UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Effects of simulated and actual caribou grazing on low-Arctic tundra vegetation O, Pamela Constance 2011

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

Item Metadata

Download

Media
24-ubc_2011_fall_o_pamela.pdf [ 5.72MB ]
Metadata
JSON: 24-1.0072027.json
JSON-LD: 24-1.0072027-ld.json
RDF/XML (Pretty): 24-1.0072027-rdf.xml
RDF/JSON: 24-1.0072027-rdf.json
Turtle: 24-1.0072027-turtle.txt
N-Triples: 24-1.0072027-rdf-ntriples.txt
Original Record: 24-1.0072027-source.json
Full Text
24-1.0072027-fulltext.txt
Citation
24-1.0072027.ris

Full Text

EFFECTS OF SIMULATED AND ACTUAL CARIBOU GRAZING ON LOW-ARCTIC TUNDRA VEGETATION  by  Pamela C. O  B.Sc.Hons., University of Toronto, 2000 M.Sc., University of Toronto, 2003  A THESIS SUBMITTED IN PARTIAL FULFULLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY  in  The Faculty of Graduate Studies  (Geography)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  August 2011  © Pamela C. O, 2011  Abstract Barren-ground caribou have been grazing and trampling the tundra for thousands of years. Because the timing of grazing and trampling is episodic, it has been theorized that their impacts at any given site are weak or absent. This study investigated if this could be verified observationally and experimentally. I conducted an experiment to examine the effects of simulated grazing and lichen removal on birch hummock - lichen heath tundra in the low-Arctic. I also examined the effects of trampling and grazing by the Bathurst Caribou Herd on the biomass of three low-Arctic plant communities. In general, the simulated grazing at intermediate and high intensities did not cause changes in vascular plants biomass or species diversity, or carbon dioxide flux. However, lichen removal caused significant reductions in lichen biomass, lichen diversity, and net ecosystem production. Ecosystem respiration rates and biomass were much lower on than off the caribou migratory trails in each of the habitats studied, due to the low amounts of biomass on migratory trails compared to off the trails. These studies show that the effects of grazing were not easily detected, but the migratory trails that have been used by caribou for thousands of years were distinctly different than the surrounding areas. The results indicate that some habitats may be resistant to change, but once they are altered, they may not readily recover.  ii  Table of Contents Abstract …………………………………………………………….…………..…….……...... ii Table of Contents …………………………………...……………….………….………...... iii List of Tables …………………………………………………………….………….….….... vi List of Figures …………………………...…………………………………..………........... viii Acknowledgements …………………..……………………………………………..……... xii 1 INTRODUCTION ………………………………………………………….… 1 1.1 The Changing Atmosphere …….……………………………….......… 1 1.2 Climate Change in the Arctic …………………...……………………. 2 1.3 Ecosystem Carbon Exchange …………………………………………. 3 1.4 Grazing ………………………………………………...………….…… 4 1.5 Compensatory Growth ……………………………………...………… 5 1.6 Northern Grazing Systems ……………………………………………. 6 1.7 Caribou (Rangifer tarandus) ………………………………………….. 8 1.8 Climate Change and Caribou ………………………………...……… 11 1.9 Caribou Grazing ………………………………………………...……. 13 1.9.1  Forage Sources ………………………………………………………..……. 13 1.10 Objectives ………………………………………………….……..…… 14 1.11 General Hypotheses ………………………………………………..…. 15  2 EFFECTS OF SIMULATED GRAZING ON PLANT GROWTH IN A MESIC BIRCH HUMMOCK HABITAT IN LOW-ARCTIC TUNDRA... 16 2.1 Introduction ……………………………………………….…………...16 2.2 Methods ………….…………………………………………….……….17 2.2.1 2.2.2 2.2.3 2.2.4  Study Sites ………………………………………………………..………… 17 Experimental Design ….………………………………………….…………. 18 Sampling Procedures …………………………………………..…………… 20 Statistical Analysis ………....………………………………..……………… 22 2.3 Results …………………………………………………………………. 23 2.3.1 Climate Data …………………...…………………………………...………. 23 2.3.2 Biomass ………………………………………………………………..……. 25 2.3.2.1 Vascular Plants …..………………………………………………..……..25 2.3.2.2 Lichens ……………………………...………………………………….. 28 2.3.2.3 Mosses …………………………………………………………………...29 2.3.2.4 Litter …………………………………………………………………….. 30 2.3.2.5 Leaf to Shoot Ratio ……………………………………………..………. 31 2.3.3 Point Cover ………………………………………………………...……….. 35 2.3.3.1 Point Frame Hits per Plot ……………………………………………….. 35 2.3.3.2 Shannon-Weaver Diversity Index (H) ………………………..………… 38 2.4 Discussion ……………………………………………...…...……...….. 39 2.5 Conclusions ……………………………………………...…………….. 41 iii  3 EFFECTS OF SIMULATED GRAZING ON THE ECOSYSTEM CARBON EXCHANGE OF A LOW-ARCTIC TUNDRA SITE …………43 3.1 Introduction ……………………………………………………………43 3.2 Methods …………………………………………………….…………..44 3.2.1 3.2.2 3.2.3 3.2.4 3.2.5  Study Area and Design …………...………………………………………… 44 Chamber-Based Methods of Measuring CO2 Flux …...…………….………. 46 Carbon Dioxide Flux Calculations ……………………………………….….47 Environmental Factors ………………………..…………..………………… 48 Statistical Analyses …………………………….…………...………………. 48 3.3 Results …………………………………………..………...…………… 49 3.3.1 Carbon Dioxide Fluxes ……………………………..…………...………….. 49 3.3.2 Ion Exchange Membranes ………………………………..………………….55 3.3.3 Soil Water Content and Temperature ……………………..…………...…….56 3.4 Discussion …………………………………………………………..…. 58 3.5 Conclusions …………………………………………………………..... 61  4 INFLUENCES OF CARIBOU MIGRATION TRAILS ON TUNDRA VEGETATION, SOILS AND CARBON FLUX …..………………………. 63 4.1 Introduction ……………………………………...…………………….63 4.2 Methods ……………………………………………...…………………64 4.2.1 4.2.2 4.2.3  Study Area and Design …………………………………..…………………. 64 Sampling …………………………………….……………..………………. 67 Statistical Analysis …………………………………………….……………. 67 4.3 Results ………………………………………….……………………… 68 4.3.1 Abiotic Variables …………………………………………………………… 68 4.3.2 Ecosystem Respiration ………………………..…………………………….. 70 4.3.3 Plant Community ……………………………..…………………………….. 71 4.4 Discussion ………………………………………..………………...….. 78 4.5 Conclusions ………………………………………….………………… 79  5 GENERAL CONCLUSIONS …………………………………………….………… 80 5.1 Summary ……………………………………………………………… 80 5.2 Future Directions …………………………………………………..…. 82 5.3 General Implications ……………………………………………….… 84 REFERENCES …………………………….…………….……………………… 85 APPENDICES ………………………………...………………………………… 93 Appendix A Net Ecosystem Production values by treatment for 2004, 2005, and 2006. n=30. Vertical dashed lines indicate the beginning and end of peak biomass, with early and late season data on either side …..…..……. 93 Appendix B Gross Ecosystem Production values by treatment for 2004, 2005, and 2006. n=30. Vertical dashed lines indicate the beginning and end of peak biomass, with early and late season data on either side .…………….94 iv  Appendix C Ecosystem Respiration values by treatment for 2004, 2005, and 2006. n=30. Vertical dashed lines indicate the beginning and end of peak biomass, with early and late season data on either side …………..……… 95 Appendix D Means and standard error of the means for soil moisture data. ………………………………………………………………………………... 96 Appendix E Daily mean, maximum and minimum soil temperature data for the three study years..…………………………………....………………….. 99  v  List of Tables Table 1.1  Recent census data on the barren-ground caribou herds in the Northwest Territories, Canada reported by the Circumarctic Rangifer Monitoring and Assessment Network (CARMA). Source: Figure 5.3-5 from NWT Environmental Audit Status of the Environment Report, 2005. ….………….. 9  Table 2.1  Seasonal rainfall and temperature summaries from 1996-2006a for Daring Lake, NWT. ………………………………………………………………………….... 24  Table 2.2  Results of the likelihood ratio tests with grazing and lichen treatments as fixed factors and temporal and period of time as random factors for live and dead vascular plant biomass (g m-2); n=3-5 plots per treatment. Significant p-values for treatment effects are shown in bold. ………...…….25  Table 2.3  Results of the likelihood ratio tests with grazing and lichen treatments as fixed factors and temporal and period of time as random factors for lichen, mosses biomass and litter standing crop (g m-2); n=3-5 plots per treatment. Significant p-values for treatment effects are shown in bold. …….……….. 28  Table 2.4  Results of the likelihood ratio tests with grazing and lichen treatments as fixed factors and temporal and period of time as random factors for leaf-toshoot ratio of three major species occurring in each experimental plot, Ledum decumbens, Vaccinium vitis-idaea, and Vaccinium uliginosum; n=35 per treatment. Significant p-values for treatment effects are shown in bold. ……………………………………….……………………………………….….. 32  Table 2.5  Results of the likelihood ratio tests with grazing and lichen treatments as fixed factors and temporal and period of time as random factors for total number of “hits” per 100 points per plot from point frame data for live, dead, and lichen standing crop; n=3-5 per treatment. Significant p-values for treatment effects are shown in bold. ……………………………………….. 36  Table 2.6  Results of the likelihood ratio tests with grazing and lichen treatments as fixed factors and temporal and period of time as random factors for Shannon-Weaver Diversity Index (H) calculated from point-frame data for live, dead, and lichen standing crop; n=3-5 per treatment. Significant pvalues for treatment effects are shown in bold. ……...…………………….… 39  Table 3.1  Results of the likelihood ratio tests with grazing and lichen treatments as fixed factors and temporal variation and period of time as random factors for net ecosystem production, gross ecosystem production, and ecosystem respiration rates; n=5 plots per treatment. Significant p-values shown in bold. …………………………………………...………………………….. 50  Table 3.2  Chi-Square results of the Kruskal-Wallis tests for Ion Exchange Membranes from summer 2004, 2005 and 2006, and overwinter 2004-2005. n=5 for grazing and n=30 for time. ………………………………………………………… 55  Table 3.3  Mean + SEM for soil nutrient flux values (mg/m2/day) measured using ion exchange membranes. n=30. Data are not separated by treatment, as there was no significant treatment effect. ………………….……………………….… 56  vi  Table 3.4  Chi-Square values for Kruskal-Wallis tests of soil water content (%) at a depth of 12 cm and temperature (°C)at a depth of 2 cm. n=5 for grazing and n=30 for time.…………………………………………………………………….…... 56  Table 4.1  F-probabilities (ANOVA) for the soil water content (%) at a depth of 12 cm and temperature (°C) at a depth of 2 cm measured. S ites were north, east, south, and west of Daring Lake; habitats were dry esker, mesic birch hummock, and wet sedge meadow; trail indicates on or off caribou migratory trails; n=5. …………………………….…………………………….…… 68  Table 4.2  F-probabilities (ANOVA) table for ecosystem respiration. Sites were north, east, south, and west of Daring Lake; habitats were dry esker, mesic birch hummock, and wet sedge meadow; trail indicates on or off caribou migratory trails n=5. ………………………………….…………………………….. 70  Table 4.3  Summary of Kruskal-Wallis analysis of live vascular plant, dead vascular plant, lichen, and moss biomass, and litter standing crop. n=5. ……………72  vii  List of Figures Figure 1.1  Barren-ground caribou resting at Daring Lake, Northwest Territories (64°52’N, 111°37’W) as they head south towards tree line at the end of summer. ………………………………………….….………………………………... 10  Figure 1.2  a) Sedge meadow and b) mesic tundra habitats in the low-Arctic are commonly used by caribou for feeding during migration in the spring and fall (photos: Daring Lake, NWT). ………………..…………………………..……. 11  Figure 2.1  Map showing location of Daring Lake, NWT (from Natural Resources, Canada 2002). …………………………………………………………………….….. 17  Figure 2.2  Map showing site location on the vegetation map for Daring Lake (from Obst 2008). …………………………………………………………………………… 19  Figure 2.3  Examples of the grazing and lichen removal treatments at Daring Lake, NWT. The collar in each photo is 50 cm x 50 cm. …………………………...... 20  Figure 2.4  An example of an experimental plot layout. The type of sampling of each subplot was randomly assigned for each experimental plot. The subplots used for biomass extraction for each sample period were also randomly chosen. .………………………………………………………...…………………...... 21  Figure 2.5  Mean (± SE) standing crop (g/m2) of live vascular plant in all treatments (CL, CN, ML, MN, HL, HN; see text for details) during pregrazing, postgrazing, peak standing crop and senescent periods in 2004, 2005 and 2006 except peak 2005; n=3-5. …………………………………….………………………..……. 26  Figure 2.6  Mean (± SE) standing crop (g/m2) of dead vascular plant standing crop (g/m2) in all treatments (CL, CN, ML, MN, HL, HN) during pregrazing, postgrazing, peak standing crop and senescent periods in 2004, 2005 and 2006 except peak 2005; n=3-5. The letters above the bars indicate results of a posteriori analysis: means sharing the same letter do not differ significantly. …………………………………………...……………………….……. 27  Figure 2.7  Mean (± SE) lichen standing crop (g/m2) from all treatments (CL, CN, ML, MN, HL, HN; see text for descriptions) during pregrazing, postgrazing, peak standing crop and senescent periods in 2004, 2005 and 2006 except peak 2005; n=3-5. The letters above the bars indicate results of a posteriori analysis: means sharing the same letter do not differ significantly. …...…. 29  Figure 2.8  Mean (± SE) moss standing crop (g/m2) from all treatments (CL, CN, ML, MN, HL, HN; see text for descriptions) during pregrazing, postgrazing, peak standing crop and senescent periods in 2004, 2005 and 2006 except peak 2005; n=3-5. …………………………………………………………………….......... 30  Figure 2.9  Mean (± SE) litter standing crop (g/m2) from all treatments (CL, CN, ML, MN, HL, HN; see text for descriptions) during pregrazing, postgrazing, peak standing crop and senescent periods in 2004, 2005 and 2006 except peak 2005; n=3-5. The letters above the bars indicate results of a posteriori analysis: means sharing the same letter do not differ significantly. …….... 31  viii  Figure 2.10  Mean (± SE) leaf-to-shoot ratios of Ledum decumbens from all treatments (CL, CN, ML, MN, HL, HN; see text for descriptions) during pregrazing, postgrazing, peak standing crop and senescent periods in 2004, 2005 and 2006 except peak 2005; n=3-5. The letters above the bars indicate results of a posteriori analysis: means sharing the same letter do not differ significantly. …………………..………….…………………………………….……. 33  Figure 2.11  Mean (± SE) leaf-to-shoot ratios of Vaccinium vitis-idaea from all treatments (CL, CN, ML, MN, HL, HN; see text for descriptions) during pregrazing, postgrazing, peak standing crop and senescent periods in 2004, 2005 and 2006 except peak 2005; n=3-5. The letters above the bars indicate results of a posteriori analysis: means sharing the same letter do not differ significantly. …………………………………..……………………………...……… 34  Figure 2.12  Mean (± SE) leaf to shoot ratios of Vaccinium uliginosum from all treatments (CL, CN, ML, MN, HL, HN; see text for descriptions) during pregrazing, postgrazing, peak standing crop and senescent periods in 2004, 2005 and 2006 except peak 2005; n=3-5. …………….……………..…… 35  Figure 2.13  Number of counts per 100 points of the point frame and Shannon-Weaver diversity indices (H) for live vascular plants in all treatments (CL, CN, ML, MN, HL, HN; see text for descriptions) during peak standing crop in 2004, 2005 and 2006. Data are means with SE bars, n=5. ……………………………37  Figure 2.14  Number of counts per 100 points of the point frame and Shannon-Weaver diversity indices (H) for dead vascular plants in all treatments (CL, CN, ML, MN, HL, HN; see text for descriptions) during peak standing crop in 2004, 2005 and 2006. Intact dead leaves and shoots that were still rooted in the ground were counted. Data are means with SE bars, n=5. The letters above the bars indicate results of a posteriori analysis: means sharing the same letter do not differ significantly. ……………………………………………….…. 37  Figure 2.15  a) Number of counts per 100 points of the point frame and b) ShannonWeaver diversity indices (H) for lichens in all treatments (CL, CN, ML, MN, HL, HN; see text for descriptions) during peak standing crop in 2004, 2005 and 2006. Data are means with SE bars, n=5. The letters above the bars indicate results of a posteriori analysis: means sharing the same letter do not differ significantly. …………………………………………………………..…. 38  Figure 3.1  Map of Daring Lake, NWT (64°52’N, 111°3 7’W)(from Natural Resources, Canada 2002). …………………………………………………….……………..…… 45  Figure 3.2  Seasonal carbon exchange values for 2004, 2005, and 2006. Each data point is the mean flux from the control treatment (CL) during that day with error bars indicating the standard error of the mean; n=30. The same pattern was observed for the grazing and lichen removal treatments (data not shown). ………………………………………….…………………………….…. 51  Figure 3.3  Net Ecosystem Production values by treatment for early, peak, and late season of 2004, 2005, and 2006; n=30. The letters above the bars indicate results of a posteriori analysis: means sharing the same letter do not differ significantly. NEP values for all the sample dates are shown in Appendix A. .………………………………………………………………………………………….. 52  ix  Figure 3.4  Gross Ecosystem Production values by treatment for early, peak, and late season of 2004, 2005, and 2006; n=30. GEP values for all the sample dates are shown in Appendix B. ………………………..………………………….……. 53  Figure 3.5  Ecosystem Respiration by treatment for early, peak, and late season of 2004, 2005, and 2006; n=30. ER values for all the sample dates are shown in Appendix C. …………………………………………….……………………………. 54  Figure 3.6  Soil water content for 2004, 2005, and 2006. Values were measured as % volumetric water content. n=5. Standard errors were too large to be included in the graph and are presented in Appendix D. ………….………… 57  Figure 3.7  Soil temperature (°C) taken at 2 cm belo w the soil surface for 2004, 2005, and 2006. n=5. Standard errors were too large to be included in the graph and are presented in Appendix E. …………………………………….…………. 58  Figure 4.1  Caribou migratory trails running through a) dry, b) mesic, and c) wet habitats near 64°52’N, 111°37’W in the Northwest Te rritories. …….……….. 66  Figure 4.2  Mean (+SEM; n=5) soil temperature (°C) a t a depth of 3 cm, and mean (+SEM; n=5) soil water content between 0-12 cm (%) on, and adjacent to, caribou migratory trails in three habitat types: dry esker, mesic birchhummock, and wet sedge meadows at four sites (north, east, south, west) within 20 km of Daring Lake, Northwest Territories. Sampling took place on August 12-13, 2006 between 0945 and 1700 hrs. Refer to Table 4.1 for the results of the statistical analyses. …………………….…………………………. 69  Figure 4.3  Mean (+SEM; n=5) ecosystem respiration values (µmol/CO2/m2/s) taken on, and adjacent to, caribou migratory trails in three habitat types: dry esker, mesic birch-hummock, and wet sedge meadows; at four sites: north, south, east and west of Daring Lake, Northwest Territories. Sampling took place on August 12-13, 2006 between 0945 and 1700 hrs. Refer to Table 4.2 for the results of the statistical analyses. ……………………..…………………….71  Figure 4.4  Mean (+SEM; n=5) live aboveground vascular plant biomass (g m-2) taken on, and adjacently to, caribou migratory trails in three habitat types: dry esker, mesic birch-hummock, and wet sedge meadows; at four sites: north, south, east and west of Daring Lake, Northwest Territories. Sampling took place on August 12-13, 2006 between 0945 and 1700 hrs. Refer to Table 4.3 for the results of the statistical analyses. …………………….………………... 73  Figure 4.5  Mean (+SEM; n=5) dead plant standing crop (g m-2) taken on, and adjacent to, caribou migratory trails in three habitat types: dry esker, mesic birchhummock, and wet sedge meadows; at four sites: north, south, east and west of Daring Lake, Northwest Territories. Sampling took place on August 12-13, 2006 between 0945 and 1700 hrs. Refer to Table 4.3 for the results of the statistical analyses. ……………………………………………………………. 74  Figure 4.6  Mean (+SEM; n=5) litter standing crop (g m-2) taken on, and adjacent to, caribou migratory trails in three habitat types: dry esker, mesic birchhummock, and wet sedge meadows; at four sites: north, south, east and west of Daring Lake, Northwest Territories. Sampling took place on August 12-13, 2006 between 0945 and 1700 hrs. Refer to Table 4.3 for the results of the statistical analyses. ……………………………………………….………….... 75  x  Figure 4.7  Mean (+SEM; n=5) lichen biomass (g m-2) taken on, and adjacent to, caribou migratory trails in three habitat types: dry esker, mesic birch-hummock, and wet sedge meadows; at four sites: north, south, east and west of Daring Lake, Northwest Territories. Sampling took place on August 12-13, 2006 between 0945 and 1700 hrs. Refer to Table 4.3 for the results of the statistical analyses. ………………………………………….……………………... 76  Figure 4.8  Mean (+SEM; n=5) moss biomass (g m-2) taken on, and adjacent to, caribou migratory trails in three habitat types: dry esker, mesic birch-hummock, and wet sedge meadow at four sites north, south, east and west of Daring Lake, Northwest Territories. Sampling took place on August 12-13, 2006 between 0945 and 1700 hrs. Refer to Table 4.3 for the results of the statistical analyses. ……………………………………………………………..….. 77  xi  Acknowledgements I thank my supervisor Greg Henry for his good nature, consistency, and support. My supervisory committee, Elyn Humphreys, Roy Turkington, and Rob Guy, provided valuable input, and I very appreciative of their support and suggestions. This endeavor would not have been possible without the incredible logistical support I received in the lab and field. For this I thank Carlie Smith, Daniel Ramos, Laura Machial, Steven Guenther, Mike Treberg, Carolyn Churchland, Rebecca Klady, Becky Zalatan, the Henry Lab work study students, the Daring Lake Tundra Ecological Research Station staff, Steve Matthews and Karin Clark from the Department of Environment and Natural Resources of the Government of the Northwest Territories, and the field pilots, especially Daryl at Great Slave Helicopters. I am most grateful for having such amazing family and friends, who provided me with whatever support I needed, whether it was pep talks, discussions about my research, proofreading, or dogsitting so that we could do field work. Many thanks to: Jackie Ngai, Odette O, Audrey Chen, Ursula Guenther, Lutz Guenther, Susan Guenther, Kara Ko, Bryan Gick, Maureen Lunn, Larry Lunn, and Deborah Peldszus. Special thanks go to my husband, Steven Guenther, and our dog, Ritter, for being my partners in every adventure. Financial support was provided by the Natural Sciences and Engineering Research Council of Canada, the Northern Scientific Training Program of Indian and Northern Affairs Canada, and the Association of Canadian Universities for Northern Studies.  