F. Huemmrich, S. Houston, C. Racine, M. Sturm, K. Tape, L. Hinzman, K. Yoshikawa, C. Tweedie, B. Noyle, C. Silapaswan, D. Douglas, B. Griffith, G. Jia, H. Epstein, D. Walker, S. Daeschner, A. Petersen, L. M. Zhou, and R. Myneni. 2004. Remote sensing of vegetation and land-cover change in Arctic Tundra Ecosystems. Remote Sensing of Environment 89:281-308. Sturm, M., J. Schimel, G. Michaelson, J. M. Welker, S. F. Oberbauer, G. E. Liston, J. Fahnestock, and V. E. Romanovsky. 2005. Winter biological processes could help convert arctic tundra to shrubland. Bioscience 55:17-26. Svoboda, J., and G. H. R. Henry. 1987. Succession in Marginal Arctic Environments. Arctic and Alpine Research 19:373-384. Tape, K., M. Sturm, and C. Racine. 2006. The evidence for shrub expansion in Northern Alaska and the Pan-Arctic. Global Change Biology 12:686-702. Thompson, C., J. Beringer, F. S. Chapin, and A. D. McGuire. 2004. Structural complexity and land-surface energy exchange along a gradient from arctic tundra to boreal forest. Journal of Vegetation Science 15:397-406. 152  Truett, J. C., and K. Kertell. 1992. Tundra disturbance and ecosystem production: implications for impact assessment. Environmental Management 16:485-494. Vavrek, M. C., N. Fetcher, J. B. McGraw, G. R. Shaver, F. S. Chapin, and B. Bovard. 1999. Recovery of productivity and species diversity in Tussock tundra following disturbance. Arctic Antarctic and Alpine Research 31:254-258. Walker, D. A. 1996. Disturbance and recovery of Arctic Alaskan vegetation. Pages 35-71 in J. F. Raynolds and J. D. Tenhunen, editors. Landscape function and disturbance in Arctic tundra. Springer-Verlag, Berlin. Walker, D. A., D. Cate, J. Brown, and C. Racine. 1987. Disturbnace and recovery of arctic Alaskan tundra terrain: a review of recent investigations. CRREL Report 87-11. US Army Corps of Engineers. Cold Regions Research and Engineering Laboratory, Hanover, NH. Walker, D. A., H. E. Epstein, G. J. Jia, A. Balser, C. Copass, E. J. Edwards, W. A. Gould, J. Hollingsworth, J. Knudson, H. A. Maier, A. Moody, and M. K. Raynolds. 2003. Phytomass, LAI, and NDVI in northern Alaska: Relationships to summer warmth, soil pH, plant functional types, and extrapolation to the circumpolar Arctic. Journal of Geophysical Research-Atmospheres 108:-. Walker, M. D., D. A. Walker, and N. A. Auerbach. 1994. Plant-Communities of a Tussock Tundra Landscape in the Brooks Range Foothills, Alaska. Journal of Vegetation Science 5:843-866. Walsh, N. E., T. R. McCabe, J. M. Welker, and A. N. Parsons. 1997. Experimental manipulations of snow-depth: effects on nutrient content of caribou forage. Global Change Biology 3(Suppl. 1) 158-164. Wein, R. W. 1975. Vegetation recovery in arctic tundra and forest tundra after fire. ALUR 74-75-62, Indian and Northern Affairs Canada, Ottawa, ON. WWF. 2002. Northwest Territories Digital Atlas. World Wildlife Fund Canada, Toronto, ON.  153  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0066494/manifest

Comment

Related Items