xii  1  INTRODUCTION  1.1  The Changing Atmosphere  The objective of this research is to provide new information on the impacts of caribou grazing on tundra ecosystems and the role these impacts play in net ecosystem carbon exchange. The effects of climate change may alter the carbon balance in Arctic systems, and these changes may be mediated by and affect caribou-plant dynamics. The atmospheric concentrations of CO2, CH4, N2O, and other biogenic trace gases have been rapidly increasing over the past two centuries and long-term records indicate that the rate and magnitude of change in these atmospheric gases is greater than any atmospheric perturbation experienced in the history of the planet (Matson & Harriss 1995; ACIA 2005), especially since the Industrial Revolution. The increases in biogenic trace gases have led to the enhancement of the “greenhouse” effect. Normally, incoming solar radiation heats the Earth-atmosphere system, and thermal radiation emissions released from the Earth to space balances the incoming radiation. Biogenic trace gases absorb some of this thermal radiation and re-radiate it, which results in the Earth’s surface becoming warmer than it would without an atmosphere. The recent increases in the concentration of these gases lead to a greater greenhouse effect and the warming climate on Earth. Albedo, or the reflectivity of a surface, has a substantial effect on the energy balance. Snow and ice have high albedo and reflect more radiation than dark surfaces. With the melting of Arctic snow and ice, the reduced global albedo decreases the reflectivity and increases the absorption of short wave radiation. This resultant heating of the Earth’s surface causes an increase in the emission of outgoing radiative energy to Earth, causing feedbacks that further increase the rate of climate change (ACIA 2005). Atmospheric CO2 concentrations were between 260-290 ppm from the end of the last ice age to the beginning of the 19th century (Maxwell 1992). Anthropogenic causes since then (e.g. landuse change, fossil fuel consumption) have resulted in an increase of about 20%, with a doubling of present day values (currently near 390 ppm) expected within 50 years (Maxwell 1992; ACIA 2005). Increases in other trace gases (e.g. CFCs, CH4, N2O) will also contribute to climate change by causing additional absorption and emission of thermal radiation in the atmosphere and/or depleting the stratospheric ozone, which increases incoming radiation. Water vapor is the most important greenhouse gas because of its concentration, and higher atmospheric temperatures increase evaporation and water vapor holding capacity of the atmosphere, which 1  further increases warming. The increase in water vapor could also increase cloud water content and reflectivity, reducing solar heating of the atmosphere and surface, lowering temperatures (Maxwell 1992; ACIA 2005). Biogenic trace gas exchange between surfaces and the atmosphere is governed by biotic and abiotic properties of ecosystems, such as physical transport of heat and materials through soils, surface-air boundaries, and microbial properties (Matson & Harriss 1995). The interactions between the biosphere and atmosphere must first be understood in order to make predictions about the future global carbon cycle and evaluate the potential outcomes in the face of climate change (Vourlitis et al. 1993, Norman et al. 1997, Longdoz et al. 2000, Drewitt et al. 2002).  1.2  Climate Change in the Arctic  Arctic ecosystems are undergoing the earliest and most intense changes for a few reasons. First, Arctic climate is highly variable. Stochastic events, such as summer freezing and storms, can kill a portion of the next generation of wildlife species, such as polar bears and caribou, due to intolerable climatic conditions or resultant food shortages (ACIA 2005). Second, some of the most drastic changes are projected to happen in the Arctic due to its location and current conditions. For example, the melting of sea and land ice is one of the first responses to increased temperatures, and the Arctic has the second largest amount of sea and land ice in the world. Positive feedback cycles related to the loss of ice and snow cover also exacerbate the effects of warming in the Arctic. The loss of ice on sea and land will result in reduced surface reflectivity of solar radiation, or albedo, which will further warm the Earth’s surface (Maxwell 1992; ACIA 2005). Also, the loss of sea and land ice could delay the onset and accelerate the end of the winter season, lead to warmer, shorter winters, result in more moderated climate during summers, longer growing seasons, and cause coastal erosion due to increased storm action and rising sea levels from warmer oceans and melting glacial ice (Maxwell 1992, ACIA 2005). Third, the Arctic has short growing seasons and lower species diversity compared to temperate areas of the world (ACIA 2005). Ecological resilience of populations and ecosystems is related to having sufficiently long recovery times following disturbance or damage and having sufficient diversity at all levels of biological organization to exist and maintain the ecosystem. The projected increases in temperature and precipitation are predicted to alter the carbon balance in northern ecosystems. Deeper active layers will result from permafrost warming, and stores of 2  frozen carbon could potentially be released into the atmosphere as carbon dioxide if soils become aerobic. If soils remain or become saturated, C may be lost as methane or through DOC transport and subsequent oxidation. The extent to which current net carbon sinks in the Arctic will become net sources is unknown, as increases in photosynthesis and respiration are also expected to occur (Welker et al. 2004b; Oberbauer et al. 2007).  1.3  Ecosystem Carbon Exchange  Carbon is the basic building block of life, and thus, its measurement is important for understanding the patterns and processes that create and maintain compounds and organisms from the cell to ecosystem scale. Carbon is transferred between organic and inorganic forms by processes in the global carbon cycle, with carbon dioxide as the main form that moves through the atmosphere, hydrosphere, and biota (Townsend et al. 2000). The sequestration of CO2 in photosynthesis and the release of the carbon contained in photosynthetic products back to the atmosphere and hydrosphere in respiration are the main driving forces of the global carbon cycle (Townsend et al. 2000). Crawley (1983) stated that “all the carbon fixed by plants in photosynthesis is destined to be eaten,” and the transfer of carbon from individual plants to animals is part of larger-scale carbon dynamics. At the ecosystem scale, the net carbon budget includes net primary productivity, soil respiration, and carbon export of litter and other particulate and dissolved organic matter (Oechel & Billings 1992). Catovsky et al. (2002) suggest that although plants fix carbon as biomass, they also translocate carbon to the soil, so that the entire net ecosystem productivity must be examined to determine precise carbon storage values. The principal controls on the carbon cycle over the longer term are soil temperature, plant community structure, reduction-oxidation boundaries associated with the water table, and the chemical composition of plant tissues and soils, including peat (Bubier et al. 1993; Whiting & Chanton 1993; Yavitt et al. 1997). Little is known about the carbon balance of tundra systems, although they cover nearly 30% of Canada and contain significant amounts of the planet’s soil carbon (Oechel & Billings 1992). The study of carbon flux, or net ecosystem exchange of carbon dioxide (NEE), in ecosystems is important in quantifying their source-sink status. This is particularly important in tundra systems, which are generally net carbon sinks, but may shift to net carbon sources if perturbed by environmental factors, such as increased atmospheric temperature (Flanagan et al. 2002; 3  Oberbauer et al. 2007). Peatlands are habitats that are rich in stored carbon and Arctic habitats, particularly northern peatlands, store large amounts of carbon. Peatlands cover 3% of the global land area (Maltby & Immirzi 1993), 12% of the land area in Canada, and store 56% of the global soil carbon supply (Tarnocai 2006, 2009). In Arctic ecosystems, low decomposition rates in a system with low primary productivity have resulted in the accumulation of large amounts of carbon in the form of soil organic matter because historically, the production of carbon exceeded its decomposition rates (Oechel & Billings 1992). Ninety percent of the carbon in Arctic ecosystems is stored in soils, and 98% percent of the carbon is stored in soils in sub-Arctic and Arctic peatlands (Oechel & Billings 1992). Grazing could cause changes in the plant community, primary productivity and the amount of carbon storage in plant biomass. Due to the large carbon storage capacity, Arctic habitats, including tundra and peatlands, may be particularly sensitive to the effects of grazing, especially when these effects are compounded with climate shifts.  1.4  Grazing  Grazing is the act of consuming leafy plant tissue, or standing forage crop, by an animal (Vallentine 1990; Krohne 1998). This process usually involves the sublethal removal of aboveground plant tissue. Responses to grazing can be at different levels: the individual plant, community, and ecosystem levels. It is expected that grazing will cause changes in each of these levels because it alters individual plant physiology and morphology, species composition and interactions, and ecosystem processes, such as net primary production and nutrient dynamics. The carbon balance of individual plants involves losses to respiration and gains from photosynthesis. Grazing causes morphological and physiological changes, which can alter the chemistry, growth, and reproduction of individuals. More specifically, grazing changes the allocation of resources to roots, shoots, and leaves, affecting photosynthetic and respiration rates (Coughenour 1985; Oechel & Billings 1992), resulting in differential survival, growth, and competition between individuals and species. A variety of studies have observed these grazinginduced plant changes, including alterations in root:shoot ratios (e.g. Lentz & Cipollini 1998), alterations in leaf and shoot turnover rates (e.g. Chapin & Shaver 1996), higher photosynthetic rates (e.g. McNaughton 1979a), and increased leaf (e.g. McIntire & Hik 2002) and shoot production (e.g. Tolvanen & Henry 2000).  4  Grazing also causes changes at the community level. Species composition can be altered when grazing reduces or eliminates sensitive species and promotes grazing-tolerant species. Grazers can shift community composition to previous successional states, e.g. from shrub- to grassdominated cover, and they can maintain species diversity in a system that would be homogeneous without grazing. The rate of vegetation succession can be slowed or quickened, depending on whether grazers select earlier or late successional species (e.g. Davidson 1993; Lesica & Cooper 1999; Dorman et al. 2000). At the ecosystem level, grazing causes changes in patterns, such as plant spatial distribution (e.g. Gibson 1988), and processes, such as nutrient cycling (e.g. Cargill & Jefferies 1984; Bazely & Jefferies 1985; van der Wal et al. 1998), net above-ground primary production (NAPP) (e.g. McNaughton 1979a & b; Hik & Jefferies 1990) and CO2 flux (Oechel & Billings 1992).  1.5  Compensatory Growth  Compensatory growth is a type of plant response to grazing and is important to assess because it reflects the ability of plants to recover after defoliation (Hik & Jefferies 1990; Raillard & Svoboda 1999). Belsky (1986) defined it as a positive response of plants to injury, ranging from a partial replacement of lost tissue to a net productivity exceeding that of the uninjured control plants. There are three categories of compensation based on a cost-benefit balance for plants: overcompensation, where the total dry weight of grazed plants is higher than ungrazed controls; exact compensation, where the dry weight of grazed and ungrazed plants are the same; and undercompensation, where the dry weight of grazed plants is less than that of ungrazed controls (Belsky 1986). While compensatory growth can occur at the individual plant level (e.g. vegetative regrowth), it is also manifested at the community (e.g. species diversity), and ecosystem levels (e.g. NAPP) (Rosenthal & Kotanen 1994; Agrawal 2000). Higher levels of compensatory growth reflect a greater tolerance to grazing, but resources need to be available for compensation to occur. Grazing can provide these resources; for example, by increasing rates of nitrogen inputs and consequent mineralization (McNaughton 1979b; Bazely & Jefferies 1985; Henry 1998).  5  1.6  Northern Grazing Systems  Grazing occurs in essentially all ecosystems, ranging from the savannahs of Africa (e.g. McNaughton 1979 a & b) to Arctic coastal marshes (e.g. Cargill & Jefferies 1984). The interactions between grazers and their forage have developed through a coevolutionary relationship (McNaughton 1984; Coughenour 1985). In grasslands, grazed vegetation becomes short, dense, and prostrate in growth form, and vegetation communities can be maintained as grazing lawns for decades (McNaughton 1984). Arctic grazing systems are characterized by short cool growing seasons, typically about 10-12 weeks, and long cold winters. The short growing seasons result in bursts of growth in plants and short periods of intense feeding by grazers. The lack of energy in Arctic systems has impacts on both plants and animals, causing low diversity of plants and grazers. There are different vegetation communities in the Low Arctic: shrub tundras, sedge-dwarf-shrub tundras, steppe tundras, tussock-dwarf-shrub tundras, mires, and coastal salt marshes (Bliss & Matveyeva 1992). Various types of grazers use Arctic habitat, including avian grazers, such as Canada, Brant, and Snow Geese (O 2003), and ungulates, such as muskoxen and caribou (Griller 2001). Geese have been a model system to study grazing in the Arctic. Pacific Black Brant Geese create and maintain the habitat in which they forage in the tidal meadow communities of western Alaska. Person et al. (1998) found that grazing by Pacific Black Brant did not reduce the NAPP of the grazing-tolerant species, Carex subspathacea. Also, Carex ramenskii, which occurs in the same landscape as C. subspathacea increased in NAPP and leaf nitrogen concentration after defoliation (Ruess et al. 1997). On Bylot Island, Greater Snow Geese feed in meadows where Dupontia fisheri, Eriophorum spp., and Carex aquatilis are the dominant species (Gauthier et al. 1995). Overall, studies indicate Greater Snow Geese impact their habitat, but the effects vary over temporal scales. For example, Gauthier et al. (1995) found Greater Snow Goose grazing lowered the NAPP of Dupontia fisheri but not Eriophorum scheuchzeri over a period of three years. However, Beaulieu et al. (1996) found that irrespective of the number and frequency of grazing episodes by Greater Snow Geese within one growing season, the NAPP and tiller production of Dupontia fisheri and Eriophorum scheuchzeri were not affected. In the past few decades, the characteristics and dynamics in Lesser Snow Goose systems in the James and Hudson Bay lowlands have drastically changed. Intensive grazing and grubbing by Lesser Snow  6  Goose have caused many areas at La Pérouse Bay with formerly lush grazing lawns to become devegetated mudflats (Handa & Jefferies 2000; Handa et al. 2002; Jefferies & Rockwell 2002). Muskoxen usually graze at moderate levels because they are adapted to survive conditions with scarce availability of low quality forage (Jefferies et al. 1992). At moderate grazing intensities, graminoid productivity was unaffected by muskoxen grazing (Smith 1996). However, when they do graze intensively, muskoxen alter plant communities by drastically reducing plant biomass and facilitating moss-dominated systems and altering species composition and abundance (McKendrik 1981; Henry 1998; Raillard & Svoboda 2000; Elliott & Henry 2011), but individual species response to grazing varies (Tolvanen & Henry 2000). Climate models predict the lengthening of the growing season in the Arctic with earlier snow melt and delayed onset of freezing and snow accumulation. Therefore, habitats will remain snow-free until autumn, at which time there is a reduction of PAR and day length. Higher temperatures may increase decomposition rates, lower water tables, and deepen the active layer, possibly resulting in a doubling of NAPP (ACIA 2005). The carbon-nutrient balance will be altered, and photosynthesis and carbohydrate production may increase, so that nutrient uptake will continue to limit plant growth as opposed to carbon fixation. Reduction in forage quality through the increased production of secondary metabolites is possible due to excess carbohydrates not needed for growth (Jefferies et al. 1992). Earlier leaf-out and delayed senescence, increased net primary production, greater reproductive success of plants, as well as changes in species composition and relative abundance are predicted to have feedbacks on gas exchange, carbon balance, surface energy balance, and nutrient cycling (ACIA 2005; Oberbauer et al. 2007). In addition, feedbacks to higher trophic levels through changes in forage quality and quantity are possible (Russell et al. 2002). Each species will respond differently to climate change, so assemblages of species will not remain intact as they migrate north when the climate shifts. Communities will reorganize, and the outcomes are dependent on herbivore modulation, forage species growth form/characteristics, and vegetation responses to changes in nutrient availability and cycling. The time scale for all these changes to happen ranges from days (i.e. physiology of individual plants) to hundreds of years (i.e. moving tree line) (e.g. Shaver et al. 2000).  7  1.7  Caribou (Rangifer tarandus)  Caribou (Rangifer tarandus) are one of the few large-bodied grazing mammals in the Arctic, with physiological and morphological adaptations to their food source, a complex social structure that varies seasonally, and complicated ecological interactions with other organisms and their environment. In North America, the periodicity of their population cycles is believed to be between 40 and 70 years, though there are few long-term records (Gunn 2005; Zalatan et al. 2006). Both direct and indirect climatic influences affect caribou abundance, and therefore, their biology, grazing patterns, and their regional population regulation may be influenced by climate change. Currently, woodland caribou (Rangifer tarandus caribou) populations are listed as “Secure” by the Northwest Territories Status Rank and of “Special Concern” by Committee on the Status of Endangered Wildlife in Canada (COSEWIC), while barren-ground caribou (Rangifer tarandus groenlandicus) populations are listed as “Sensitive” by the Northwest Territories Status Rank and of “Special Concern” by COSEWIC, and Peary caribou (Rangifer tarandus pearyi) populations are listed as “At Risk” by the Northwest Territories Status rank and “Endangered” by COSEWIC (NWT 2006). Given the current status of these caribou populations, they are sensitive to changes in weather or forage quality. The most recent census numbers of the barren-ground caribou herds in the Northwest Territories show that most herds are in a declining phase of their population cycles (Table 1.1).  8  Table 1.1 Recent census data on the barren-ground caribou herds in the Northwest Territories, Canada reported by the Circumarctic Rangifer Monitoring and Assessment Network (CARMA). Source: Figure 5.3-5 from NWT Environmental Audit Status of the Environment Report, 2005.  Last Census  NWT Herd  Current size (n)  Estimated Size Status  Year  Proposed  Estimated Rate of  Next Census  Harvest (n)  Bathurst  186,400  Declining  2003  Unknown  Estimate of ~14,000 in 1996  Beverly  286,000  Unknown  1994  2000  14,0001  Qaminirjuaq  496,000  Unknown  1994  2000  -  Ahiak  200,000  -  1996  Unknown  Unknown  Bluenose East  66,600  Declining  2005  -  5,0002  Bluenose West  20,800  Declining  2005  -  5,0002  Cape Bathurst  2,400  Declining  2005  -  -  Porcupine  123,000  Declining  2001  -  3,000  www.carmanetwork.com 1  – estimate is probably for both the Beverly and Qamanirjuaq herds.  2  – total estimated harvest of Bluenose East, Bluenose West and Cape Bathurst herds.  Barren-ground caribou migrate in the spring and fall from wintering to summer feeding and calving grounds and back (Bean 2003; Zalatan et al. 2006). They are selective feeders foraging on a mixture of shrub tundras, sedge-dwarf-shrub tundras and wetland vegetation types (Bliss & Matveyeva 1992). The time period for migration of barren-ground caribou coincides with the phenology of vegetation and, thus, is short because they are sensitive to local variation in forage and preferentially use sites with higher nutrients and minerals, potentially enhancing graminoid growth in sites that are inherently more productive (Post & Klein 1996).  9  Figure 1.1 Barren-ground caribou resting at Daring Lake, Northwest Territories (64°52’N, 111°37’W) as they head south towards tree line at the end of summer.  Photo by P.O Caribou forage in a variety of plant communities. Wetlands such as bogs, wetland muskeg or fens (sedge meadows) have plant communities dominated by sedges and grasses, and/or dwarf shrub and forbs species (Bliss & Matveyeva 1992). In these ecosystems, primary species consumed are Carex aquatilis, Eriophorum angustifolium, and Eriophorum vaginatum. Wetlands can be sources or sinks for C (Blodau 2002) but are generally assumed to be net carbon sinks for atmospheric CO2 and have potential for long-term soil carbon sequestration (Oechel & Billings 1992; Oberbauer et al. 2007). Tussock and wet sedge tundra soils contain most of the carbon stores in Arctic soils because of their expansive areas and relatively large NPP (Oechel & Billings 1992). Typical mesic tundra can have ground cover consisting of dwarf heath shrubs, grasses, forbs, mosses, and lichens. In this habitat, caribou forage on deciduous shrubs, such as Betula nana, Salix spp., and grasses and sedges, such as Eriophorum vaginatum and Carex spp. Their primary source of non-vascular forage is lichens that take up to 15 years to regrow after being grazed (Jefferies et al. 1992). Caribou consume between 3 and 50 % NAPP of sedge meadows (Jefferies et al. 1992) but estimates of consumption from other tundra communities are 10  unknown. It was traditionally believed that because the frequency and intensity of caribou grazing at any particular site is episodic, their effects on ecosystem processes, consequent plant production, and nutrient feedback loops are weak or even absent (Jefferies et al. 1992). However, there are observed strong impacts along caribou migratory trails (see Chapter 4), such as vegetation loss and soil compaction, so that this statement may not be verified upon further research. In fact, this study challenges the view of Jefferies et al. (1992) by asserting that constant low level grazing pressure over thousands of years has shaped tundra systems. Because caribou have large population cycles and travel long distances across the tundra every year on generally similar migration routes, their impacts are diffuse but relatively constant.  Figure 1.2 a) Sedge meadow and b) mesic tundra habitats in the low-Arctic are commonly used by caribou for feeding during migration in the spring and fall (photos: Daring Lake, NWT).  a)  1.8  b)  Climate Change and Caribou  Many, if not all, of the predicted changes in climate will affect caribou populations. Caribou roam over wide areas to find adequate food, and shortages of forage in one season at a particular 11  site can be compensated for by the exploitation of the same vegetation communities at other locations. To date, studies have shown the ability of vegetation to recover from caribou grazing, with the most compensation occurring in preferred plant species (e.g. Henry & Gunn 1991, Brathen and Oksanen 2001). Although this can mostly be attributed to the co-evolution of vegetation to repeated defoliation, enhanced nutrient cycling by indirect effects such as fecal deposition, also aids in plant regrowth. Caribou and other ungulates alter nutrient cycling, particularly nitrogen, by changing litter quality, which affects N mineralization by adding N through urine and feces that are readily taken up by plants and microbes. Ungulates also influence N availability to plants by reducing microbial competition (Jefferies et al. 1992, Hobbs 1996). Changes in nutrient cycling will have feedback effects at all levels of organization from individuals to ecosystems and from detritivores to top consumers. If the climate becomes warmer and wetter, increased snow depth could increase lichen abundance, and warming could change plant community structure and composition. This may positively and negatively affect caribou cows and calves in their calving grounds, respectively. The recovery time for lichens could be extended or may not even occur in areas of extreme environmental change. The size and success of vascular plants increase with warming, leading to loss of lichens through shading (Cornelissen et al. 2001; Walker et al. 2006). Also, with the migration of the tree line northward and the replacement of dwarf shrubs by tall shrubs, subArctic habitats will affect the timing and availability of forage for caribou. Schaefer (1996) found that percent lichen cover in Manitoba forest was positively correlated with maximum thickness and maximum vertical hardness of snow because more snow lead to more moisture through the spring, leading to greater lichen growth. These results show that it is possible that caribou will have sufficient lichen availability, if there is higher snow accumulation with climate change. Changes in vegetation affect caribou survival and reproduction. Many studies have shown that failure of cows to accumulate sufficient reserves over the summer leads to reduced conception in the fall, and low body weight of cows in the spring directly affects calf survival because of reduced milk quality (White 1983). Animal movement, including some caribou herds (e.g. Dolphin and Union Herd in Nunavut), will be restricted by lack of sea ice between islands and lack of consecutive suitable grazing habitat patches, resulting in increased costs for migration and feeding. The tragic occurrence of the loss of caribou that were stranded on Rideout Island is likely to become increasingly 12  common as sea ice thins and recedes earlier (Henry & Gunn 1991). Also, the thinning or loss of lake ice and the potential increase in river discharge from increased snow melt will affect migration. Increased insect harassment will likely occur as temperature increases. This will decrease caribou feeding and resting and increase movements and insect-avoidance behaviors when harassed by insects (Harper 1955, Downes 1985).  1.9  Caribou Grazing  Biological productivity in Arctic ecosystems is limited by the availability of nutrients, water, and temperature. Herbivores in southern ecosystems generally consume low amounts of plant NAPP (ca. 10% NAPP), but Arctic herbivores consume between 50 and 90% NAPP (Jefferies et al. 1992). According to Jefferies et al. (1992), caribou graze using a Selective-Exploitation-and-Idling Model. The rates of forage intake depend on the growth form and abundance of the food source. The metabolic rate of caribou falls in the winter (idling), and caribou can sustain a negative nitrogen balance in the winter by using body protein reserves.  1.9.1  Forage Sources  Caribou feed on lichens at the beginning (Oct-Nov) and end of winter (May-June), then switch to foraging on vascular plants, such as Carex, Dupontia, and Eriophorum predominantly, but also on Salix spp., Betula spp. and some mosses, in the summer (Skogland 1984, Adamczewski et al. 1988, Jefferies et al. 1992). Caribou prefer lichens at the beginning and end of winter because, although they have low nitrogen, they are rich in energy, such as polysaccharides and lipids (Jefferies et al. 1992). It is during the summer that caribou accumulate almost all of their weight (Brathen & Oksanen, 2001), with the caloric intake of caribou 35-45% lower in winter than summer (McEwan & Whitehead 1970). Over winter, however, caribou often have a varied diet consisting of woody plants, such as Empetrum, as well as sedges, Dryas, Equisetum, and some lichens (Skogland 1984, Adamczewski et al. 1988, Jefferies et al. 1992, Storeheier et al. 2002). The reason for the switches in their diet is because forage nutrition varies seasonally. In the  13  summer (July-beginning August), plants have high nitrogen, but this rapidly declines in the autumn and over winter, with the lowest plant nitrogen in the spring (Adamczewski et al. 1988). Lichen-dominated vegetation covers 8% of the world’s terrestrial surface (Storeheier et al. 2002), which bodes well for caribou diet. However, quite extensive damage to lichens occurs both by consumption and trampling. The recovery of tundra vegetation to grazing and trampling damage by large numbers of caribou takes many years (Henry & Gunn 1991, Arseneault et al. 1997, Brathen & Oksanen, 2001). Jefferies et al. (1992) cited one study where it took 15-20 years for recovery of a winter lichen pasture after grazing and cited another study where 22 years after a reindeer population crash, lichens had only recovered 10% relative to an adjacent area with no grazing history. From 1989-1992, Arseneault et al. (1997) reported that even though there was no significant reduction in lichen cover in younger stands, caribou grazing caused a 5% per year decrease in lichen cover in stands > 50 years post-fire. Overall, the number of caribou exceeded the carrying capacity of the habitat in the study area (lichens grow 1%/yr) through both consumption and collateral damage by trampling in the feeding areas (Arseneault et al. 1997). The hundreds of thousands of caribou that migrate through low-Arctic tundra affect the landscape by trampling and grazing along the migratory pathways they use semi-annually (Bean 2003). However, the effects of caribou grazing on tundra are poorly studied and understood. The impacts by caribou have implications for the carbon uptake and release in the vegetation and soils of these migratory trails and the areas used by caribou for grazing. Plant species may exhibit differential responses that would contribute to changes in species composition, and consequently, NEP.  1.10 Objectives The purpose of this research was to examine the net ecosystem exchange of carbon in low-Arctic tundra systems used by caribou and the changes to the carbon balance caused by grazing by large populations of caribou. Specifically, the research focused on the effects of the Bathurst caribou herd foraging on lichens, herbaceous plants, and deciduous shrubs in central Northwest Territories, Canada. Previous studies on general plant responses to grazing at different scales and net ecosystem carbon exchange will serve as a foundation for this little-researched topic: the effects of caribou 14  grazing on vegetation in Arctic ecosystems. Although CO2 flux has been studied in the Canadian high-Arctic (e.g. Welker et al. 2004b; Oberbauer et al. 2007) and low-Arctic (e.g. Lafleur & Humphreys 2008), no study has yet examined the effects of grazing on CO2 flux on tundra systems. This research will provide new information on the impacts of caribou grazing on tundra ecosystems and the role these impacts play in net ecosystem carbon exchange. The objective of this research is to quantify the impacts of caribou grazing on the habitat they use in the lowArctic during spring and fall migrations. Particularly, I examined responses at the individual plant, community, and ecosystem levels. It is expected that grazing will cause changes in each of these levels because alterations in individual plant physiology and morphology as a result of grazing could contribute to changes in species composition and interactions, and consequently, ecosystem processes, such as net primary production, nutrient dynamics, and net ecosystem carbon exchange (Catovsky et al. 2002).  1.11 General Hypotheses A variety of research approaches were used to test some general hypotheses: 1) I used defoliation experiments to simulate grazing of all plant species in order to test the following predictions: a) lichens will not re-grow after being removed, b) the recovery of plants to high intensity defoliation treatments will be lower than those in low intensity treatments, and c) there will be more dramatic differences in the high level treatments in species composition, plant biomass, and net ecosystem exchange compared to the experimental control plots due to the inability of plants to recover from heavy tissue loss. 2) I used surveys to examine natural caribou trampling and grazing. I predicted that ecosystem respiration rates would be higher off caribou migratory trails because there is little vegetation on trails as a result of caribou trampling and grazing.  15  2  EFFECTS OF SIMULATED GRAZING ON PLANT GROWTH IN A MESIC BIRCH HUMMOCK HABITAT IN LOW-ARCTIC TUNDRA  2.1  Introduction  Caribou grazing has altered Arctic landscapes for thousands of years (e.g. Henry & Gunn 1991; Jefferies et al. 1992). In systems such as this, where grazers and plants have the opportunity to co-evolve, it can be expected that grazing could alter plant responses such that plants would compensate for lost tissues by changing their growth rates. This compensatory growth can be negative, exact or positive (Belsky 1986). If the net primary production of grazed plants is less than that of ungrazed controls the response is undercompensation. In certain relatively rare situations, moderately grazed plants can produce more biomass than ungrazed plants over a fixed time period and this is termed overcompensation. Exact compensation is where the dry weight of grazed plants equals that of ungrazed controls. As caribou only graze in the same area twice a year at most (Bean 2003; Gunn 2005; Zalatan et al. 2006), the impacts they have may or may not be sufficiently intense to result in compensatory growth. Typical mesic Arctic tundra is composed of dwarf shrubs, grasses, forbs, mosses, and lichens (Bliss & Matveyeva 1992; Walker et al. 2003). In this habitat during the early spring and through the fall when vascular plants are nutrient poor, the primary source of forage for caribou is lichens, which take up to 15 years to regrow after being grazed (Jefferies et al. 1992). During the summer, barren ground caribou mostly forage on shrubs, such as Betula nana, Salix spp., and grasses and sedges, such as Eriophorum vaginatum and Carex spp. Caribou consume between 3 and 50 % of net aboveground primary production (NAPP) of sedge meadows (Jefferies et al. 1992), but estimates of consumption in shrub tundra are unknown. The estimates are difficult because the frequency and intensity of caribou grazing at a particular site is episodic. As a result, their effects on ecosystem processes, consequent plant production, and nutrient feedback loops are weak or even absent (Jefferies et al. 1992). However, caribou migratory trails exhibit impacts, such as vegetation loss and soil compaction on a local scale. The objective of this study was to examine the effects of caribou grazing on plant biomass in a low Arctic dwarf shrub tundra to determine what type of compensatory growth is occurring.  16  2.2  Methods  2.2.1  Study Site  Grazing effects were examined in a small part of the migration route of the Bathurst caribou herd, and this was largely a logistical constraint in order to make the study area manageable. The experimental work was done in the low Arctic tundra communities around Daring Lake, Northwest Territories (64°52’N, 111°37’W)(Figure 2.1). Climate data were from a weather station that was installed and monitored by Indian and Northern Affairs Canada, approximately 1.5 km southwest of the study site.  Figure 2.1 Map showing location of Daring Lake, NWT (from Natural Resources, Canada 2002).  The study was conducted at the Tundra Ecology Research Station at Daring Lake, which is maintained by the Government of the NWT. Caribou forage in habitats along a soil moisture continuum, ranging from dry barrens on tops of eskers and outcrops to wetland communities at the base of slopes and near water bodies. Site selection criteria for the study were: 1) homogeneous conditions within the habitat type, i.e. similar soil type and vegetation composition; and 2) relatively low disturbance by other grazers (e.g. microtines, hares, geese). In summer 2004, the experiment was established in mesic tundra habitat. 17  2.2.2 Experimental Design The goal of this experiment was to assess plant responses under simulated grazing that is similar to natural conditions. Important caribou forage species, such as lichens, herbaceous plants, graminoids (e.g. Eriophorum spp. and Carex spp.) and deciduous shrubs (e.g. Salix spp. and Betula spp.) were manipulated to emulate different grazing intensities (Ouellet et al. 1994). Thirty experimental plots were established in birch hummock - heath lichen tundra habitat on a north facing esker at Yamba Lake (Figure 2.2). The vegetation classification used was based on the system developed in this area by the Government of the NWT (Obst 2008).The experiment used six levels of simulated grazing (1 control and 5 treatments) to assess foraging impacts in a 3 X 2 factorial design. The factors were: three levels of simulated vegetation grazing and two levels of simulated lichen grazing (Figure 2.3). The grazing treatments were: control (C) having no grazing, medium (M) with 25% defoliation, and high (H) with 75% defoliation. The percentage of vegetation removed was a visual estimate based on the existing vegetative cover in each plot at the time of removal. Leaves were clipped from individual stems, which remained rooted, as commonly occurs during caribou foraging (White 1983). The grazing treatment was evenly distributed throughout each study plot. For the lichen treatments, lichens were left intact (L) or 100% were removed (N) from the assigned plots using a fork. The combinations of the six experimental treatments were: 1) no grazing and lichens left intact (CL), 2) no grazing with lichens removed (CN), 3) 25% defoliation with lichens left intact (ML), 4) 25% defoliation with lichens removed (CN), 5) 75% defoliation with lichens left intact (HL), and 6) 75% defoliation with lichens removed (HN). There were five replicates of each of the six experimental treatments (5 x 6 = 30 plots). The treatments were randomly assigned to the plots. In each year from 2004 to 2006, the simulated grazing treatments were applied to the entire experimental plot, 6.25 m2, during the early growing season, to emulate the timing of caribou grazing on their migratory trails.  18  Figure 2.2 Obst 2008)  Map showing site location on the vegetation map for Daring Lake (from  19  Figure 2.3 Examples of the grazing and lichen removal treatments at Daring Lake, NWT. The collar in each photo is 50 cm x 50 cm. n=5 for each of the 6 experimental treatments.  2.2.3 Sampling Procedures Each plot was 2.5 m by 2.5 m, and contained 25 subplots, each 0.25 m2, which were used for repeated destructive sampling at the plot level, with different subplots harvested in each sample period, or for long-term monitoring of biophysical attributes. At the beginning of the study, each 0.25 m2 subplot in each experimental plot was randomly assigned for either destructive biomass sampling (n = 18) or repeated plant cover measurements (n = 2, CO2 flux measurements (n = 4) and soil measurements (n = 2)(Figure 2.4). Samples were taken at one of three sample periods, i.e. early, peak, late in the growing season. The vegetation biomass and cover data are presented in this chapter; the CO2 flux and soil data are provided in Chapter 3.  20  Figure 2.4 An example of an experimental plot layout. The type of sampling of each subplot was randomly assigned for each experimental plot. The subplots used for biomass extraction for each sample period were also randomly chosen.  2.5 m  Destructive biomass sampling subplots Ion exchange and soil data subplots  2.5 m  Permanent point frame subplot CO2 flux collar and point frame subplot CO2 flux collar buffer area  Point frame measurements were taken from two 0.25 m2 subplots that were designated for longterm monitoring of vegetation cover within each experimental plot (Figure 2.4). One hundred systematically located points per subplot were sampled (Figure 2.4). At each point, the entire canopy was measured by recording all species intercepted from the topmost layer to bare ground. The species, plant part (leaves, flowers, stems), and whether tissues were living or dead were recorded for each vascular plant. In most subplots, there was a multi-layer canopy structure which resulted in multiple counts at each point. For example, a recording at one point was: Vaccinium vitis-idea live leaf, Vaccinium vitis-idea dead leaf, Vaccinium vitis-idea live stem, Cladina rangiferina, Vaccinium vitis-idea live leaf, litter. These data were used to assess species composition (identified to the species level) and canopy cover (number of counts per subplot) to determine if biodiversity changes or compensatory regrowth occurred after grazing. Aboveground biomass of vascular plants, mosses, and lichens was destructively sampled from randomly selected subplots throughout the experiment. One subplot per plot was sampled before and after grazing treatments were applied at the beginning of each growing season (n = 2 periods x 3 years = 6 subplots), at peak biomass accumulation in late July (n = 3 subplots) and at senescence in mid-August (n = 3 subplots); however, no harvests were made at the peak biomass period in 2005 (total = 12 -1 = 11 subplots). For the biomass removal, plants were clipped with scissors, and lichens were removed from the ground surface with a fork. The biomass samples that were clipped from each plot were dried in a 30°C oven for three days to preserve the samples for sorting. The vascular plants were sorted by species and then separated into total live biomass, summation of the leaves, shoots, and reproductive parts, and dead standing crop. After samples were sorted, they were dried again to remove all water content before weighing then 21  weighed to the nearest 0.001 g. Separating the biomass into components allowed for the examination of leaf: shoot per species.  2.2.4  Statistical Analysis  Ratios of leaf biomass to shoot biomass were calculated as: [2.2.4.1]  Leaf: shoot =  leaf standing crop shoot standing crop  These ratios were calculated from the weights of the dried sorted biomass in order to determine whether growth form changed over time with the grazing treatments for the three dominant shrub species: Ledum decumbens, Vaccinium vitis-idaea, and Vaccinium uliginosum. Shannon-Weaver Diversity Indices (H) were also calculated as (Krohne 1998): n  [2.2.4.2]  Heterogeneity (H) = − ∑ (pilnpi) i =1  where p is the proportion of the ith species and n is the total number of species. Data analyses were performed using statistical packages JMP 4.0 (SAS Institute Inc., 2000) and R version 2.12.1 (R Development Core Team, 2010). The analyses used for the data were linear mixed-effects models. For biomass and leaf-to-shoot ratios, preliminary analysis showed there was no significant difference between sampling years (2004, 2005, and 2006), so the effect of year is not included in this model. The full model was the sum of fixed treatment effects (grazing and lichen) and random effects: Response  = Effect of grazing + Effect of lichen + Random effect due to variation of plots + Random effect of period within year + Random effect due to temporal variation  Temporal variation in the model included the between-years variation due to repeated sampling in three years and the between-periods variation due to repeated sampling in three periods within the same year. For the species counts from the point frames and Shannon-Weaver Diversity Index, which were only sampled once per annum, the model was: 22  Response  = Effect of grazing + Effect of lichen + Random effect due to variation of plots + Random effect of year  The likelihood ratio test was used to test the significance of the fixed effects of interest (grazing and lichen). This test was conducted by fitting the data to a reduced model with only the random effects, and then comparing this reduced model to the full model using the likelihood ratio test. Tukey tests for multiple comparisons were used when grazing was a significant factor.  2.3  Results  Overall, analyses showed few significant differences due to the grazing treatments on any of the measured variables, and there were no lichen by grazing interactions. Conversely, effects of the lichen removal were detected.  2.3.1  Climate Data  Rainfall was higher in June 2006 than any other month, but all other data were within the decadal means (Table 2.1). The annual and July mean temperature and the number of days above 0ºC were very similar between years, with 2004 having fewer days above 0ºC than 2006 (Table 2.1).  23  Table 2.1 Lake, NWT.  Months  April May June July August September October Total  Seasonal rainfall and temperature summaries from 1996-2006a for Daring  Rainfall (mm) for June – August Mean 2004 2005 2006 (1996 2006) 29.3 16.4 35.4 94.9 31.0 17.8 11.2b 30.2 49.0 37 n/a 35.3 109.3 71.2 46.6c 160.4  Number of Days with Rain Mean 2004 2005 2006 (19962006) 8.1 11 11 9 10.9 8 11b 14 15.9 18 n/a 13 34.9 37 22c 36  Notes: a data taken from an Indian and Northern Affairs Canada weather station approximately 1.5 km from experimental plots b  data logger malfunction from July 27, 2005 – February 13, 2006  c  total is calculated using incomplete data but was likely 31.  Number of Days Above 0ºC Mean 2004 2005 2006 (1996 2006) 0.9 0 3 1 9.3 1 4 28 28.9 30 30 30 b 31 31 26 31 31 31 n/a 31 25.3 19 n/a 29 3.1 0 n/a 4 129.6 112 63c 154  24  2.3.2  Biomass  2.3.2.1 Vascular Plants Empetrum nigrum occurred rarely and sporadically in the samples, yet it contributed a disproportionately large amount of biomass when it did occur. Empetrum is not a preferred forage type by caribou; therefore, it was removed from the data set. No significant treatment effects were detected in live biomass during any of the sample periods (Table 2.2)(Figure 2.4). Dead standing crop was significantly higher in the plots with lichens intact than in plots where lichens were removed (Table 2.2)(Figure 2.5).  Table 2.2 Results of the likelihood ratio tests with grazing and lichen treatments as fixed factors and temporal and period of time as random factors for live and dead vascular plant biomass (g m-2); n=3-5 plots per treatment. Significant p-values for treatment effects are shown in bold.  Biomass Measurement  Live  Dead  Effect Grazing Lichen Grazing x Lichen Period Temporal Variation Grazing Lichen Grazing x Lichen Period Temporal Variation  Test Statistic (χ2-distributed) 2.330 0.700  df  p-value  2 1  0.312 0.403  2.492  2  0.288  7.490  5  0.187  44.506  1  2.166 x 10-10  2.659 16.455  2 1  0.265 4.982 x 10-5  1.303  2  0.521  4.467  5  0.484  10.810  1  4.494 x 10-3  25  Figure 2.5 Mean (± SE) standing crop (g/m2) of live vascular plants in all treatments (CL, CN, ML, MN, HL, HN; see text for details) during pregrazing, postgrazing, peak standing crop and senescent periods in 2004, 2005 and 2006 except peak 2005; n=3-5.  Live Plant Standing Crop (gm-2)  a) Pregrazing  b) Postgrazing  450  450  400  2004 2005 2006  350  400  300  300  250  250  200  200  150  150  100  100  50  50  0  0 CL CN ML MN HL HN  CL CN ML MN HL HN  Live Plant Standing Crop (gm-2)  c) Peak 450 400  2004 2005 2006  350  d) Senesence 2004 2006  450 400  350  350  300  300  250  250  200  200  150  150  100  100  50  50  0  0 CL CN ML MN HL HN  2004  CL CN ML MN HL HN  26  Figure 2.6 Mean (± SE) standing crop (g/m2) of dead vascular plant standing crop 2 (g/m ) in all treatments (CL, CN, ML, MN, HL, HN) during pregrazing, postgrazing, peak standing crop and senescent periods in 2004, 2005 and 2006 except peak 2005; n=3-5. The letters above the bars indicate results of a posteriori analysis: means sharing the same letter do not differ significantly.  Dead Plant Standing Crop (gm-2)  a) Pregrazing  b) Postgrazing  150  150 2004 2005 2006  100  a  a  2004 2005 2006  100  b  a  50  b  a  50  a b  0  b  a b  0 CL CN ML MN HL HN  CL CN ML MN HL HN  c) Peak Dead Plant Standing Crop (gm-2)  b  d) Senesence  150  150 2004 2006  100  2004  100 a  50  b  a  b  50  a b  0  a  a b  a  b  b  0 CL CN ML MN HL HN  CL CN ML MN HL HN  27  2.3.2.2 Lichens Lichen biomass was significantly different among the treatments, where lichen controls had significantly more lichen biomass than the lichen removal treatments (Table 2.3)(Figure 2.6).  Table 2.3 Results of the likelihood ratio tests with grazing and lichen treatments as fixed factors and temporal and period of time as random factors for lichen, mosses biomass and litter standing crop (g m-2); n=3-5 plots per treatment. . Significant p-values for treatment effects are shown in bold.  Biomass Measurement  Lichens  Mosses  Litter  Effect Grazing Lichen Grazing x Lichen Period Temporal Variation Grazing Lichen Grazing x Lichen Period Temporal Variation Grazing Lichen Grazing x Lichen Period Temporal Variation  Test Statistic (χ2distributed) 0.902 51.935 1.047 6.921 19.547 0.629 0.537 0.003 48.585 19.512 4.907 19.265 0.155 0.404 40.355  df  p-value  2 1 2 5 1 2 1 2 5 1 2 1 2 5 1  0.637 5.736 x 10-13 0.593 0.227 5.693 x 10-3 0.730 0.464 0.999 2.698 x 10-9 5.794 x 10-5 0.086 1.138 x 10-5 0.925 0.995 1.726 x 10-9  28  Figure 2.7 Mean (± SE) lichen standing crop (g/m2) from all treatments (CL, CN, ML, MN, HL, HN; see text for descriptions) during pregrazing, postgrazing, peak standing crop and senescent periods in 2004, 2005 and 2006 except peak 2005; n=3-5. The letters above the bars indicate results of a posteriori analysis: means sharing the same letter do not differ significantly. a) Pregrazing  b) Postgrazing  Lichen Standing Crop (gm-2)  300 250  300 a  200  2004 2005 2006  b  a  a  200  b  150  150  a  a  a 100  100 b  50 0  CL CN ML MN HL HN d) Senesence 300  2004 2006  250 200  2004  250 200  a a  150  a  a  a a  100  100 b  50  b  b  0  300  150  b  50  CL CN ML MN HL HN c) Peak  Lichen Standing Crop (gm-2)  2004 2005 2006  250  b  b  b  0  50  b  b  0 CL CN ML MN HL HN  CL CN ML MN HL HN  2.3.2.3 Mosses There were no treatment effects on moss biomass (Table 2.3)(Figure 2.7).  29  Figure 2.8 Mean (± SE) moss standing crop (g/m2) from all treatments (CL, CN, ML, MN, HL, HN; see text for descriptions) during pregrazing, postgrazing, peak standing crop and senescent periods in 2004, 2005 and 2006 except peak 2005; n=3-5.  a) Pregrazing  b) Postgrazing  Mosses Standing Crop (gm-2)  500  500 2004 2005 2006  400  400  300  300  200  200  100  100  0  0 CL CN ML MN HL HN c) Peak  Mosses Standing Crop (gm-2)  500  2004 2005 2006  CL CN ML MN HL HN d) Senesence 500  2004 2006  2004  400  400  300  300  200  200  100  100  0  0 CL CN ML MN HL HN  CL CN ML MN HL HN  2.3.2.4 Litter The lichen controls had significantly more litter than the treatments where lichens were removed (Table 2.3)(Figure 2.8).  30  Figure 2.9 Mean (± SE) litter standing crop (g/m2) from all treatments (CL, CN, ML, MN, HL, HN; see text for descriptions) during pregrazing, postgrazing, peak standing crop and senescent periods in 2004, 2005 and 2006 except peak 2005; n=3-5. The letters above the bars indicate results of a posteriori analysis: means sharing the same letter do not differ significantly.  a) Pregrazing  b) Postgrazing  800  800  Litter Standing Crop (gm-2)  a  2004 2005 2006  600 b 400  600  b a  a  b  400 b  a  200  200  0  0 CL CN ML MN HL HN c) Peak  800 Litter Standing Crop (gm-2)  a  2004 2005 2006  b  b a  CL CN ML MN HL HN d) Senesence 800  2004 2006  2004  600  600  400  400 a  a 200  b  a b b  0  200  a  a b  b  a b  0 CL CN ML MN HL HN  CL CN ML MN HL HN  2.3.2.5 Leaf- to-Shoot Ratio There were significant effects of grazing on the leaf-to-shoot ratios of L. decumbens (Table 2.4)(Figure 2.9) and V. vitis-idaea (Table 2.4)(Figure 2.10). The control grazing treatments had significantly higher leaf-to-shoot ratios than the medium and high intensity grazing treatments. The leaf-to-shoot ratios of V. uliginosum were not significantly different between treatments (Table 2.4)(Figure 2.11). 31  Table 2.4 Results of the likelihood ratio tests with grazing and lichen treatments as fixed factors and temporal and period of time as random factors for leaf-to-shoot ratio of three major species occurring in each experimental plot, Ledum decumbens, Vaccinium vitis-idaea, and Vaccinium uliginosum; n=3-5 per treatment. Significant p-values for treatment effects are shown in bold.  Leaf-to-shoot Ratio  L. decumbens  V. vitis-idaea  V. uliginosum  Effect Grazing Lichen Grazing x Lichen Period Temporal Variation Grazing Lichen Grazing x Lichen Period Temporal Variation Grazing Lichen Grazing x Lichen Period Temporal Variation  Test Statistic (χ2distributed) 16.771 0.262 1.569 3.855 3.101 20.553 0.608 0.006 5.695 38.193 0.490 0.004 0.585 25.487 0.131  df  p-value  2 1 2 5 1 2 1 2 5 1 2 1 2 5 1  2.281 x 10-4 0.609 0.456 0.570 0.212 2.443 x 10-5 0.436 0.997 0.337 5.086 x 10-9 0.783 0.948 0.747 1.122 x 10-4 0.937  32  Figure 2.10 Mean (± SE) leaf-to-shoot ratios of Ledum decumbens from all treatments (CL, CN, ML, MN, HL, HN; see text for descriptions) during pregrazing, postgrazing, peak standing crop and senescent periods in 2004, 2005 and 2006 except peak 2005; n=3-5. The letters above the bars indicate results of a posteriori analysis: means sharing the same letter do not differ significantly.  a) Pregrazing  b) Postgrazing  0.7 Leaf to Shoot Ratio  0.6  0.7 a a  0.5  b  0.4  2004 2005 2006  0.6  b  0.4  0.5 b  0.3  b  0.2  0.1  0.1  0.7 Leaf to Shoot Ratio  0.3  b b  b  CL CN ML MN HL HN d) Senesence 0.7  0.6  0.4  b  0.0 CL CN ML MN HL HN c) Peak  0.5  a  0.3  0.2  0.0  2004 2005 2006  a  2004 2006  a  0.6 0.5  a b  b  b b  0.4 0.3  0.2  0.2  0.1  0.1  0.0  a  2004  a b b b  b  0.0 CL CN ML MN HL HN  CL CN ML MN HL HN  33  Figure 2.11 Mean (± SE) leaf-to-shoot ratios of Vaccinium vitis-idaea from all treatments (CL, CN, ML, MN, HL, HN; see text for descriptions) during pregrazing, postgrazing, peak standing crop and senescent periods in 2004, 2005 and 2006 except peak 2005; n=3-5. The letters above the bars indicate results of a posteriori analysis: means sharing the same letter do not differ significantly.  a) Pregrazing  Leaf to Shoot Ratio  4  b) Postgrazing 4  a 2004 2005 2006  3 a b  2  2004 2005 2006  b 3  b  a  b  a  b b  2  1  1  0  0 CL CN ML MN HL HN c) Peak  4  b b  CL CN ML MN HL HN d) Senesence 4  Leaf to Shoot Ratio  2004 2006  3  3 a  a a  a b  2  2004  b  b  b  1  2  b b  b  b  1  0  0 CL CN ML MN HL HN  CL CN ML MN HL HN  34  Figure 2.12 Mean (± SE) leaf-to-shoot ratios of Vaccinium uliginosum from all treatments (CL, CN, ML, MN, HL, HN; see text for descriptions) during pregrazing, postgrazing, peak standing crop and senescent periods in 2004, 2005 and 2006 except peak 2005; n=3-5.  a) Pregrazing  b) Postgrazing  0.7  0.7 2004 2005 2006  Leaf to Shoot Ratio  0.6 0.5  0.6 0.5  0.4  0.4  0.3  0.3  0.2  0.2  0.1  0.1  0.0  0.0 CL CN ML MN HL HN c) Peak  0.7 Leaf to Shoot Ratio  CL CN ML MN HL HN d) Senesence 0.7  2004 2006  0.6  0.6  0.5  0.5  0.4  0.4  0.3  0.3  0.2  0.2  0.1  0.1  0.0  0.0 CL CN ML MN HL HN  2.3.3  2004 2005 2006  2004  CL CN ML MN HL HN  Point Cover  2.3.3.1 Point Frame Hits per Plot The cover of live vascular plants was not significantly different among treatments, but total cover of dead vascular plants was significantly higher in the lichen control plots than in the lichen removal plots (Table 2.5)(Figures 2.12a & 2.13a). Also, there was more lichen cover in 35  the CL, ML, and HL treatments than any of the lichen removal treatments (CN, HN, MN) (Table 2.5)(Figure 2.14a). The counts for live, dead, and lichen standing crop were significantly higher in 2004 than 2005 and 2006 (Table 2.5)(Figures 2.12a, 2.13a, & 2.14a).  Table 2.5 Results of the likelihood ratio tests with grazing and lichen treatments as fixed factors and temporal and period of time as random factors for total number of “hits” per 100 points per plot from point frame data for live, dead, and lichen standing crop; n=3-5 per treatment. Significant p-values for treatment effects are shown in bold.  Counts of occurrence of species Live  Dead  Lichen  Effect  Test Statistic (χ2distributed)  df  p-value  Grazing Lichen Grazing x Lichen Year Grazing Lichen Grazing x Lichen Year Grazing Lichen Grazing x Lichen Year  4.919 0.239 1.777 116.854 4.923 5.933 2.269 135.011 2.688 77.16 3.097 98.071  2 1 2 7 2 1 2 7 2 1 2 7  0.085 0.625 0.411 < 2.2 x 10-16 0.106 0.015 0.322 < 2.2 x 10-16 0.261 < 2.2 x 10-16 0.193 < 2.2 x 10-16  36  Figure 2.13 Number of counts per 100 points of the point frame and Shannon-Weaver diversity indices (H) for live vascular plants in all treatments (CL, CN, ML, MN, HL, HN; see text for descriptions) during peak standing crop in 2004, 2005 and 2006. Data are means with SE bars, n=5.  a) Vascular Live Counts 2004 2005 2006  200 175 150 125 100 75 50  2.25  2004 2005 2006  2.00 Diversity Index (H)  Number of Hits Per Plot  225  b) Diversity Index of Vascular Live  25  1.75 1.50 1.25 1.00 0.75 0.50 0.25  0  0.00 CL CN ML MN HL HN  CL CN ML MN HL HN  Figure 2.14 Number of counts per 100 points of the point frame and Shannon-Weaver diversity indices (H) for dead vascular plants in all treatments (CL, CN, ML, MN, HL, HN; see text for descriptions) during peak standing crop in 2004, 2005 and 2006. Intact dead leaves and shoots that were still rooted in the ground were counted. Data are means with SE bars, n=5. The letters above the bars indicate results of a posteriori analysis: means sharing the same letter do not differ significantly.  a) Vascular Dead Counts 2004 2005 2006  200 175 150 125 100 75 50  a  a b  a b  b  2.25  2004 2005 2006  2.00 Diversity Index (H)  Number of Hits Per Plot  225  b) Diversity Index of Vascular Dead  1.75 1.50 1.25 1.00  a  a  b  a  b  b  0.75 0.50 0.25  25  0.00  0 CL CN ML MN HL HN  CL CN ML MN HL HN  37  Figure 2.15 a) Number of counts per 100 points of the point frame and b) ShannonWeaver diversity indices (H) for lichens in all treatments (CL, CN, ML, MN, HL, HN; see text for descriptions) during peak standing crop in 2004, 2005 and 2006. Data are means with SE bars, n=5. The letters above the bars indicate results of a posteriori analysis: means sharing the same letter do not differ significantly.  b) Diversity Index of Lichens  a) Lichen Counts 2004 2005 2006  200 175 150  a  a  a  125 100 75 50  b  b  b  25  2.25 2.00 Diversity Index (H)  Number of Hits Per Plot  225  1.75  a a  a  1.50  2004 2005 2006  1.25 1.00 0.75  b  b  b  0.50 0.25  0  0.00 CL CN ML MN HL HN  CL CN ML MN HL HN  2.3.3.2 Shannon-Weaver Diversity Index (H) Based on total hits for individual species in the sub-plots, the Shannon-Weaver Diversity Index (H) was not significantly different between treatments for the live vascular plants. However, there was significantly higher diversity of dead standing crop in treatments with lichens left intact (CL, ML, HL) compared to those where lichens were removed (CN, MN, HN) in the dead vascular category in 2006 (Table 2.6)(Figures 2.12b & 2.13b). As would be expected, the lichen removal treatments had significantly less diversity than all of the lichen control treatments (Table 2.6)(Figure 2.14b). The Shannon-Weaver Diversity Index for live, dead, and lichen standing crop were significantly higher in 2004 than 2005 and 2006 (Table 2.6)(Figures 2.12b, 2.13b, & 2.14b).  38  Table 2.6 Results of the likelihood ratio tests with grazing and lichen treatments as fixed factors and temporal and period of time as random factors for Shannon-Weaver Diversity Index (H) calculated from point-frame data for live, dead, and lichen standing crop; n=3-5 per treatment. Significant p-values for treatment effects are shown in bold.  ShannonWeaver Diversity Index Live  Dead  Lichen  2.4  Effect  Test Statistic (χ2distributed)  df  p-value  Grazing Lichen Grazing x Lichen Year Grazing Lichen Grazing x Lichen Year Grazing Lichen Grazing x Lichen Year  2.765 0.000027 4.069 85.916 3.919 13.309 2.502 86.272 1.358 35.014 0.108 69.722  2 1 2 7 2 1 2 7 2 1 2 7  0.251 0.996 0.131 8.512 x 10-16 0.141 2.641 x 10-4 0.286 7.197 x 10-16 0.507 3.273 x 10-9 0.948 1.682 x 10-12  Discussion  Overall, results show that the simulated grazing treatments did not alter the vegetative biomass or species composition. However, the lichen removal treatment significantly reduced lichen biomass where the treatment was applied, but the grazing treatments combined with the lichen treatments did not alter the vegetative biomass or composition. Therefore, lichens undercompensated in regrowth, while vascular plants appeared to compensate and return the lost biomass. The lichen removal treatment was detected likely because lichens are slow growing and take years to recover from intense defoliation (Henry & Gunn 1990; Jefferies et al. 1992). The experimental treatments with lichens left intact had greater dead standing crop than those where lichens were removed. This pattern was consistent for the diversity, cover, and litter standing crop. The loss of litter and dead plant crop could be due to their inadvertent removal while removing lichens for the lichen removal treatments, though caution was taken to prevent this. Leaf-to-shoot ratios were higher in ungrazed than highly grazed and medium intensity grazed plots for L. decumbens and V. vitis-idaea. This indicated that, although the grazing applications 39  were not detected by changes in biomass, they were evident in leaf-to-shoot ratios. This suggests that the two grazing treatments removed significant amounts of leaves, but the stem weights dominated the biomass in all treatments. Inter-annual variability in climate did not cause differences in biomass between years. However, species diversity was significantly higher in 2004 than 2005 and 2006 for both vascular plants and lichens. It is possible that the warmer, wetter climate in the latter two sample years could have altered the competitive abilities of plant species in this community, shifting composition to fewer dominant species, such as Ledum decumbens and Vaccinium vitis-idaea. This result is similar to Hudson & Henry (2009) found that increased temperatures caused changes in community composition; specifically, evergreen shrubs became more abundant in High Arctic coastal lowland as a result of climate warming over 13 years. There are several possible reasons why there were no major changes in aboveground biomass or species diversity of the vascular vegetation in this system despite the relatively large disturbance caused by the simulated grazing treatments. One possibility is that there is a high degree of resistance to change and/or resiliency in this ecosystem, which prevented the detection of any experimental results. Golladay et al. (1992) stated that mechanisms of resilience promote recovery, while mechanisms of resistance minimize the effects of disturbance or prevents them from happening. As there are few plant defenses in tundra plants, it appears that mechanisms of resilience may be operating. If the system were resilient to the grazing treatments, changes in biomass values would recover to control values over time, and this is what was observed. Typically, high biodiversity is associated with more resilient and resistant ecosystems, which makes it seem unlikely that the low diversity Arctic systems would be resistant or resilient to disturbance, as community changes would directly alter ecosystem processes (Allison & Martiny 2008). Chapin et al. (1997) propose that high biodiversity protects the functioning of ecosystems by: 1) increasing the efficiency of resource use, and 2) increasing the chances of including species with strong ecosystem effects. However, it has recently been found that only a few fungal species are needed to attain a functional and morphological variability of soil fungal assemblages in a Mediterranean grassland (Persiani et al. 2008). This concept may apply to the functioning of the plant and soil communities in the Low Arctic tundra ecosystem of this study, as we did not observe any changes in soil condition and quality (O, pers. obs.)(see Chapter 3). Recently, the concept of resistance in Arctic habitats has been reported. Hudson and Henry (2010) found that in evergreen–shrub heath in the high-Arctic tundra, there is plant community40  level resistance to experimental warming, even though it was applied for 15 years. This finding suggests that some unproductive systems can also be resistant to change, while our study suggests that these systems can be resilient to change. Often in ecosystems where grazers and plant communities have co-evolved, grazers maintain the quality and productivity of the vegetation (McNaughton 1979b; McNaughton 1984). This has been observed in other systems, ranging from ungulates in the Serengeti Plains (Wilmherst et al. 1995) to prairie dogs in Central North America (Painter et al. 1993) to geese in the coastal marshes in northern Canada (Hik & Jefferies 1990) and in Europe (van der Waal et al. 1998) and muskoxen in wet sedge tundra in the High Arctic (Elliott & Henry 2011). It is possible that the experimental treatments did not have an effect because the plants in our study system have coevolved with grazers and have sufficient below-ground reserves to recover from defoliation (Chapin 1980). Ouellet et al. (1994) reported that when they simulated caribou grazing, deciduous shrubs and graminoids which are regularly grazed by caribou had a better ability to compensate for tissue loss than evergreen shrubs, which are seldom grazed. This is because the latter allocate more resources to leaves and shoots than roots when compared to the former. Most of our study species were deciduous shrubs or graminoids, but the two species that showed the response in leaf:shoot ratio were both evergreen shrubs. The spatial heterogeneity of the habitat may also have precluded the experimental treatments from having detectable grazing effects. The 0.25 m2 sampling plots may not have been large enough to encompass the spatial variability of plant composition in the area. This could have resulted in the high variation around the mean for each experimental condition, making it difficult to detect differences between control and treatment mean values. Although this is a possibility, there was replication of the smaller plot size (n=5), and species richness did not differ significantly between the 0.25 m2 scale and the 6.25 m2 scale.  2.5  Conclusions  Lichen biomass and diversity, litter standing crop, and leaf-to-shoot ratios of the three major plant species differed when lichen removal treatments were applied, but not when plants were clipped to simulate use by caribou. This confirms that lichens are more sensitive to grazing than vascular plants, as they cannot recover lost tissue as quickly. The lack of detectable changes in vascular plant manipulations as a result of the experiment could be because the ecosystem is 41  resistant to disturbance; specifically, because the experimental manipulations did not sufficiently alter soil conditions or other ecosystem processes (see Chapter 3).  42  3  EFFECTS OF SIMULATED GRAZING ON THE ECOSYSTEM CARBON EXCHANGE OF A LOW-ARCTIC TUNDRA SITE  3.1  Introduction  The study of the net ecosystem exchange of carbon dioxide (CO2) in ecosystems is important in quantifying their carbon (C) source-sink status. Although Arctic tundra covers nearly 30% of Canada and contains 10-20% of the planet’s soil C, the C balance of this system remains poorly quantified and understood (Oechel & Billings 1992; Tarnocai 2009). In these ecosystems, low decomposition rates coupled with low primary productivity have resulted in the accumulation of large amounts of C in the form of soil organic matter because historically, the production of C exceeded its decomposition (Oechel & Billings 1992; Tarnocai 2009). Ninety percent of the C in Arctic tundra ecosystems is stored in soils (Oechel & Billings 1992; Tarnocai 2009). Tundra systems are generally net C sinks, but may shift to net C sources, if perturbed by environmental factors, such as increased air and soil temperature, nitrogen deposition, or changes in grazing pressure (Lund et al. 2010). Grazing could cause changes in the species diversity, primary productivity and the amount of C storage in plant biomass (Belsky 1986, Sjögersten et al. 2008). Due to large C storage capacity, Arctic habitats, including tundra and peatlands, may be particularly sensitive to the effects of grazing, especially when these effects are compounded with climate shifts (Post & Pedersen 2008; ACIA 2005; Chapin 1983). The intensity of grazing affects the nutrient availability in a system, which affects plant productivity (Potter et al. 2001). Consequently, nutrient availability will be important in explaining the net ecosystem exchange of CO2 or net ecosystem production (NEP = gross ecosystem production (GEP) + ecosystem respiration (ER)), at the plant and plot scales. For the purposes of this study, net ecosystem production will refer solely to CO2 flux. However, it should be noted that other C fluxes, such as methane (CH4) exchange and dissolved organic C are also important components of the total annual net C balance but represent a small component in most systems (Lund et al. 2010). Typically NEP and GEP are negative values as CO2 is taken up by the system, while ER is a positive value due to CO2 being released into the air. In Arctic grazing systems, where grazers and plants have the opportunity to co-evolve over time, it can be expected that grazing could alter plant responses such that plants would attempt to 43  compensate for lost tissues by changing their growth rates (Jefferies et al. 1992). This compensatory growth can be negative, exact or positive: if plants recover the same biomass as they lost from grazing, this is exact compensation, while if they fail to recover or produce more biomass than ungrazed plants, this is known as under and over compensation, respectively (Belsky 1986). There have been a few studies that examined the effect of grazing on CO2 fluxes in various tundra habitat types ranging from steppe (Holst et al. 2008) to alpine (Welker et al. 2004a, Schmitt et al. 2010) and Arctic (Welker et al. 2004b, Sjögersten et al. 2008) environments. These studies found that CO2 efflux was greater in grazed than ungrazed areas. Because of the limited number of studies on grazing and CO2 fluxes done to-date, there is a need for similar studies in the low-Arctic. The objective of this study was to assess the changes in the net ecosystem production of a lowArctic birch hummock and heath lichen habitat in response to simulated caribou grazing. Lichens as well as aboveground vegetation were manipulated, as these are the primary food sources for caribou in the spring and summer, respectively. Specifically, plant community changes, such as species composition and density, as well as the presence or absence of lichens, are expected to affect the C balance of this ecosystem. If equal or over-compensatory growth occurs following grazing, it is expected that net CO2 uptake will be as high or higher (more negative NEP) than in control plots due to increased photosynthesis. If there is no compensatory growth, then net CO2 uptake may be lower (more positive NEP) potentially as a result of reduced photosynthesis. Further, if simulated grazing decreases both photosynthesis and plant respiration rates while belowground respiration remains the same, then net CO2 uptake will again be lower with the potential for a net release of CO2 from the system.  3.2  Methods  3.2.1  Study Area and Design  The purpose of this experiment was to assess plant responses under simulated grazing that is similar to natural conditions. Important caribou forage species, such as lichens, herbaceous plants, graminoids (e.g. Eriophorum spp. and Carex spp.) and deciduous shrubs (e.g. Salix spp. and Betula spp.) were manipulated to emulate different grazing intensities (Ouellet et al. 1994). Thirty 6.25 m2 experimental plots were randomly established on a North facing esker at Yamba Lake, Northwest Territories (64°52’N, 111°33’W) near the Daring Lake Tundra Ecological 44  Research Station, which is a government-run research facility (Figure 3.1). These plots were placed in birch hummock - heath lichen tundra habitat (vegetation classification by Obst 2008, see Chapter 2). Twenty-five 0.25 m2 subplots were partitioned within each experimental plot, and different subplots were randomly assigned to each sample period, allowing for repeated destructive sampling at the plot scale. Aluminum collars were used to prevent gas exchange at the soil-air interface during CO2 flux measurements. The collars (0.25 m2) were placed 5-10 cm deep in the ground in the middle of four of the 0.25 m2 subplots to prevent edge effects from the destructive sampling (see Chapter 2). The subplots with the collars were randomly assigned within the experimental plots. The experimental manipulations included simulated grazing of the vascular vegetation (leaves) using scissors and removal of lichens with forks from the plots. There were: three levels of grazing (C = 0%, M = 25%, and H = 75% ground cover removed) and presence (L) or absence (N) of lichens. These treatments (n=6) are identified by their combinations of grazing and lichen removal: CL, CN, ML, MN, HL, HN. The treatments were applied to the entire experimental plot (6.25 m2) in the early growing season of each study year.  Figure 3.1 Map of Daring Lake, NWT (64°52’N, 111°3 7’W)(from Natural Resources, Canada 2002).  45  3.2.2 Chamber-Based Methods of Measuring CO2 Flux Chamber based methods for measuring CO2 flux can be used to assess fluxes for small plots or for individual components of an ecosystem (e.g. belowground respiration, foliage respiration or NEP). The rate of change in CO2 within a chamber is measured over a period of time and the CO2 flux is proportional to this rate of change (Drewitt et al. 2002). This method is well suited for controlled field experiments where measurements need to be taken in the same place over time (Davidson et al. 2002). In this study, NEP and ER rates from the plots were measured with a closed-chamber nonsteadystate portable infrared gas analyzer (IRGA) system (see Livingston & Hutchinson 1995; Humphreys 2004). Our system was built by the Biometeorology and Soil Physics Group of UBC. The system consisted of an infrared CO2 and H2O gas analyzer (IRGA) (model LI–840, LI-COR, Lincoln, NE), a pump to circulate air from the chamber to IRGA, temperature sensors inside and outside of the chamber, and a solar radiation sensor mounted outside of the chamber and leveled for measuring (model LI-190 quantum sensor, LI-COR, Lincoln, NE). This system simultaneously measured and recorded CO2 and H2O concentrations, temperature, andphotosynthetically active radiation on a datalogger at 1 second intervals (model CR-21X, Campbell Scientific, Inc., Logan, UT). During a measurement, a 50 cm x 50 cm x 30 cm transparent acrylic chamber with weather stripping along the base was clamped to the aluminum soil collars to create an airtight seal. During NEP measurements, changes in CO2 concentrations were measured over a 30 second interval. Each NEP measurement per collar was made twice with a 30 second break between measurements allowing ambient conditions to be restored. The same procedure was followed for ER measurements, except an opaque cloth was placed over the clear chamber to block the light. Flux measurements began on all plots one week after defoliation and continued throughout the remainder of the growing season from June to September 2004, 2005, and 2006. The frequency of measurements ranged from 2 days to 2 weeks, depending on the weather, as clear skies and low wind conditions are required for optimal measurements.  46  3.2.3  Carbon Dioxide Flux Calculations  Carbon dioxide flux values from the light (NEP) and dark (ER) chamber measurements were calculated as:  Fc =  ρ DVtot dC dry 100 Acham  dt  where Fc is CO2 flux D is dilution = 1 − Cdry =  W , where W is the sample cell [H2O] 1000  C , where C is the sample cell [CO2] D  1000 P ρ is density of the air, where ρ = 8.314(T + 237.15) , P is barometric pressure (measured at the c nearby weather station) Tc is the chamber temperature D = averaged dilution value  Vtot = Vsys + Achamh, where Vtot is the total volume (cm3) Vsys is the system volume Acham is exposed chamber area, h is the height of the chamber base above the ground (the offset) dC dry dt  is the averaged change in Cdry over time (µmol of CO2 mol-1 s-1)  Negative CO2 flux values indicate an uptake of CO2 by the system, while positive CO2 flux values indicate an efflux or release of CO2 to the atmosphere. These data were used to calculate GEP, where GEP = NEP – ER for a given set of NEP and ER measurements per collar. Note that NEP and ER were always measured within a few minutes for any given collar.  47  3.2.4  Environmental Factors  A series of environmental variables were also assessed. Soil moisture (%), was measured in situ (model CS620 Hydrosense® Soil Water Measurement Probe, Campbell Scientific Inc., Logan, UT) by inserting the probe prongs 12 cm into the soil and taking a reading approximately 15 cm next to the aluminum collars. Soil surface temperature (°C) was taken with a digital thermometer at 2.2 cm below the soil surface in the same area just outside of the aluminum collars. These measurements were taken every time chamber measurements were taken. Also, nutrient flux, using ion exchange membranes (IEMs) (model PRS™-probes, Western Ag Innovations, Saskatoon, SK), was measured seasonally. Two cation and two anion IEMs were buried in two randomly selected 0.25 m2 subplots that were not scheduled for harvesting after treatments were applied each spring, and the results from these membranes were pooled to get average nutrient flux values in the 6.25 m2 plots. The IEMs were put in the ground for the entire growing season in each of the three study years, and overwinter from 2004-2005. Each year two cation and two anion IEMs were separated and analyzed as blanks. Concentrations of total nitrogen, NO3-N, NH4-N, Ca, Mg, K, P, Fe, Mn, Cu, Zn, B, S, Pb, and Al were measured. These measurements were used as indicators of ecosystem nutrient availability and serve as supplementary characteristics for examining ecosystem changes over time.  3.2.5  Statistical Analyses  Data analyses were performed using statistical packages R version 2.12.1 (R Development Core Team, 2010) and JMP 4.0 (SAS Institute Inc., 2000). CO2 flux values were categorized into early season (Day 160-195), peak growth (Days 196-219) and late season (Days 220-255). These seasons were selected based on the temporal patterns in CO2 fluxes and roughly correspond to patterns in seasonal weather and phenology with a leafing out or ‘green-up’ period, peak (maximum) aboveground biomass/leaf area period, and early senescence period, respectively. The soil ions, temperature and water content data were analyzed using non-parametric statistics because the data were non-normal and had uneven variances that could not be transformed. For soil temperature and moisture, all measurement days within these seasons were pooled per treatment to increase sample size. The analyses used for the CO2 flux data (ER, NEP and GEP) were linear mixed-effects models. Preliminary analysis showed there was no significant difference between sampling years (2004, 48  2005, and 2006), so the effect of year is not included in this model. The full model was the sum of fixed treatment effects (grazing and lichen) and random effects: Response  = Effect of grazing + Effect of lichen + Random effect due to variation of plots + Random effect of period within year + Random effect due to temporal variation  Temporal variation in the model included the between-years variation due to repeated sampling in three years, the between-periods variation due to repeated sampling in three periods within the same year (early, peak, late), and the between-day variation due to repeated sampling within the same period. The likelihood ratio test was used to test the significance of the fixed effects of interest (grazing and lichen). This test was conducted by fitting the data to a reduced model with only the random effects, and then comparing this reduced model to the full model using the likelihood ratio test.  3.3  Results  3.3.1  Carbon Dioxide Fluxes  There were significant differences in ER, NEP, and GEP between plots and seasons (Table 3.1)(Figure 3.2). The variation between plots represents the heterogeneity of the habitat. The peak and late seasons had more CO2 exchange than the early season, with CO2 exchange rates being greatest during the peak period in each year (Figures 3.3-3.5). The lichen removal treatment had a significant effect on NEP, where the treatments with lichens intact had greater CO2 uptake than the treatments where lichens were removed (Table 3.1). Though there were no significant lichen by grazing interaction effects, there was a general pattern of more CO2 uptake (more negative NEP and GEP) in ungrazed, non-lichen removed (CL) and 25% grazed plots (M) than those which were highly grazed (H) and/or had lichens removed (N) plots (Table 3.1)(Figures 3.3 & 3.4). Grazing and lichen removal caused slightly but not significantly lower ER values in HL, HN and MN plots compared to CL, CN and ML (Table 3.1)(Figure 3.5).  49  Table 3.1 Results of the likelihood ratio tests with grazing and lichen treatments as fixed factors and temporal variation and period of time as random factors for net ecosystem production, gross ecosystem production, and ecosystem respiration rates; n=5 plots per treatment. Significant p-values shown in bold.  Carbon Measurement Net Ecosystem Production  Gross Ecosystem Production  Ecosystem Respiration  Effect Grazing Lichen Grazing x Lichen Period Temporal Variation Grazing Lichen Grazing x Lichen Period Temporal Variation Grazing Lichen Grazing x Lichen Period Temporal Variation  Test Statistic (χ2distributed) 0.071 3.937 3.295 73.532 649.717 0.565 0.525 2.511 82.671 811.199 0.215 0.004 0.734 102.205 1102.208  df  p-value  2 1 2 5 1 2 1  0.965 0.047 0.193 1.883 x 10-14 <2.2 x 10-16 0.754 0.469 0.285 2.316 x 10-16 <2.2 x 10-16 0.898 0.949 0.693 <2.2 x 10-16 <2.2 x 10-16  5 1 2 1 2 5 1  50  Figure 3.2 Seasonal carbon exchange values for 2004, 2005, and 2006. Each data point is the mean flux from the control treatment (CL) during that day with error bars indicating the standard error of the mean; n=30. The same pattern was observed for the grazing and lichen removal treatments (data not shown).  2004  2005  2006  -2  -1  CO2 exchange (µmol m s )  4  2  0  -2  -4  Ecosystem Respiration Net Ecosystem Production Gross Ecosystem Production  -6 210  230  250 160  180  200  220  240 160 180 200 220 240  Day of Year  51  Figure 3.3 Net Ecosystem Production values by treatment for early, peak, and late season of 2004, 2005, and 2006; n=30. The letters above the bars indicate results of a posteriori analysis: means sharing the same letter do not differ significantly. NEP values for all the sample dates are shown in Appendix A.  0.0  -2 -1  CO2 exchange (µmol m s )  2004  CL CN ML MN HL HN  -0.5  b  b  b -1.0  b  a  b  b  a  a a  a  -1.5  a -2.0  Early  Peak  Late  0.0  -2 -1  CO2 exchange (µmol m s )  2005 b -0.5  a  CL CN ML MN HL HN  b  a b a  -1.0  a  b -1.5  b  a  a  b  b a  b  b a  a -2.0  Early  Peak  Late  0.0  -2 -1  CO2 exchange (µmol m s )  2006  CL CN ML MN HL HN  -0.5  b  b b  -1.0  b  a b  -1.5  a  a a  b a  b a  -2.0  Early  b  Peak  a  a  a  b  Late  Season  52  Figure 3.4 Gross Ecosystem Production values by treatment for early, peak, and late season of 2004, 2005, and 2006; n=30. GEP values for all the sample dates are shown in Appendix B. 0  -2 -1  CO2 exchange (µmol m s )  2004  CL CN ML MN HL HN  -1  -2  -3  -4  -5  Early  Peak  Late  0  -2 -1  CO2 exchange (µmol m s )  2005  CL CN ML MN HL HN  -1  -2  -3  -4  -5  Early  Peak  Late  0  -2 -1  CO2 exchange (µmol m s )  2006  CL CN ML MN HL HN  -1  -2  -3  -4  -5  Early  Peak  Late  Season  53  Figure 3.5 Ecosystem Respiration by treatment for early, peak, and late season of 2004, 2005, and 2006; n=30. ER values for all the sample dates are shown in Appendix C.  4  -2 -1  CO2 exchange (µmol m s )  2004  CL CN ML MN HL HN  3  2  1  0  Early  Peak  Late  4  -2 -1  CO2 exchange (µmol m s )  2005  CL CN ML MN HL HN  3  2  1  0  Early  Peak  Late  4  -2 -1  CO2 exchange (µmol m s )  2006  CL CN ML MN HL HN  3  2  1  0  Early  Peak  Late  Season  54  3.3.2  Ion Exchange Membranes  None of the rates of adsorption on the IEMs for the measured nutrients were significantly different between the experimental treatments during any of the sample periods. However, there were significant interannual variations in absorption rates for most of the nutrients, except Cu, Mn, and Pb (Table 3.2). In general, 2004 had lower rates of nutrient adsorption (or availability) than all other sample periods (Table 3.3). The exception to this pattern was with nitrogen. Total nitrogen availability was as low in 2006 as in 2004, while ammonium availability was significantly lower in 2006 than all other sample periods, and nitrate availability was higher in summer 2005 than the other years. Table 3.2 Chi-Square results of the Kruskal-Wallis tests for Ion Exchange Membranes from summer 2004, 2005 and 2006, and overwinter 2004-2005. n=5 for grazing and n=30 for time.  Ion  Treatment  df  Year  Df  Aluminum (Al)  5.57  5  67.88***  3  Boron (B)  6.38  5  42.61***  3  Calcium (Ca)  9.28  5  39.35***  3  Copper (Cu)  5.81  5  1.21  3  Iron (Fe)  7.56  5  51.78***  3  Potassium (K)  5.19  5  41.62***  3  Magnesium (Mg)  5.08  5  32.65***  3  Manganese (Mn)  3.24  5  14.87  3  Ammonium (NH4-N)  2.23  5  39.51***  3  Nitrate (NO3-N)  1.93  5  27.76***  3  Lead (Pb)  0.00  5  0.00  3  Sulphur (S)  4.47  5  16.57***  3  Zinc (Zn)  10.85  5  27.17***  3  Total Nitrogen (N)  1.15  5  31.47***  3  Phosphorus (P)  10.98  5  14.62**  3  *p<0.05, **p<0.01, ***p<0.001  55  Table 3.3 Mean +SEM for soil nutrient flux values (mg/m2/day) measured using ion exchange membranes. n=30. Data are not separated by treatment, as there was no significant treatment effect.  Nutrient Total N NO3--N NH4+-N Ca Mg K P Fe Mn Cu Zn B S Pb Al  3.3.3  Summer 2004 140.0 217.5 120.0 5,930 2,158 1,525 45.0 40.0 65.0 7.5 40.0 5.0 305.0 0 265.0  SE Mean 14.3 12.0 5.37 1964 248 241 7.48 4.35 11.8 2.03 3.75 0.42 22.0 n/a 30.0  Winter 2004-2005 31.9 9.1 27.0 1,131 660.4 699.6 11.6 25.6 15.8 0.7 8.8 0.7 131.2 0 89.8  SE Mean 2.37 1.44 2.05 154 85.4 98.0 2.10 5.85 3.18 0.07 1.02 0.04 33.2 n/a 11.6  Summer 2005 185.5 63.8 121.7 3,961 1,794 1,486 24.6 75.4 34.8 7.2 29.0 7.2 408.7 0 360.9  SE Mean 8.34 6.06 5.91 537 261 143 8.04 27.6 5.02 2.03 3.38 0.66 162 n/a 76.7  Summer 2006 61.3 44.0 42.7 7,553 3,291 2,583 72.0 198.7 86.7 4.0 46.7 4.0 310.7 0 949.3  SE Mean 5.96 5.38 4.06 701 323 220 9.14 63.2 18.0 0.36 6.11 0.37 127 n/a 72.1  Soil Water Content and Temperature  Soil water content was significantly different between treatments, where CN and HN had drier soil than the other treatments at peak 2005, CL and ML were wetter than the other treatments in late 2005, and HL and ML had wetter soils than the other treatments at peak 2006 (Table 3.4)(Figure 3.6). Soil temperatures varied significantly between treatments during peak 2005, with CN and HN plots having warmer soil than the other treatments (Table 3.4)(Figure 3.7). Table 3.4 Chi-Square values for Kruskal-Wallis tests of soil water content (%) at a depth of 12 cm and temperature (°C)at a depth of 2 cm. n=5 for grazing and n=30 for time.  Early χ -Value Water 2004 n/a Content 2005 6.61 2006 7.34 Temperature 2004 n/a 2005 2.19 2006 5.91 *p<0.05, **p<0.01, ***p<0.001 Year  2  df 5 5 5 5  Peak χ -Value 1.70 29.92*** 16.19** 4.70 17.44** 8.05 2  df 5 5 5 5 5 5  Late χ -Value 4.79 13.52* 6.41 3.52 10.72 5.38 2  df 5 5 5 5 5 5  56  Figure 3.6 Soil water content for 2004, 2005, and 2006. Values were measured as % volumetric water content. n=5. Standard errors were too large to be included in the graph and are presented in Appendix D. 60  2004  CL CN ML MN HL HN  50  2005  2006  Soil Moisture (%)  40  30  20  10  0 160  190  220  250  160  190  220  250  160  190  220  250  Day of Year  57  Figure 3.7 Soil temperature (°C) taken at 2 cm belo w the soil surface for 2004, 2005, and 2006. n=5. Standard errors were too large to be included in the graph and are presented in Appendix E.  16 CL CN ML MN HL HN  14  o  Soil Temperature ( C)  12  10  8  6  4  2  2005  2004  2006  0 160  190  220  250  160  190  220  250  160  190  220  250  Day of Year  3.4  Discussion  There were inter-annual variations in the measured variables: soil nutrients, water content, and temperature. These were likely due to the differences in weather between years, with 2004 having lower values of rainfall and temperature than 2005 and 2006 (refer to Chapter 2 for description of the weather data). Drier soil conditions in the control treatments could be due to greater plant water uptake, if there was higher plant biomass in these plots. Also, removing lichens would have altered the albedo of the soil from a more reflective pale colour to the darker brown colour of soil, which absorbs more solar radiation (Brodo et al. 2001). Although significant differences in soil temperature were generally not seen between treatments, lichen removal in the CN and HN treatments reduced soil water content, possibly because the darker surface increased evaporation rates, which could have affected plant productivity and microbial respiration and subsequent CO2 flux. 58  The only significant temperature result was that CN and HN had higher temperatures at peak 2005, probably because of the decrease in albedo. In 2005, CN and HN were both drier, even though the HN treatment should have much lower biomass for water uptake than CN. Because this study did not find significantly higher vascular plant biomass in control compared to grazed plots (see Chapter 2), and since there was not a consistent pattern of increasing soil moisture with decreasing biomass, it is possible that the observed differences in soil moisture were either caused by the lichen removal treatments or spatial variability. Our results differ with those of Sjögersten et al. (2008), Welker et al. (2004a) and Holst et al. (2008) who found that CO2 efflux was greater in grazed than ungrazed areas in Arctic, alpine, and steppe habitats, respectively. Also, Schmitt et al. (2010) found that grazing and mowing increased carbon efflux. Similar to our study, Strebel et al. (2010) found patterns but no significant effect of goose grazing on soil respiration in wet sedge tundra in the high Arctic. Overall, grazing and/or lichen removal treatments did not result in significant changes in the CO2 exchanges of the Daring Lake birch hummock and heath lichen habitat. ER showed no significant treatment effects during any of the sample periods. However, NEP was significantly affected by lichen removal treatments during the productive peak biomass period. Studies have shown that LAI is positively related to NEP during the growing season (Aurela et al. 2009, Lund et al. 2010, Schmitt et al. 2010), so if grazing reduced NEP, it was likely due to lower plot-level  photosynthetic rates due to lower leaf area. The lack of statistical significance in the grazing treatments on CO2 flux and lichen treatments on ER and GEP may be due to rather large spatial variations among plots so that greater sample sizes and perhaps higher productivity would be required to detect differences between treatments. The lack of significant effects of treatments on ER and GEP compared to NEP could be because the roots of the heavily grazed plants were still in the ground and were decomposing and continuing to contribute respiration, such that ER remained relatively unchanged. Our study found there was sufficient soil moisture for decomposition so that ER was similar in all experimental conditions, which is contrary to Nobrega and Grogan (2008), who found ER rates in a dwarf birch habitat during the summer at the same study area to be limited by low soil moisture content. There were no differences between the grazing treatments in any of the CO2 flux values, and the lack of plant responses in C uptake and release support the idea of compensatory growth in plants (Belsky 1986). The CO2 flux values (NEP, GEP and ER) of the 25% grazed treatments 59  (ML) were similar to the controls, showing plants that are moderately grazed try to recover the loss by replacing lost tissue and productivity. Though not significant, the 75% grazed treatments, and occasionally the 25% grazed treatments (MN), were lower in C uptake rates than the controls, which indicates that the highly grazed plants may “under-compensate” for the loss of the biomass. In theory, lichen photosynthesis could have contributed to the significant differences in CO2 fluxes between the removal and control treatments. The ML treatment (removal of 25% of vascular biomass with lichens unaffected) had similar NEP, GEP, and ER results to CL, while the lichen removal treatment with 25% grazing (MN) was similar to the 75% grazed treatment values. For lichens to photosynthesize, they need to have a thallus saturation of 50-70% water (Brodo et al. 2001). When this saturation level is not achieved, lichens are dormant. Typically, air-dried lichens have 15-30% water content (Brodo et al. 2001). When we measured lichen water content, the average values ranged from 10-30% during the 2005 growing season (O, unpublished data). However, it is likely the moisture status of the lichen was quite dynamic with  increases immediately following rainfall and/or dewfall. Prior to saturation, lichen photosynthetic rates are greater than respiration rates, whereas at complete thallus saturation, lichen respiration rates exceed photosynthetic rates (Brodo et al. 2001). These patterns are consistent with our results, as the lichens were not saturated (O, unpublished data). As a result, lichen photosynthesis likely almost always exceeded respiration, when they were not dormant. However, given the fact that the lichens were below the saturation level required for photosynthesis for most of the growing season, they may have been dormant for long periods, leading to few statistically significant results. Significant effects of lichen removal were observed for NEP but not GEP, probably as GEP is a derived variable and any uncertainty or mismatch in NEP and ER could prevent detection of the small GEP that would be attributed to lichen removal. Another possibility for the lack of significant results is that there is a high degree of resistance to change and/or resiliency in this ecosystem. Golladay et al. (1992) state that mechanisms of resilience promote recovery, while mechanisms of resistance minimize the effects of disturbance. If the system was resilient to the grazing treatments, changes in CO2 fluxes would have been detected initially, with these values recovering back to control values over time. This was not observed. Seasonal variation and the natural decline in CO2 exchange through the growing season could have prevented recovery from being observed, but typically, high 60  biodiversity is associated with more resilient and resistant ecosystems (Chapin et al. 1997; Allison & Martiny 2008). This suggests that the low diversity Arctic systems would not be resistant or resilient to disturbance, as community changes would directly alter ecosystem processes (Allison & Martiny 2008). However, Hudson & Henry (2010) concluded that an evergreen heath community in the high Arctic was resistant to long-term experimental warming, and linked their study to similar results in other ecosystems. In addition, it has recently been found that only a few fungal species are needed to attain a functional and morphological variability of soil fungal assemblages in a Mediterranean grassland (Persiani et al. 2008). This concept may apply to the functioning of the plant and soil communities in the Low Arctic tundra ecosystem of this study, as we observed few changes in soil condition and nutrient availability as a result of simulated grazing and/or lichen removal. Another possibility for not observing experimental treatment effects could be that removing plant tissue caused a shift in competitive interactions so that remaining plants were more prolific. Treberg and Turkington (2010) found that per-plant mass was negatively related to density, suggesting biotic density-dependent processes, such as competition and facilitation, are also responsible for structuring plant communities. With less plant cover per area, more resources become available to the remaining plants so that they were no longer resource limited. With the availability of more nutrients and light, remaining plant tissues could maximize photosynthesis and recover biomass more quickly, whether it be increased growth of remaining leaves, growth of new leaves or investment into reproductive tissues.  3.5  Conclusions  Ecosystem CO2 flux was affected by lichen removal, where the treatments with lichens intact had greater CO2 uptake than the treatments where lichens were removed. Though not statistically significant, the control grazing treatments had higher NEP and GEP than highly grazed treatments, with no difference in ER. The medium intensity grazed treatments were more similar to control treatments when lichens were left intact, and more similar to highly grazed treatments when lichens were removed. Soil nutrients and temperature generally had no response to experimental manipulations, while soil water content was reduced in control plots, likely due to plant uptake. Soil water content was also reduced in the plots with high grazing pressure and lichen removal probably because of spatial variability, with some contribution from reduced 61  albedo. As lichens are the primary food source for caribou in the spring and seemed to affect the difference between the exact and under-compensation threshold for CO2 uptake, it would be useful to examine the relationships between lichen biomass and ecosystem CO2 exchange more closely.  62  4  INFLUENCES OF CARIBOU MIGRATION TRAILS ON TUNDRA VEGETATION, SOILS AND CARBON FLUX  4.1  Introduction  Arctic ecosystems are expected to undergo the earliest and most intense changes as a result of global climate change (ACIA 2005), with altered carbon balance resulting from the predicted increases in temperature and changes in precipitation (Oberbauer et al. 2007; Welker et al. 2004b). In fact, changes are already being detected throughout the Arctic that are consistent with predictions of a warming climate (Sturm et al. 2001; Loya & Grogan 2004; ACIA 2005; Hudson & Henry 2009; Hill & Henry 2011). Tussock and wet sedge tundra soils contain most of the carbon stores of Arctic tundra ecosystems because of their large area and productivity, and the relatively low decomposition rates in the cold permafrost soils (Oechel & Billings 1992; Tarnocai 2009). Wet tundra and peatlands can be either sources or sinks for carbon (Blodau 2002; Harazono et al. 2003) but are generally assumed to be net carbon sinks for atmospheric CO2 and have the potential for long-term soil carbon sequestration (Oechel & Billings 1992; Welker et al. 2004b). However, we cannot fully predict the absolute and relative changes in photosynthesis and respiration rates in tundra ecosystems that will result from climate change. Consequently, the extent to which the current Arctic carbon budget will change is unknown. An additional source of uncertainty is the impact of herbivory, which is known to cause changes in the abundance and composition of plant communities (McNaughton 1984; Coughenour 1985) and hence, affect ecosystem processes. Arctic systems are particularly sensitive to the effects of grazing because of their limited ability to recover (Jefferies 1997). Moreover, climate change may exacerbate this vulnerability to herbivores because of changes in plant species diversity and communities (Henry 1998; Post & Pedersen 2008; Elliott & Henry 2011). These effects are of special concern because of the large stores of carbon in Arctic soils (Bliss & Matveyeva, 1992; Welker et al. 2004b; Tarnocai 2009) that can potentially be released under these changing biotic and abiotic conditions. Barren-ground caribou (Rangifer tarandus) are a major herbivore in the Arctic (Henry & Gunn 1990; Jefferies et al. 1992) that forage on a mixture of shrub tundra, sedge-dwarf-shrub tundra and mire vegetation (Bliss & Matveyeva 1992). In these habitats, caribou forage on shrubs, such as Betula nana and Salix spp., and grasses and sedges, such as Hierochloe alpina, Eriophorum 63  vaginatum and Carex spp. Caribou also depend on lichens as a major winter forage, which take up to 15 years to fully regrow after being grazed (Henry & Gunn 1990; Jefferies et al. 1992). While caribou impacts are patchy in the landscape and are temporally episodic, the feeding activity of these herds can have substantial effects at the landscape scale, particularly along migratory trails or in areas of high population densities, such as the calving grounds. For example, the George River Caribou Herd (RGH) in northern Quebec and Labrador grazed some areas very heavily, such that the caribou population was negatively affected (Arseneault et al. 1997). Caribou usually consume between 3 and 50% of net aboveground primary production (NAPP) of wet sedge tundra (Jefferies et al. 1992) although estimates of consumption of other tundra plant communities are poorly known. Caribou populations in the circumpolar Arctic are regulated by different processes in space and time. In general, all barren-ground caribou populations (in North America) are migratory and undergo population cycles that create a host of effects on survival and reproduction, including forage availability and predator numbers. Regardless of the population size of each herd, which can range from hundreds to hundreds of thousands, caribou have managed to exist in the Arctic for thousands of years (Bean 2003; Zalatan et al. 2006). Through trampling and grazing over time, they have created and maintained extensive migratory trail networks that lead them to calving grounds in the spring, then south to boreal forest in the fall, with some trails dating as far back as three hundred years (Zalatan et al. 2006). The impacts by caribou on the low-Arctic landscape have yet to be assessed. It is important to understand these impacts at the present time, as climate change may play a role in shifting the caribou-landscape dynamics in the future (Gough et al. 2008; Sharma et al. 2009). We examined the effects of trampling and grazing by the Bathurst Caribou Herd on the biomass of three low-Arctic plant communities and determined the consequences of trampling and grazing on ecosystem respiration. We predicted that respiration rates would be lower on caribou migratory trails because there is little vegetation as a result of caribou trampling and grazing.  4.2  Methods  4.2.1  Study Area and Design  The study took place in mid-August 2006 in low-Arctic tundra communities around the Tundra Ecological Research Station, Daring Lake, Northwest Territories (64°52’N, 111°37’W). 64  Comparative surveys were conducted at sites along the migratory route of the Bathurst Caribou Herd. The general areas for the sites were visually selected from a helicopter based on trail densities that were representative of the surrounding habitat types. Sites were included based on the following criteria: 1) visually homogeneous conditions within habitat types (same vegetation, similar elevation and aspect); and 2) relatively low disturbance by other grazers. Four sites were selected within 18.5 km (10 nautical miles) of the field station: north (N), east (E), south (S) and west (W). Sampling took place on August 12 and13, 2006 between 0945 and 1700 hrs. The weather was consistent during both sampling days with sunshine, light wind, and an air temperature of approximately 17°C. The regional geomorphology is shield rock and was consistent among the four sites. At each of the sites, three habitat types were sampled along a moisture gradient: wet sedge meadow, mesic birch hummock, and dry esker (Figure 4.1).  65  Figure 4.1 Caribou migratory trails running through a) dry, b) mesic, and c) wet habitats near 64°52’N, 111°37’W in the Northwest Te rritories.  a)  b)  c)  66  4.2.2  Sampling  At each site, samples were taken at five points along one trail that was representative of the area in each habitat, both on, and adjacent to, the caribou migratory trails (n=5 for on and off trails within each habitat type). Transects were started at a point along the trail that was representative of the area, then points along each transect were spaced 3 m apart. Off-trail samples were taken approximately 0.3 m away from the edge of the trail. This distance was optimal, as it was distinctly off of a trail without encroaching on adjacent trails. In some areas where migratory trails occurred in high densities, trails were spaced less than 1 m apart from each other. We were unable to replicate trails due to logistical and financial constraints. We measured plant biomass, ecosystem respiration and soil moisture and temperature at each of these points. Respiration rates (CO2 flux) were measured during daylight hours in a 10 cm2 diameter, 20 cm high opaque chamber using a non-steady state flow through system (see Chapter 3 for details) and a portable infra-red gas analyzer (model LI-840, LI-COR, Lincoln, NE) (see Drewitt et al. 2002). Measurements of ecosystem respiration on and off trails were made by placing the chamber directly on top of whatever matter was present on the ground surface (i.e. soil and vegetation, if present). Shrubs that were too large to fit into the chamber were not sampled, but this situation was rarely encountered. Respiration measurements were made for 120 seconds per sample. To relate vegetation differences to abiotic factors, soil water content was measured using a 12 cm long conductivity probe (model CS620 Hydrosense® Soil Water Measurement Probe, Campbell Scientific Inc., Logan, UT), and soil temperature measurements taken at a depth of 3 cm using a digital thermometer. These measurements were taken simultaneously in the plot immediately after the soil respiration measurements. Following the respiration measurements, plant aboveground biomass was measured by clipping all vegetation above ground level that was sampled using the respiration chamber. All vegetation was dried at 60 °C for 36 hours to preserve the samples. The dried biomass was sorted into living and dead vascular plants, mosses, lichens, or litter. After samples were sorted, they were dried again to remove all water content before weighing then weighed to the nearest 0.001 g.  4.2.3  Statistical Analysis  Data analyses were done using the S-Plus (TIBCO Software Inc., 2002) and JMP 4.0 (SAS Institute Inc., 2000) statistical packages. For each dependent variable, the data were analyzed 67  using three-factor analysis of variance (ANOVA), with on or off the migratory trail, habitat type (esker, birch hummock, sedge meadow), and site (north, east, south or west of Daring Lake) treated as fixed factors. Carbon flux and soil water content data were square-root transformed before analysis to normalize their distributions. When ANOVAs were significant, post hoc Bonferroni analyses were done to determine the relative differences among the treatments. Non-parametric Kruskal-Wallis one-way analysis of variance tests were conducted on the biomass data because they could not be normalized with transformations. Site, habitat and on/off trails were analyzed separately for live biomass, lichen biomass, moss biomass, dead standing crop, and litter. Significant tests were followed with post hoc Mann-Whitney U-tests.  4.3  Results  4.3.1  Abiotic Variables  Soil temperature was statistically different between sites (p<0.00001) and habitats (p<0.00001) (Table 4.1). Soils at sites E and W were significantly warmer than N and S. Also, soils in esker and birch hummock habitats were at least 2.5°C warmer than soils in sedge meadow habitats (Figure 4.2). There was no significant effect (p=0.111) of on/off trail on soil temperature, but there were significant site x habitat (p<0.00001), site x trail (p=0.001), habitat x trail (p=0.026), and site x habitat x trail interactions (p=0.047)(Table 4.1).  Table 4.1 F-probabilities (ANOVA) for the soil water content (%) at a depth of 12 cm and temperature (°C) at a depth of 2 cm measured. S ites were north, east, south, and west of Daring Lake; habitats were dry esker, mesic birch hummock, and wet sedge meadow; trail indicates on or off caribou migratory trails; n=5.  Factor  Soil Temperature (°C) Soil Water Content (%) F Df F df Site 24.35***** 3,96 7.02*** 3,96 Habitat 85.09***** 2,96 383.41***** 2,96 Trail 2.58 1,96 63.91***** 1,96 Site*Habitat 6.35**** 6,96 10.00***** 6,96 Site*Trail 6.40*** 3,96 1.02 3,96 Habitat*Trail 3.81* 2,96 29.49***** 2,96 Site*Habitat*Trail 2.23* 6,96 1.97 6,96 *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001, *****P<0.00001  68  Soil Temperature ( C)  100  o  17  Soil Water Content (%)  Figure 4.2 Mean (+SEM; n=5) soil temperature (°C) a t a depth of 3 cm, and mean (+SEM; n=5) soil water content between 0-12 cm (%) on, and adjacent to, caribou migratory trails in three habitat types: dry esker, mesic birch-hummock, and wet sedge meadows at four sites (north, east, south, west) within 20 km of Daring Lake, Northwest Territories. Sampling took place on August 12-13, 2006 between 0945 and 1700 hrs. Refer to Table 4.1 for the results of the statistical analyses.  15 13 11 9 7  East North South West  80 60 40 20 0  On  Off Dry  Off  On Mesic  Off  On Wet  There were significant effects of all three factors: trail (p<0.00001), habitat (p<0.00001), and site (p<0.00001), on soil moisture (Table 4.1). Sites N and E were significantly wetter than W but not different from S sites. Sedge meadows were significantly wetter than esker and birch hummock sites (Figure 4.2). Soils were significantly wetter on migratory trails than off trails. There were also significant site x habitat (p<0.00001) and habitat x trail (p<0.00001) interactions (Table 4.1).  69  4.3.2  Ecosystem Respiration  Ecosystem respiration (ER) was 21 to 140 % higher (p=0.005) off the migratory trails than on the trails (Table 4.2). The site locations had a significant (p=0.024) effect on respiration rates, with N sites having significantly lower rates than E sites (Table 4.2, Figure 4.3).We also observed differences in the respiration rates of the different habitats, with mesic birch hummocks having significantly higher respiration rates than either dry or wet habitats (Table 4.2, Figure 4.3). There were also significant site x habitat (p<0.00001) and habitat x trail (p=0.045) interactions (Table 4.2).  Table 4.2 F-probabilities (ANOVA) table for ecosystem respiration. Sites were north, east, south, and west of Daring Lake; habitats were dry esker, mesic birch hummock, and wet sedge meadow; trail indicates on or off caribou migratory trails n=5.  Factor  Ecosystem Respiration F Site 3.29** Habitat 10.45*** Trail 8.40** Site*Habitat 7.01**** Site*Trail 0.96 Habitat*Trail 3.20* Site* Habitat*Trail 1.77 *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001, *****P<0.00001  df 3,96 2,96 1,96 6,96 3,96 2,96 6,96  70  Figure 4.3 Mean (+SEM; n=5) ecosystem respiration values (µmol/CO2/m2/s) taken on, and adjacent to, caribou migratory trails in three habitat types: dry esker, mesic birchhummock, and wet sedge meadows; at four sites: north, south, east and west of Daring Lake, Northwest Territories. Sampling took place on August 12-13, 2006 between 0945 and 1700 hrs. Refer to Table 4.2 for the results of the statistical analyses.  Carbon Flux (µmol CO2 / m2 / s)  0.28 East North South West  0.24 0.20 0.16 0.12 0.08 0.04 0.00  on  off Dry  4.3.3  off  on  off  Mesic  on Wet  Plant Community  Caribou grazing and trampling along their migratory trails clearly reduced the amount of live biomass that was present in plots on the trails relative to off the trails (p<0.00001)(Table 4.3, Figure 4.4). Mesic sites had significantly more live biomass than wet sites (p=0.028)(Table 4.3, Figure 4.4). The more abundant vegetation away from trails was also reflected in the significantly higher amounts of dead standing crop at these sites (p<0.00001)(Figure 4.5), as well as higher amounts of plant litter (p<0.00001)(Figure 4.6). The differences for vascular plants were, not surprisingly, also seen in lichen biomass (Figure 4.7). Moss biomass alone did not differ significantly between the plots on or off trails (Table 4.3, Figure 4.8).  71  Table 4.3 Summary of Kruskal-Wallis analysis of live vascular plant, dead vascular plant, lichen, and moss biomass, and litter standing crop. n=5.  Standing Crop Factor Chi-Square Df Live plant Site 0.8901 3 Live plant Habitat 1.9625* 2 Live plant Trail 76.9558**** 1 Dead plant Site 0.9848 3 Dead plant Habitat 17.7071*** 2 Dead plant Trail 50.1016**** 1 Lichens Site 0.9318 3 Lichens Habitat 0.8290 2 Lichens Trail 26.1257**** 1 Mosses Site 25.3870**** 3 Mosses Habitat 5.8689 2 Mosses Trail 2.9242 1 Litter Site 0.3534 3 Litter Habitat 19.6575**** 2 Litter Trail 39.4016**** 1 *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001, *****P<0.00001  72  Figure 4.4 Mean (+SEM; n=5) live aboveground vascular plant biomass (g m-2) taken on, and adjacently to, caribou migratory trails in three habitat types: dry esker, mesic birch-hummock, and wet sedge meadows; at four sites: north, south, east and west of Daring Lake, Northwest Territories. Sampling took place on August 12-13, 2006 between 0945 and 1700 hrs. Refer to Table 4.3 for the results of the statistical analyses. 800 East North South West  Live biomass (g m-2)  600  400  200  0  -200 off  Dry  on  off  Mesic  on  off  Wet  on  73  Figure 4.5 Mean (+SEM; n=5) dead plant standing crop (g m-2) taken on, and adjacent to, caribou migratory trails in three habitat types: dry esker, mesic birch-hummock, and wet sedge meadows; at four sites: north, south, east and west of Daring Lake, Northwest Territories. Sampling took place on August 12-13, 2006 between 0945 and 1700 hrs. Refer to Table 4.3 for the results of the statistical analyses.  East North South West  600  -2  Dead Standing Crop (g m )  800  400  200  0  -200  off  on Dry  off  on Mesic  off  on Wet  74  Figure 4.6 Mean (+SEM; n=5) litter standing crop (g m-2) taken on, and adjacent to, caribou migratory trails in three habitat types: dry esker, mesic birch-hummock, and wet sedge meadows; at four sites: north, south, east and west of Daring Lake, Northwest Territories. Sampling took place on August 12-13, 2006 between 0945 and 1700 hrs. Refer to Table 4.3 for the results of the statistical analyses.  800 East North South West  -2  Litter (g m )  600  400  200  0  -200  off  on Dry  off  on Mesic  off  on Wet  75  Figure 4.7 Mean (+SEM; n=5) lichen biomass (g m-2) taken on, and adjacent to, caribou migratory trails in three habitat types: dry esker, mesic birch-hummock, and wet sedge meadows; at four sites: north, south, east and west of Daring Lake, Northwest Territories. Sampling took place on August 12-13, 2006 between 0945 and 1700 hrs. Refer to Table 4.3 for the results of the statistical analyses.  800  East North South West  -2  Lichens (g m )  600  400  200  0  -200  off  on Dry  off  on Mesic  off  on Wet  76  Figure 4.8 Mean (+SEM; n=5) moss biomass (g m-2) taken on, and adjacent to, caribou migratory trails in three habitat types: dry esker, mesic birch-hummock, and wet sedge meadow at four sites north, south, east and west of Daring Lake, Northwest Territories. Sampling took place on August 12-13, 2006 between 0945 and 1700 hrs. Refer to Table 4.3 for the results of the statistical analyses.  800 East North South West  -2  Mosses (g m )  600  400  200  0  -200  off  on Dry  off  on Mesic  off  on Wet  We also found variation between habitats in terms of standing crop. Sedge meadows had more dead plant biomass (p=0.006) and less litter than either the mesic or esker sites (p<0.00001) (Table 4.3, Figures 4.5 & 4.6). Lichens and litter were less abundant on than off trails (p<0.00001)(Figure4.6 & 4.7). Moss biomass, in contrast, was present in similar amounts among habitats, but varied among the sites with less moss in the east than in the south (p=0.024)(Table 4.3, Figure 4.8).  77  4.4  Discussion  Caribou have created trails that were wetter than the surrounding off-trail areas in all habitats. Despite the wetter conditions on the trails, there was no difference in soil temperatures on or off the trails. However, we found that ecosystem respiration rates were much less on the trails than off in each of the habitats at all sites. Our ecosystem respiration results are consistent with those of Nobrega and Grogan (2008), who also reported that mesic birch hummock had significantly higher respiration rates than either dry or wet habitats, although they also found the latter two habitats as different from each other. As expected, there was little biomass on migratory trails compared to off the trails. This was true for live and dead plant biomass, litter, and lichens. The higher amounts of vegetation off migratory trails accounts for the higher amounts of respiration measured, confirming our predictions. Mosses did not show the same trend, and this is partially attributable to the small sample size (n=5 at each site), although the wetter soils of the trails likely helped to maintain moss cover. The small sample size probably also contributed to the site differences in moss biomass. Taken together, these results show that the compaction and depression of the soil by caribou causes more water to remain at the soil surface, as increased compaction reduces soil porosity, which leads to decreased hydraulic conductivity, and thus increased moisture retention (Singer & Munns 2002). These changes may prevent vegetative reestablishment on migratory trails, especially if the trails are used regularly. So, although caribou herds may alternate which specific trails they use over a span of years (Gunn, pers. comm., Sharma et al. 2009), the trails they create may be maintained for decades or even centuries (Zalatan et al. 2006) unless sufficient time is allowed for vegetation recovery (Figure 4.9). The photosynthetic rate of birch hummock habitat near the survey areas was approximately 1.5 times that of the ecosystem respiration rate in ungrazed areas (see Chapter 3 ) and similar to the rates measured by other studies in the same area (Lafleur & Humphreys 2007; Nobrega & Grogan 2008). This indicates that while soils on the trails continually release CO2, locations off the trail have net ecosystem uptake, storing carbon in plant tissues. Thus, the differences between on and off trails are significant at a local scale. At the landscape scale, the trails contribute to CO2 emissions, but they cover a relatively small area; therefore, the net effect at the landscape scale is continued uptake of carbon through the summer. Trails created by the migration of barren-ground caribou persist for many years (LeResche & Linderman 1975), and the recovery of lichens in heavily used areas is slow (Henry & Gunn 78  1990; Jefferies et al. 1992; Boudreau & Payette 2004). Our results are consistent with those of Boudreau & Payette (2004): grazing and trampling of lichens by large herds of migratory caribou reduce lichen abundance and alter ground vegetation composition. Vistnes and Nelleman (2008) also found that grazing and trampling resulted in more bare ground. However, the increase in bare ground they observed was less than expected because over time the density of mosses increased and changes in plant diversity occurred, thus influencing the forage communities. These changes could be due to changes in litter decomposition or changes in soil nutrient availability, and the decrease in lichen abundance changes invertebrate populations that use lichens (Gough et al. 2008). Thus, through their use of habitat, caribou are key species in Arctic ecosystems. Furthermore, changes in their distribution due to events, such as climate change, will have biological, social and economic implications (Sharma et al. 2009). The warming climate will cause increased soil respiration that could cause vegetative changes (Jackson et al. 2009; Loya & Grogan 2004), and warming and nutrient increases will decrease lichen abundance (Cornelissen et al. 2001), which will be further reduced by herbivory (Gough et al. 2008). This loss of lichen biomass could have impacts on caribou populations, as they rely almost solely on lichens in the fall and winter (Vistnes & Nelleman 2008; Sharma et al. 2009). The added impact of caribou may alter selective areas disproportionately, as climate and physical habitat conditions are the main predictors of caribou migratory trails in all seasons, but temporal shifts in migration route are due to selections they make in a given year (Sharma et al. 2009). For this reason, we can project that the responses we observed at a local scale could have broader implications across a landscape scale.  4.5  Conclusions  This study confirms the hypothesis that trampling and grazing by caribou along their migration trails reduces soil respiration rates because a) there is little vegetation on trails and b) increased water content on the trails can inhibit aerobic respiration. Caribou migratory trails have an impact on low-arctic tundra landscape. The current usage of landscapes by caribou is estimated at between 10 and 50% per year of a given area (O, pers. obs.). This usage may change in the future, as climate change alters arctic landscapes. The interactions between climate change and caribou grazing and their possible synergistic effects on changes in low arctic vegetation composition and diversity need to be further examined. 79  5  GENERAL CONCLUSIONS  5.1  Summary  Caribou have been grazing and trampling low-Arctic tundra landscapes for thousands of years, where large populations of caribou primarily forage on lichens, herbaceous plants, and deciduous shrubs (Bean 2003; Gunn 2005; Zalatan et al. 2006). In these systems where plants and grazers have co-evolved, plants have adapted by attempting to compensate for lost tissue and can exceed, meet, or attain less biomass relative to ungrazed plants (Belsky 1986). Grazing causes changes in individual plant physiology and morphology and can contribute to changes in species composition and interactions, and consequently, ecosystem processes, such as net primary production, nutrient dynamics, and net ecosystem carbon exchange (Catovsky et al. 2002). Studies of caribou plant consumption have focused on calving grounds, not areas along their migratory routes, and the vegetation studies that have been done have used conventional ecological measures, such as plant biomass. While there have been studies of CO2 fluxes in the Canadian high-Arctic (e.g. Welker et al. 2004b) and low-Arctic (e.g. Lafleur & Humphreys 2008), there have not been any studies that examine the effects of grazing on CO2 flux in tundra systems. In the first parts of my study, I experimentally manipulated vegetation in birch hummock/ heath lichen tundra habitat to determine whether simulated caribou grazing causes changes in low-Arctic plant communities and whether these changes affect the CO2 fluxes of this system. I also explored the characteristics of migratory trails relative to the surrounding habitat by surveying vegetation and soil respiration on and off the natural migratory trails that caribou use in the low-Arctic during spring and fall migrations. In the following sections, the original hypotheses of this study will be revisited, and the results will be summarized and synthesized. Hypothesis 1: a) Lichens will not re-grow after being removed; b) the recovery of plants to high intensity defoliation treatments will be lower than those in low intensity treatments; and c) there will be more dramatic differences in the high level treatments in species composition, plant biomass, and net ecosystem exchange compared to the experimental control plots due to the inability of plants to recover from heavy tissue loss (Chapters 2 and 3). Positive compensatory regrowth has been observed in previous studies that simulated caribou grazing, but this was dependent on when during the growing season the treatments were applied 80  and how much vegetation was removed (Ouellet et al. 1994). For this reason, it was expected that the intensity of grazing will affect responses, so that plants experiencing high intensity grazing will compensate less than plants experiencing low intensity grazing. However, in this study, simulated grazing did not affect the aboveground biomass or species diversity of the vascular vegetation in this system. There were higher leaf to shoot ratios in the control plots than the medium and highly grazed plots, showing that there were effects of simulated grazing at the individual species level, even though it was not detected in other variables. As expected, lichen removal significantly reduced both the biomass and diversity of lichens. Plant community changes, such as species composition, diversity, and density, as well as the presence or absence of lichens, will likely affect the carbon balance of this ecosystem. So it is predicted that if compensatory growth occurs following grazing, NEP (net effect of photosynthesis and ecosystem respiration rates) will be as high as or higher than control values. If there is no compensatory growth, then overall NEP will be lower. However, if reduced plant biomass decreases both photosynthesis and respiration rates but soil respiration, which is the major contributor to respiration, remains the same, there will be a net increase in carbon release from the system. In our study, simulated grazing of vascular plants had no effect on CO2 fluxes, but NEP was lower in plots where lichens were removed relative to intact lichen controls. There were no effects of simulated grazing or lichen removal on ecosystem respiration or gross photosynthetic rates. The lower NEP indicates undercompensation by the lichen community in response to removal, and this result was supported by the lower lichen biomass and diversity. Hypothesis 2: Ecosystem respiration rates will be higher off caribou migratory trails because there is little vegetation on trails as a result of caribou trampling and grazing (Chapter 4). Ecosystem respiration rates and biomass were much lower on the trails than off in each of the habitats studied, as expected, due to the low amounts of biomass on migratory trails compared to off the trails. These studies show that the effects of grazing were not easily detected, but the migratory trails that have been used by caribou for thousands of years were distinctly different than the surrounding areas. The results indicate that some habitats may be resistant to change, but once they are altered, they may not readily recover.  81  5.2  Future Directions  One topic of interest is identifying Arctic plant communities that are resistant to change. It is typically believed that because of increased efficiency in resource use and higher chances of including species with strong ecosystem effects, high diversity leads to a higher degree of resistance or resilience to change (Chapin et al. 1997). So systems with lower diversity and productivity, such as Arctic tundra, will be less resistant or be slower to recover from change. Welker et al. (1999) and Allison and Martiny (2008) have found that plant and microbial communities respond quickly to experimental change in Arctic systems. However, Henry and Hudson (2010) recently reported that evergreen-shrub heath in the high-Arctic was unaffected by 15 years of experimental warming. Our experimental results (Chapters 2 and 3) also showed little change from the simulated grazing that was applied for 3 years. So, it is possible that only substantial changes in resource supply or disturbance will affect some unproductive Arctic plant communities. Further research should be conducted to identify which specific communities are sensitive to change and which others are resistant. Also, it is important to assess multi-factorial changes (e.g. climate change and grazing), as these factors may interact synergistically to affect otherwise resistant systems. Recovery from trampling and grazing is another interesting topic. Chapters 2 and 3 showed that even after 3 years of grazing treatments, there was no change in vascular plant biomass or soil temperature, and there was a difference in soil moisture between experimental treatments for one time period. This indicates that short-term grazing had little effect on this ecosystem. Conversely, Chapter 4 revealed that there were significant differences on and off migratory trails in biomass, soil moisture, soil temperature, and ecosystem respiration. This indicates that longterm use by caribou has ecosystem effects. Caribou stay within their home range during migration, but within this range, they often do not return to the same trails in consecutive years, yet evidence of trails remains for many years after their establishment (Bean 2003; Zalatan et al. 2006). So, it would be interesting to determine what timing and intensity of trampling and grazing is required to form migratory trails and once established, how long it would take for these trails to recover back to a pre-trail state. Another question stemming from this research is how climate change scenarios will act in conjunction with grazing on low-Arctic plant communities. Climate change will likely cause increased temperature and unpredicted moisture changes, and will alter melt rates, which can 82  delay or accelerate and extend the growing seasons which will affect CO2 uptake. Specific habitats may be favored grazing areas over others. For example, if climate change causes melt to occur earlier in low-lying sedge valleys, caribou may use these areas more intensely than in the historical past. The grazing in sedge meadows could thus decrease live plant biomass and subsequently CO2 uptake. However, if these sedges experience compensatory growth and match or surpass ungrazed biomass (Belsky 1986; Raillard and Svoboda 1999), there could be an increase in CO2 uptake by grazed plants. If climate change causes the melting and release of belowground stored CO2, then there could be a net CO2 release from the system regardless of the grazing impact, thus further contributing to the input of greenhouse gases at the global scale. Areas that melt earlier and are used most of the year, like esker top areas, may experience heavier grazing, but with warmer summer temperatures, vegetation on this habitat may proliferate. On the contrary, early melt of all habitats would eventually lead to an availability of more varieties of forage, and there may be more rapid recovery in areas that were more heavily grazed and trampled. More diffuse use on any given area could allow vegetation to re-establish on some migratory trails. It would be interesting to further examine grazing impacts on different habitat types in the low-Arctic, along with nutrient addition and snow manipulations that would mimic climate change. The habitat shifts caused by climate change could affect the forage quality and timing of food availability for caribou in their wintering, migratory, and calving grounds. Caribou accumulate almost all of their weight during the summer (Brathen & Oksanen, 2001), so they are sensitive to habitat changes along their migratory routes and in their calving grounds. Habitat changes could alter caribou population dynamics, so more studies on the characteristics of these habitats and their potential changes, including plant nutrition, would be useful in predicting how the population size and health of caribou herds might be affected, and how migratory routes might shift with climate change. Changes in caribou densities could affect the number of trails and thus the CO2 (respiration) rates of the habitat in the areas used by caribou. My survey was the first study to assess on and off migratory trail characteristics, and it would be useful to do the same study across a larger sampling regime. The goal would be to scale these results up to a regional and ultimately circumpolar level.  83  5.3  General Implications  There have been numerous studies that examine grazing, CO2 flux, and climate change independently in Arctic ecosystems. Grazing is anticipated to become an increasingly important topic as climate change continues, and a wider range of processes and possible effects may shift the structure and function of these systems. Grazing may buffer or exacerbate the effects of climate change, and predicting the interactions between these two factors will be important in the conservation and management of Arctic habitats and the organisms that live within these habitats. Measuring CO2 flux has become an invaluable tool in monitoring the Arctic’s carbon source-sink status, and this method is also useful for assessing grazing and/or trampling by herbivores.  84  REFERENCES Adamczewski, J.Z., Gates, C.C., Soutar, B.M., and R.J. Hudson. 1988. Limiting effects of snow on seasonal habitat use and diets of caribou (Rangifer tarandus-groenlandicus) on Coats Island, Northwest Teritories. Canadian Journal of Zoology, 66(9): 1986-1996. Agrawal, A.A. 2000. Overcompensation of plants in response to herbivory and the by-product benefits of mutualism. Trends in Plant Science, 5(7): 309-313. Allison, S.D. and J.B.H. Martiny. 2008. Resistance, resilience, and redundancy in microbial communities. Proceedings of the National Academy of Sciences of the United States of America, 105(suppl 1): 11512-11519. Aurela, M., Lohila, A., Tuovinen, J-P., Hatakka, J., Riutta, T., and T. Laurila. 2009. Carbon dioxide exchange on a northern boreal fen. Boreal Environment Research, 14: 669-710. Bazely, D.R. and R.L. Jefferies. 1985. Goose faeces: a source of nitrogen for plant growth in a grazed salt-marsh. Journal of Applied Ecology, 22: 693-703 Bean, D. 2003. Barren-ground caribou calving grounds: vegetation and implications for change. M.Sc. thesis, University of British Columbia. Beaulieu J., Gauthier, G., and Rochefort, L. 1996. The growth response of graminoid plants to goose grazing in a High Arctic environment. Journal of Ecology, 84: 905-914. Belsky, A.J. 1986. Does herbivory benefit plants? A review of the evidence. The American Naturalist, 127(6): 870-892. Bliss, L.C. and N.V. Matveyeva. 1992. Circumpolar Arctic Vegetation. In Arctic Ecosystems in a Changing Climate. Edited by: Chapin, F.S. III, Jefferies, R.L., Reynolds, J.F., Shaver, G.R., and Svoboda, J. Academic Press, New York, pp.59-89. Blodau, C. 2002. Carbon cycling in peatlands – a review of processes and controls. Environmental Reviews, 10: 111-134. Boudreau, S. and S. Payette. 2004. Growth performance of Cladina stellaris following caribou disturbance in subarctic Quebec. Ecoscience, 11(3): 347-355. Brathen, K.A. and J. Oksanen. 2001. Reindeer reduce biomass of preferred plant species. Journal of Vegetation Science, 12(4): 473-480. Brodo, I.M, Sharnoff, S.D and S. Sharnoff. 2001. Lichens of North America. Yale University Press, New Haven, USA. Pp. 3-61. Bubier, J. L., Moore, T. R., and N.T.Roulet. 1993. Methane emissions from wetlands in the midboreal region of northern Ontario, Canada. Ecology, 74: 2240-2254. Cargill, S.M. and R.L. Jefferies. 1984. The effects of grazing by lesser snow geese on the vegetation of a sub-arctic salt marsh. Journal of Applied Ecology, 21: 669-686. Catovsky, S., Bradford, M.A., and A. Hector. 2002. Biodiversity and ecosystem productivity: implications for carbon storage. Oikos, 97(3): 443-448. Chapin, F. S., III , B.H. Walker, R.J. Hobbs, D.U. Hooper, J.H. Lawton, O.E. Sala, and D. Tilman. 1997. Biotic control over functioning ecosystems. Science, 277(5325): 500-504.  85  Chapin, F. S., III. 1983. Patterns of nutrient absorption and use by plants from natural and manmodified environments. In Disturbance and ecosystems. Components of response. Edited by: H.A. Mooney and M. Godron. Springer-Verlag, Berlin. Pp. 175-187. Chapin, F. S., III. and G.R. Shaver. 1996. Physiological and growth responses of arctic plants to a field experiment simulating climatic change. Ecology, 77(3): 822-840. Chapin, F.S., III. 1980. Nutrient allocation and response to defoliation in tundra plants. Arctic and Alpine Research, 12: 553-563. Cornelissen, J.H.C., Callaghan, T.V., Alatalo, J.M., Michelsen, A., Graglia, E., Hartley, A.E., Hik, D.S., Hobbie, S.E., Press, M.C., Robinson, C.H., Henry, G.H.R., Shaver, G.R., Phoenix, G.K., Jones, D.G., Jonasson ,S., Chapin, F.S., Molau, U., Neill, C,. Lee, J.A., Melillo, J.M., Sveinbjornsson, B., and R. Aerts. 2001. Global change and arctic ecosystems: is lichen decline a function of increases in vascular plant biomass? Journal of Ecology, 89(6): 984-994. Coughenour, M. B. 1985. Graminoid responses to grazing by large herbivores: adaptations, exaptations, and interacting processes. Annals of the Missouri Botanical Gardens, 72: 852863. Crawley, M.J. 1983. Herbivory: the dynamics of plant-animal interactions. In Studies in Ecology, Volume 10. Blackwell Scientific Publications, Oxford. Davidson, D.W. 1993. The effects of herbivory and granivory on terrestrial plant succession. Oikos, 68: 23-35. Davidson, D.W., Bowker, M., George, D., Phillips, S.L., and J. Belnap. 2002. Treatment effects on performance of N-fixing lichens in disturbed soil crusts of the Colorado Plateau. Ecological applications, 12(5): 1391-1405. Dormann, C.F., van der Wal, R., and J.P. Bakker. 2000. Competition and herbivory during salt marsh succession: the importance of forb growth strategy. Journal of Ecology, 88: 571-583. Downes, C.M. 1985. Hilltop aggregation sites and behaviour of male Cephenemyia-Trompe (Diptera, Oestridae). Canadian Entomologist, 117(3): 321-326. Drewitt, G.B., Black, T.A., Nesic, Z., Humphreys, E.R., Jork, E.M., Swanson, R., Ethier, G.J., Griffis, T., and K Morgenstern. 2002. Measuring forest floor CO2 fluxes in a Douglas-fir forest. Agricultural and Forest Meterology, 110(4): 299-317. Elliott, T. and G.H.R. Henry. 2011. Muskox grazing in the Canadian High Arctic. In prep. Flanagan, L.B., Wever, L.A., and P.J. Carlson. 2002. Seasonal and interannual variation in carbon dioxide exchange and carbon balance in a northern temperate grassland. Global Change Biology, 8(7): 599-615. Gauthier, G., Hughes, R.J., Reed, A., Beaulieu, J., and L. Rochefort. 1995. Effect of grazing by greater snow geese on the production of graminoids at an arctic site (Bylot Island, NWT, Canada). Journal of Ecology, 83: 653-664. Gibson, D.J. 1988. The relationship of sheep grazing and soil heterogeneity to plant spatial patterns in dune grassland. Journal of Ecology, 76: 233-252. Griller, N. 2001. Recovery of plants after intensive grazing. M.Sc. thesis, University of British Columbia. 86  Golladay, S.W., Webster, J.R., Benfield, E.F., and W.T. Swank. 2002. Changes in stream stability following forest clearing as indicated by storm nutrient budgets. Archiv fur Hydrobiologie, 1(Suppl 90): 1-33. Gough, L., Shrestha, K., Johnson, D.R., and B. Moon. 2008. Long-term mammalian herbivory and nutrient addition alter lichen community structure in Alaskan dry heath tundra. Arctic, Antarctic, and Alpine Research, 40(1): 65-73. Gunn, A. 2005. Voles, lemmings and caribou – population cycles revisited? Rangifer Special Issue, 14: 105-112. Handa, I.T. and R.L. Jefferies. 2000. Assisted revegetation trials in degraded salt marshes. Journal of Applied Ecology, 37: 944-958. Handa, I.T., Harmsen, R., and R.L. Jefferies. 2002. Patterns of vegetation change and the recovery potential of degraded areas in a coastal marsh system of the Hudson Bay lowlands. Journal of Ecology, 90: 86-99. Harazono Y, Mano M, Miyata A, Zulueta RC, and Oechel WC. 2003. Inter-annual carbon dioxide uptake of a wet sedge tundra ecosystem in the Arctic. TELLUS, 55B: 215-231. Harper, F. 1955. The Barren Ground Caribou of Keewatin. Lawrence, Kansas: The University of Kansas, Miscellaneous Publication No. 6. 240 pgs. Henry, G.H.R. 1998. Environmental influences on the structure of sedge meadows in the Canadian High Arctic. Plant Ecology, 134(1): 119-129. Henry, G.H.R. and A. Gunn. 1991. Recovery of tundra vegetation after overgrazing by caribou in Arctic Canada. Arctic, 44(1): 38-42. Hik, D.S. and R.L. Jefferies. 1990. Increases in the net above-ground primary production of a salt-marsh forage grass: a test of the predictions of the herbivore-optimization model. Journal of Ecology, 78: 180-195. Hill, G.B. and G.H.R. Henry. 2011. Responses of High Arctic wet sedge tundra to climate warming since 1980. Global Change Biology, 17(1): 276-287. Hobbs, N.T. 1996. Modification of ecosystems by ungulates. Journal of Wildlife Management, 60(4): 695-713. Holst J., Liu, C., Yao, Z., Brüggemann, N., Zheng, X., Giese, M., and K. Butterbach-Bahl. 2008. Fluxes of nitrous oxide, methane and carbon dioxide during freezing–thawing cycles in an Inner Mongolian steppe. Plant Soil, 308: 105-117. Hudson, J.M.G. and G.H.R. Henry. 2009. Increased plant biomass in a High Arctic heath community from 1981 to 2008. Ecology, 90(10): 2657–2663 Hudson, J. and G.H.R. Henry. 2010. High Arctic plant community resists 15 years of experimental warming. Journal of Ecology, 98:1035-1041. Humphreys, E.R. 2004. Net ecosystem production of three coastal Douglas-fir stands at different stages of development after harvesting. Ph.D. thesis, University of British Columbia. Jackson, R.B., Cook, C.W., Pippen, J.S., and S.M. Palmer. 2009. Increased belowground biomass and soil CO2 fluxes after a decade of carbon dioxide enrichment in a warmtemperate forest. Ecology, 90(12): 3352-3366. 87  Jefferies, R.L. and R.F. Rockwell. 2002. Foraging geese, vegetation loss and soil degradation in an Arctic salt marsh. Applied Vegetation Science, 5: 7-16. Jefferies, R.L. 1997. Long-term damage to sub-Arctic coastal ecosystems by geese: ecological indicators and measures of ecosystem dysfunction. In Disturbance and Recovery in Arctic Lands. Edited by R.M.M. Crawford. Kluwer Academic Publishers, Netherlands. Pp. 151165. Jefferies, R.L., Svoboda, J., Henry, G., Raillard, M., and R. Ruess. 1992. Tundra Grazing Systems and Climate Change. In Arctic Ecosystems in a Changing Climate. Edited by: Chapin, F.S. III, Jefferies, R.L., Reynolds, J.F., Shaver, G.R., and Svoboda, J. Academic Press, New York, pp.391-408. Krohne, D.T. 1998. General Ecology. Wadsworth Publishing Company, New York. Lafleur, P.M. and E.R. Humphreys. 2007. Spring warming and carbon dioxide exchange over low arctic tundra in central Canada. Global Change Biology, 14: 740-756. Lafleur, P.M. and E.R. Humphreys. 2008. Spring warming and carbon dioxide exchange over low-arctic tundra in central Canada. Global Change Biology, 14: 740-756. Lentz, K.A. and D.F. Cipollini, Jr. 1998. Effects of light and simulated herbivory on growth of endangered notheastern bulrush, Scirpus ancistrochaetus Schuyler. Plant Ecology, 139: 125-131. Leresche, R.E. and S.A. Linderman. 1975. Caribou trail systems in northeastern Alaska. Arctic, 28(1): 54-61. Lesica, P. and S.V. Cooper. 1999. Succession and disturbance in sandhills vegetation: constructing models for managing biological diversity. Conservation Biology, 13(2): 293302. Livingston, G.P. and G.L. Hutchinson. 1995. Enclosure-based measurement of trace gas exchange: Applications and sources of error. In Biogenic Trace Gases: Measuring Emissions from Soil and Water. Edited by P.A. Matson and R.C. Harriss. Blackwell Scientific Publishing, Cambridge, pp. 14-51. Lund, M., LaFleur, P.M., Roulet, N.T., Lindroth, A., Christensen, T.R., Aurelas, M., Bogdan, H., Chojnicki, B.H., Flanaga, L.B., Humphreys, E.R., Laurilas, T., Oechel, W.C., Olejnik, J., Rinne, J., Schubert, P., and M.B. Nilsson. 2010. Variability in exchange of CO2 across 12 northern peatland and tundra sites. Global Change Biology, 16: 2436-2448. Matson, P.A. and R.C. Harriss. 1995. Trace gas exchange in an ecosystem context: multiple approaches to measurement and analysis. In Biogenic Trace Gas Emissions: Measuring Emissions from Soil and Water. Edited by P. A. Matson and R.C. Harriss. Blackwell Scientific Publishing, Cambridge pp. 1-13. Maxwell, B., 1992: Arctic climate: potential for change under global warming. In Arctic Cosystems in a Changing Climate: An Ecophysiological Perspective. Edited by Chapin, E S., III, Jeffries, R. L., Reynolds, J. E, Shaver, G. R., and J. Svoboda. Academic Press, San Diego, 11-34. McEwan, E.H. and P.E. Whitehead. 1970. Seasonal changes in energy and nitrogen intake in reindeer and caribou. Canadian Journal of Zoology, 48(5): 905-913.  88  McIntire, E.J.B. and D.S. Hik. 2002. Grazing history versus current grazing: leaf demography and compensatory growth of three alpine plants in response to a native herbivore (Ochotona collaris). Journal of Ecology, 90: 348-359. McKendrick, J.D. 1981. Responses of arctic tundra to intensive muskox grazing. AgroborealisAlaska-Agric-Exp-Stn-Fairbanks, 13: 49-55. McNaughton, S.J. 1979a. Grazing as an optimization process: grass-ungulate relationships in the Serengeti. The American Naturalist, 113(5): 691-703. McNaughton, S.J. 1979b. Grassland-herbivore dynamics. In Serengeti: dynamics of an ecosystem. Edited by A.R.E. Sinclair and M. Norton-Griffiths. University of Chicago Press, Chicago. pp. 46-81. McNaughton, S.J. 1984. Grazing lawns: animals in herds, plant form, and coevolution. The American Naturalist, 124(6): 937-939. Nobrega, S. and P. Grogan. 2008. Landscape and ecosystem-level controls on net carbon dioxide exchange along a natural moisture gradient in Canadian low arctic tundra. Ecosystems, 11: 377-396. Norman, J.M., Garcia, R., and S.B. Verma. 1992. Soil surface CO2 fluxes and the carbon budget of a grassland. Journal of Geophysical Research, 97: 18845–18853. O, P.C. 2003. Responses of Festuca rubra to natural and simulated foraging by geese on Akimiski Island, Nunavut Territory. M.Sc. thesis, University of Toronto. Oberbauer, S.F., Tweedie, C.E., Welker, J.M., Fahnestock, J.T., Henry, G.H R., Webber, P.J., Hollister, R.D., Walker, M.D., Kuchy, A., Elmore, E., and G. Starr. 2007. Tundra CO2 fluxes in response to experimental warming across latitudinal and moisture gradients. Ecological Monographs, 77(2): 221-238. Oechel, W.C. and W.D. Billings. 1992. Effects of Global Change on the Carbon Balance of Arctic Plants and Ecosystems. In Arctic Ecosystems in a Changing Climate. Edited by: Chapin, F.S. III, Jefferies, R.L., Reynolds, J.F., Shaver, G.R., and Svoboda, J. Academic Press, New York, pp.139-164. Ouellet, J.- P., Boutin, S ., and Heard, D .C. 1994. Responses to simulated grazing and browsing of vegetation available to caribou in the Arctic. Canadian Journal of Zoology, 72: 14261435. Painter, E.L., Detling, J.K., and Steingraeber, D.A., 2003. Plant morphology and grazing history: relationships between native grasses and herbivores. Vegetatio, 106(1): 37-62. Persiani, A.M., Maggi, O., Montalvo, J., Casado, M.A., and F.D. Pineda. 2008. Mediterranean grassland soil fungi: Patterns of biodiversity, functional redundancy and soil carbon storage. Plant Biosystems, 142(1): 111-119. Person, B.T., Babcock, C.A., and R.W. Ruess. 1998. Forage variation in brood-rearing areas used by pacific Brant Geese on the Yukon-Kuskokwim delta, Alaska. Journal of Ecology, 86: 243-259. Post, E.S. and D.R. Klein. 1996. Relationships between graminoid growth form and levels of grazing by caribou (Rangifer tarandus) in Alaska. Oecologia, 107(3): 364-372.  89  Post, E. and C. Pederson. 2008. Opposing plant community responses to warming with and without herbivores. Proceedings of the National Academy of Sciences of the United States of America, 105(34): 12353-12358. Potter, K.N., Daniel, J.A., Altom, W. and H.A. Torbert. 2001. Stocking rate effect on soil carbon and nitrogen in degraded soils. Journal of Soil and Water Conservation, 56(3): 233-236. Raillard, M.C. and J. Svoboda. 1999. Exact growth and increased nitrogen compensation the by arctic sedge Carex aquatilis var. stans after simulated grazing. Arctic, Antarctic, and Alpine Research, 31(1): 21-26. Raillard, M. and J. Svoboda. 2000. High grazing impact, selectivity, and local density of muskoxen in Cental Ellesmere Island, Canadian high Arctic. Arctic, Antarctic, and Alpine Research, 32(3): 278-285. Rosenthal, J.P. and P.M. Kotanen. 1994. Terrestrial plant tolerance to herbivory. Trends in Ecology and Evolution, 9(4): 145-148. Ruess, R.W., Uliassi, D.D., C.P.H. Mulder, and B.T. Person. 1997. Growth responses of Carex ramenskii to defoliation, salinity, and nitrogen availability: implications for geeseecosystem dynamics in Western Alaska. Écoscience, 4: 170-178. Russell, D.E., Griffith, B. and G.P. Kofinas. 2002. Caribou in a changing tundra environment. Integrative and Comparative Biology, 42(6): 1304-1304. Schmitt, M., Bahn, M., Wohlfarht, G., Tappeiner, U., and A. Cernusca. 2010. Land use affects the net ecosystem CO2 exchange and its components in mountain grasslands. Biogeosciences, 7: 2297–2309. Sharma, S., Couturier, S. and S.D. Cote. 2009. Impacts of climate change on the seasonal distribution of migratory caribou. Global Change Biology, 15(10): 2549-2562. Shaver, G.R., Canadell, J., Chapin, F.S., Gurevitch, J., Harte, J., Henry, G., Ineson, P., Jonasson, S., Melillo, J., Pitelka, L. and L. Rustad. 2000. Global warming and terrestrial ecosystems: A conceptual framework for analysis. Bioscience, 50(10): 871-882. Singer, M.J. and D.N. Munns. 2002. Soils: An Introduction, 5th edition. Prentice Hall, Inc., New Jersey, USA. Sjögersten, S., van der Wal, R., and S. Woodin. 2008. Habitat type determines herbivory controls over CO2 fluxes in a warmer arctic. Ecology, 89(8): 2103-2116. Skogland, T. 1984. Wild Reindeer Foraging-Niche Organization. Holarctic Ecology 7(4): 345379. Smith, D.L. 1996. A test of herbivore optimization hypothesis using muskoxen and a graminoid meadow plant community. Rangifer, 16(2): 69-77. Storeheier, P.V., Mathiesen, S.D., Tyler, N.J.C., Schjelderup, I. and M.A. Olsen. 2002. Utilization of nitrogen- and mineral-rich vascular forage plants by reindeer in winter. Journal of Agricultural Science, 139: 151-160. Strebel, D., Eberling, B., Morgner, E., Knicker, H.E., and E.J. Cooper. 2010. Cold-season soil respiration in response to grazing and warming in High-Arctic Svalbard. Polar Research, 29: 46-57.  90  Sturm, M., Racine, C. and K. Tape. 2001 Climate change - Increasing shrub abundance in the Arctic. Nature, 411(6837): 546-547. Tarnocai, C. 2006. The effect of climate change on carbon in Canadian peatlands. Global and Planetary Change, 53: 222–232. Tarnocai, C. 2009. The impact of climate change on Canadian peatlands. Canadian Water Resources Journal, 34(4) Special Issue: 453-466. Tolvanen, A. and G.H.R. Henry. 2000. Population structure of three dominant sedges under muskox herbivory in the High Arctic. Arctic, Antarctic and Alpine Research, 32(4): 449455. Townsend, C.R, Harper, J.L., and M. Begon. 2000. Essentials of Ecology. Blackwell Scientific, Inc., Massachusetts, USA. Treberg, M.A. and R. Turkington. 2010. Density dependence in an experimental boreal forest understory community. Botany, 88: 753–764 Vallentine, J.F. 1990. Grazing Management. Academic Press, San Diego, California. van der Wal, R., van de Koppel, J., and M. Sagel. 1998. On the relation between herbivore foraging efficiency and plant standing crop: an experiment with barnacle geese. Oikos, 82: 123-130. Vistnes, I.I. and C. Nellemann. 2008. Reindeer winter grazing in alpine tundra: Impacts on ridge community composition in Norway. Arctic, Antarctic, and Alpine Research, 40:1 215-224. Vourlitis, G.L., Oechel, W.C., Hastings, S.J., and M.A. Jenkins. 1993. The Effect of SoilMoisture and Thaw Depth on CH4 Flux From Wet Coastal Tundra Ecosystems on the North Slope of Alaska. Chemosphere, 26(1-4): 329-337. Walker, D.A., Jia, G.J., Epstein, H.E., Raynolds, M.K., Chapin, F.S., Copass, C., Hinzman, L.D., Knudson, J.A., Maier, H.A., Michaelson, G.J., Nelson, F., Ping, C.L., Romanovsky, V.E., and N. Shiklomanov. 2003. Vegetation-soil-thaw-depth relationships along a Low-Arctic bioclimate gradient, Alaska: Synthesis of information from the ATLAS studies. Permafrost and Periglacial Processes, 14(2): 103-123. Walker, M.D., Wahren, C.H., Hollister, R.D., Henry, G.H.R., Ahlquist, L.E., Alatalo, J.M., BretHarte, M.S., Calef, M.P., Callaghan, T.V., Carroll, A.B., Epstein, H.E., Jónsdóttir, I.S., Klein, J.A., Magnússon, B., Molau, U., Oberbauer, S.F., Rewa, S.P., Robinson, C.H., Shaver, G.R., Suding, K.N., Thompson, C.C., Tolvanen, A., Totland, O., Turner, P.L., Tweedie, C.E., Webber, P.J., and P.A. Wookey. 2006. Plant community responses to experimental warming across the tundra biome. Proceedings of the National Acadamy of Sciences of the United States of America, 103(5): 1342–1346. Welker, J.M. 1999. CO2 flux in Arctic and alpine dry tundra: Comparative field responses under ambient and experimentally warmed conditions. Arctic, Antarctic, and Alpine Research, 31(3): 272-277. Welker, J.M., Fahnestock, J.T., Povirk, K.L, Bilbrough, C.J., and R.E. Piper. 2004a. Alpine grassland CO2 exchange and nitrogen cycling: grazing history effects, Medicine Bow Range, Wyoming,U.S.A. Arctic, Antarctic, and Alpine Research, 36(1): 11-20.  91  Welker, J.M., Fahenstock, J.T., Henry, G.H.R., O’Dea, K.W. and R.A. Chimners. 2004b. CO2 exchange in three Canadian High Arctic ecosystems: response to long-term experimental warming. Global Change Biology, 10: 1981-1995. White, R.G. 1983. Foraging patterns and their multiplier effects on productivity of northern ungulates. Oikos, 40: 377-384. Whiting, G. J. and J.P. Chanton. 1993. Primary production control of methane emissions from wetlands. Nature, 364: 794-795. Wilmherst, J.F., Fryxell, J.M., and R.J. Hudson. 1995. Forage quality and patch choice by wapiti (Cervus elaphus). Behavioral Ecology, 6(2): 209-217. Yavitt, J.B., Williams, C.J., and R.K. Wieder. 1997. Production of methane and carbon dioxide in peatland ecosystems across North America: Effects of temperature, aeration, and organic chemistry of the peat. Geomicrobiology Journal, 14: 299-316. Zalatan, R., Gunn, A., and G.H.R. Henry. 2006. Long-term Abundance Patterns of Barrenground Caribou Using Trampling Scars on Roots of Picea Mariana in the Northwest Territories, Canada. Arctic, Antarctic, and Alpine Research, 38(4): 624-630.  92  Appendix A Net Ecosystem Production values by treatment for 2004, 2005, and 2006. n=30. Vertical dashed lines indicate the beginning and end of peak biomass, with early and late season data on either side  CO2 exchange (μmol m  -2 -1 s )  0.0 -0.5 -1.0 -1.5 -2.0  CL CN ML MN HL HN  -2.5 -3.0  2004 -3.5 204  220  232  255  CO2 exchange (μmol m-2 s-1)  0.0 -0.5 -1.0 -1.5 -2.0  CL CN ML MN HL HN  -2.5 -3.0  2005 -3.5 169  190  193  196  200  210  CO2 exchange (μmol m-2 s-1)  0.0  213  217  228  231  236  238  244  Julian Day  -0.5 -1.0 -1.5 -2.0 CL CN ML MN HL HN  -2.5 -3.0  2006 -3.5 160  173  177  186  190  195  201  208  214  217  220  228  232  240  Day of Year (Categorical)  93  Appendix B Gross Ecosystem Production values by treatment for 2004, 2005, and 2006. n=30. Vertical dashed lines indicate the beginning and end of peak biomass, with early and late season data on either side  CO2 exchange (μmol m  -2 -1 s )  0 -1 -2 -3 CL CN ML MN HL HN  -4 -5 -6  2004 204  220  232  255  CO2 exchange (μmol m-2 s-1)  0 -1 -2 -3 CL CN ML MN HL HN  -4 -5 -6  2005 169  190  193  196  200  210  213  217  228  231  236  238  244  CO2 exchange (μmol m-2 s-1)  0 -1 -2 -3 CL CN ML MN HL HN  -4 -5 -6  2006 160  173  177  186  190  195  201  208  214  217  220  228  232  240  Day of Year (Categorical)  94  Appendix C Ecosystem Respiration values by treatment for 2004, 2005, and 2006. n=30. Vertical dashed lines indicate the beginning and end of peak biomass, with early and late season data on either side 2004  CL CN ML MN HL HN  CO2 exchange (μmol m  -2 -1 s )  4  3  2  1  0 204  CO2 exchange (μmol m-2 s-1)  4  220  232  255  2005  CL CN ML MN HL HN  3  2  1  0 169  CO2 exchange (μmol m-2 s-1)  4  190  193  196  200  210  213  217  228  231  236  238  244  CL CN ML MN HL HN  2006  3  2  1  0 160  173  177  186  190  195  201  208  214  217  220  228  232  240  Day of Year (Categorical)  95  Appendix D Means and standard error of the means for soil moisture data 2004  2005  Day of Year  Soil Moisture %  CL CN ML MN HL HN  204 204 204 204 204 204  11 10 10 12 12 11  Std Error of the Mean 2.35 3.60 2.07 3.33 3.19 2.35  CL CN ML MN HL HN  220 220 220 220 220 220  19 19 24 27 26 21  CL CN ML MN HL HN  232 232 232 232 232 232  CL CN ML MN HL HN  255 255 255 255 255 255  Treatment  CL CN ML MN HL HN  2006  Day of Year  Soil Moisture %  169 169 169 169 169 169  42 29 56 44 40 35  Std Error of the Mean 10.84 8.43 5.68 7.89 6.31 6.73  5.07 5.14 5.44 6.10 7.25 2.70  190 190 190 190 190 190  30 28 40 32 28 33  22 17 21 28 32 23  5.89 3.69 2.32 5.85 7.94 3.60  193 193 193 193 193 193  30 21 28 29 35 23  10.40 3.84 2.27 6.50 8.85 3.05  Day of Year  Soil Moisture %  160 160 160 160 160 160  46 35 58 39 36 43  Std Error of the Mean 4.91 9.37 9.07 13.07 4.60 6.59  6.86 6.64 9.14 8.24 4.50 5.79  173 173 173 173 173 173  28 28 43 30 43 24  4.71 7.83 9.34 7.96 13.47 4.30  32 17 39 31 30 24  6.92 6.09 5.37 5.81 10.68 4.76  177 177 177 177 177 177  22 24 35 24 39 28  5.04 6.53 7.19 5.26 11.85 4.49  196 196 196 196 196 196  24 23 47 29 36 29  5.78 7.16 9.37 3.57 10.99 6.10  186 186 186 186 186 186  34 34 34 40 36 39  7.43 11.43 8.48 10.10 7.37 6.17  200 200 200 200 200 200  27 19 33 23 35 32  6.07 7.26 7.21 4.92 11.23 6.34  190 190 190 190 190 190  34 29 44 34 33 36  7.98 9.64 6.19 6.53 7.46 6.85  96  2004  2005 Std Error of the Mean  2006  Day of Year  Soil Moisture %  CL CN ML MN HL HN  210 210 210 210 210 210  23 15 31 25 34 21  Std Error of the Mean 4.93 5.89 8.83 3.54 8.97 4.20  CL CN ML MN HL HN  213 213 213 213 213 213  27 19 29 23 30 19  CL CN ML MN HL HN  217 217 217 217 217 217  CL CN ML MN HL HN  Day of Year  Soil Moisture %  195 195 195 195 195 195  27 28 42 33 32 28  Std Error of the Mean 8.77 9.91 2.80 4.18 8.15 6.49  6.91 3.81 9.17 6.66 7.97 4.18  201 201 201 201 201 201  24 27 90 26 31 26  7.29 8.64 52.80 5.94 6.44 4.59  28 16 31 28 33 20  7.62 4.32 7.00 4.96 10.24 3.15  208 208 208 208 208 208  29 24 33 28 35 28  6.53 7.12 6.41 3.14 8.99 6.27  228 228 228 228 228 228  30 26 35 29 34 30  5.86 6.54 7.80 3.53 8.79 6.86  217 217 217 217 217 217  23 19 29 28 22 24  5.18 6.19 5.61 4.80 6.95 5.15  CL CN ML MN HL HN  231 231 231 231 231 231  32 19 30 23 26 24  7.11 6.07 3.50 4.29 7.11 4.72  220 220 220 220 220 220  21 17 29 22 27 21  6.31 5.62 5.31 7.31 9.40 4.69  CL CN ML MN HL HN  236 236 236 236 236 236  24 23 30 27 25 28  6.08 6.50 4.44 4.75 7.13 4.62  214 214 214 214 214 214  21 23 37 20 29 26  5.87 6.87 6.40 4.15 11.20 5.69  Treatment  Day of Year  Soil Moisture %  97  2004  2005 Std Error of the Mean  2006  Day of Year  Soil Moisture %  CL CN ML MN HL HN  238 238 238 238 238 238  34 16 26 22 25 25  Std Error of the Mean 6.65 5.83 5.13 2.07 6.88 5.21  CL CN ML MN HL HN  240 240 240 240 240 240  28 19 28 23 25 24  CL CN ML MN HL HN  244 244 244 244 244 244  31 21 30 26 32 25  Treatment  Day of Year  Soil Moisture %  Day of Year  Soil Moisture %  228 228 228 228 228 228  18 21 28 24 25 23  Std Error of the Mean 3.87 7.48 5.62 6.89 5.92 3.19  5.15 5.72 4.66 4.29 7.64 4.83  232 232 232 232 232 232  22 15 26 17 25 20  6.95 5.47 5.30 4.35 8.91 4.14  6.57 6.06 5.68 3.00 9.03 4.34  240 240 240 240 240 240  18 16 19 22 24 25  5.27 3.51 4.23 7.13 7.77 4.51  98  Appendix E Daily mean, maximum and minimum soil temperature data for the three study years 2004  2005  2006  CL  Day of Year 204  CN  204  8.0  9.4  11.1  169  4.7  6.2  7.2  160  7.3  10.1  12.3  ML MN  204 204  5.8 7.2  8.7 8.7  11.0 9.9  169 169  3.3 4.2  4.8 5.3  5.6 7.8  160 160  6.4 7.8  8.8 9.3  10.3 10.7  Treatment  Minimum (˚C)  Mean (˚C)  Maximum (˚C)  4.5  8.8  10.2  Day of Year 169  Minimum (˚C)  Mean (˚C)  Maximum (˚C)  1.8  4.7  6.3  Day of Year 160  Minimum (˚C)  Mean (˚C)  Maximum (˚C)  8.5  9.6  11.2  HL  204  6.5  8.6  11.1  169  4.5  5.4  6.2  160  9.2  10.4  11.9  HN  204  6.0  9.4  11.2  169  3.4  6.1  7.3  160  8.5  9.5  10.7  CL  220  5.6  8.1  9.9  190  6.4  8.5  9.9  173  4.8  9.6  11.8  CN  220  6.5  8.3  9.1  190  8.2  10.2  12.9  173  8.0  11.1  12.6  ML MN  220 220  6.0 5.9  9.2 7.3  12.1 9.6  190 190  7.4 7.3  9.4 9.6  11.0 11.5  173 173  7.4 8.1  9.8 10.8  11.7 14.5  HL  220  4.5  7.4  9.4  190  7.7  9.4  11.8  173  9.1  10.2  11.3  HN  220  6.5  8.7  10.2  190  5.9  9.5  11.2  173  6.6  11.4  14.3  CL  232  1.9  4.2  5.6  193  8.5  30.8  110.8  177  8.9  13.7  17.3  CN  232  3.4  4.4  5.0  193  12.3  13.1  14.3  177  11.6  14.0  15.2  ML  232  2.9  4.4  5.7  193  9.9  11.6  12.8  177  10.8  13.4  15.4  MN  232  2.8  3.4  4.3  193  10.8  11.9  13.8  177  10.7  13.6  15.0  HL  232  3.4  4.1  5.3  193  8.4  10.8  13.1  177  10.7  13.5  18.6  HN  232  3.0  4.4  5.3  193  11.2  12.4  13.4  177  13.8  15.2  16.7  99  2004  2005  Minimum (˚C)  Mean (˚C)  Maximum (˚C)  0.6  1.8  3.0  Day of Year 196  2006  Minimum (˚C)  Mean (˚C)  Maximum (˚C)  5.6  9.3  11.9  Day of Year 186  CL  Day of Year 255  CN  255  1.3  2.1  3.2  196  9.1  10.2  11.2  186  Treatment  Minimum (˚C)  Mean (˚C)  Maximum (˚C)  8.3  12.8  15.6  1.6  12.1  16.8  ML  255  1.3  2.5  3.3  196  6.6  9.4  11.6  186  9.5  10.8  12.5  MN  255  1.1  1.5  1.9  196  8.6  9.4  10.2  186  10.3  12.1  14.9  HL  255  1.3  1.9  2.4  196  8.0  9.3  11.3  186  9.9  12.2  14.3  HN  255  1.3  2.1  3.3  196  8.1  9.8  11.0  186  8.7  11.4  12.5  CL  200  4.3  7.8  10.0  190  5.3  10.0  13.7  CN  200  7.4  9.0  10.0  190  10.2  10.8  12.1  ML  200  6.8  7.8  9.4  190  8.2  9.1  10.9  MN  200  7.7  8.3  8.7  190  7.6  9.2  11.0  HL  200  7.0  7.8  8.6  190  9.2  11.3  15.6  HN  200  7.2  8.3  9.3  190  6.4  9.6  11.2  CL  210  5.5  7.7  9.8  195  7.1  10.8  12.5  CN  210  7.5  8.6  9.8  195  10.9  13.6  17.2  ML  210  6.3  8.3  9.5  195  8.9  11.6  15.2  MN  210  6.8  7.4  8.0  195  9.0  12.7  17.4  HL  210  6.0  7.3  7.9  195  9.7  12.3  16.3  HN  210  6.9  8.4  9.3  195  8.4  11.5  13.2  CL  213  4.7  8.2  10.2  201  9.7  14.2  18.8  CN  213  8.4  9.9  11.4  201  12.6  15.5  17.2  ML  213  7.4  9.1  11.1  201  12.1  14.4  18.1  MN  213  8.3  9.2  10.6  201  12.1  14.1  19.4  HL  213  6.5  8.1  9.1  201  12.4  14.1  16.4  HN  213  6.9  9.4  10.4  201  10.0  15.2  17.7  100  2004  2005  2006  Minimum (˚C)  Mean (˚C)  Maximum (˚C)  4.9  8.0  10.0  Day of Year 208  CL  Day of Year 217  CN  217  8.7  9.4  10.7  208  10.9  12.5  14.9  Treatment  Day of Year  Minimum (˚C)  Mean (˚C)  Maximum (˚C)  Minimum (˚C)  Mean (˚C)  Maximum (˚C)  7.6  11.5  14.5  ML  217  6.6  8.6  10.7  208  8.5  11.9  14.4  MN  217  8.1  8.8  10.3  208  8.5  10.8  16.1  HL  217  6.3  8.0  9.0  208  10.4  11.9  15.0  HN  217  7.3  9.1  10.4  208  8.0  11.3  13.4  CL  228  3.3  6.1  7.4  217  7.8  12.6  17.5  CN  228  6.2  7.5  8.6  217  11.6  13.6  16.1  ML  228  5.0  7.0  8.5  217  9.6  13.2  16.2  MN  228  5.3  6.2  7.7  217  11.0  12.6  16.5  HL  228  4.8  5.8  6.9  217  11.6  13.0  16.2  HN  228  4.6  7.0  8.1  217  9.3  14.4  17.1  CL  231  3.5  6.6  8.5  220  9.5  13.0  15.7  CN  231  6.5  7.4  8.5  220  12.6  15.3  16.9  ML  231  5.1  6.9  9.1  220  11.5  13.7  16.7  MN  231  5.6  6.4  7.2  220  11.2  13.7  18.5  HL  231  5.1  6.0  6.6  220  13.8  14.8  16.0  HN  231  5.2  6.9  8.2  220  10.5  14.8  17.5  CL  236  2.9  5.9  7.2  214  10.2  12.4  15.8  CN  236  5.5  6.2  7.2  214  11.0  12.5  15.1  ML  236  4.3  5.6  7.1  214  9.3  11.5  13.8  MN  236  4.5  5.0  5.6  214  9.8  10.8  12.5  HL  236  4.4  5.1  5.9  214  9.5  11.6  13.7  HN  236  4.5  6.1  7.3  214  8.9  11.9  14.2  101  2004  2005  CL  Day of Year 238  CN  238  Treatment  Day of Year  Minimum (˚C)  Mean (˚C)  Maximum (˚C)  2006  Minimum (˚C)  Mean (˚C)  Maximum (˚C)  4.7  8.4  10.2  Day of Year 228  8.8  9.5  10.5  Minimum (˚C)  Mean (˚C)  Maximum (˚C)  6.6  10.5  14.0  228  9.7  11.7  14.1  ML  238  6.8  9.0  10.7  228  8.0  10.4  13.4  MN  238  7.9  8.3  8.6  228  8.3  10.8  14.9  HL  238  6.5  7.7  8.7  228  9.5  11.1  12.5  HN  238  7.1  9.3  10.5  228  8.0  11.1  13.4  CL  240  4.5  6.7  7.7  232  5.9  8.9  10.1  CN  240  6.3  7.4  8.2  232  9.5  11.3  14.5  ML  240  5.7  6.8  7.9  232  6.9  9.2  10.7  MN  240  5.9  6.8  9.0  232  7.2  9.0  12.3  HL  240  5.4  6.5  7.4  232  7.8  8.7  9.7  HN  240  6.0  7.1  7.8  232  6.8  10.3  13.0  CL  244  2.1  3.8  5.0  240  5.6  6.5  8.1  CN  244  3.2  4.3  5.5  240  6.1  7.9  9.3  ML  244  2.9  3.9  5.2  240  5.1  6.7  8.7  MN  244  3.0  3.3  3.6  240  5.0  6.6  8.8  HL  244  2.8  3.3  3.9  240  5.3  7.1  9.5  HN  244  2.7  3.9  4.5  240  4.8  7.1  8.9  102  

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.24.1-0072027/manifest

Comment

Related Items