Carbon Cycle Changes in a Changing Climate

MISSING IMAGE

Material Information

Title:
Carbon Cycle Changes in a Changing Climate Effects of Permafrost Thaw on Soil Carbon Accumulation, Ecosystem Respiration, and Decomposition
Physical Description:
1 online resource (205 p.)
Language:
english
Creator:
Pries, Caitlin E. Hicks
Publisher:
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:

Thesis/Dissertation Information

Degree:
Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Botany, Biology
Committee Chair:
Schuur, Edward A
Committee Members:
Lichstein, Jeremy W
Mack, Michelle C
Martin, Ellen E
Curtis, Jason H

Subjects

Subjects / Keywords:
accumulation -- alaska -- carbon -- climate -- decomposition -- delta13c -- moisture -- permafrost -- radiocarbon -- respiration -- temperature -- thaw -- tundra
Biology -- Dissertations, Academic -- UF
Genre:
Botany thesis, Ph.D.
Electronic Thesis or Dissertation
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )

Notes

Abstract:
Permafrost ecosystems have historically been carbon (C)sinks, but have the potential to become C sources to the atmosphere as climate change causes permafrost to thaw. Over thousands of years, generally freezing temperatures have allowed permafrost ecosystems to accumulate a large soil C pool. In a warming climate, this C pool is at risk of being respired into CO2,causing more warming via the greenhouse effect. The magnitude of this permafrost thaw climate change feedback is currently unknown. In tundra soils near Healy, Alaska, soil C stocks are substantial, >50 kg C m-2 in the top meter (e.g., m3), and the average ecosystem carbon balance over the past 50 years has been 14.4 g C m-2 y-1, indicating this tundra was a C sink. However, measurable vertical soil accretion ceased within the past two decades, indicating this tundra may now be becoming a C source. As permafrost thaws, ecosystem respiration C losses increase. If old soil respiration drives this increase, a large positive feedback is likely. To investigate drivers of this increase, ecosystem respiration was partitioned into autotrophic, young soil, and old soil sources using C isotopes in a permafrost thaw gradient and a warming experiment in Healy, Alaska. Permafrost thaw and warmer soils increased old soil C loss by 34 to 150%, but autotrophic respiration also increased because thaw enhanced primary production. Old soil respiration losses are currently balanced by production increases. However,future increases in old soil respiration will result in a positive feedback because the soil C pool has greater potential for C loss than the plant C pool has potential to gain. Permafrost thaw and climate change also affect C loss atthe soil surface (top 20 cm) by changing soil moisture. Surface decomposition rates increased when water tables were shallower in the warming experiment and when growing season precipitation increased at the thaw gradient. Overall, warmer soils increase heterotrophic respiration, which will lead to a positive feedback to climate change. However, this effect will be modified by soil moisture, which can both decrease and increase as a result of permafrost thaw.
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis:
Thesis (Ph.D.)--University of Florida, 2012.
Local:
Adviser: Schuur, Edward A.
Electronic Access:
RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2013-06-30
Statement of Responsibility:
by Caitlin E. Hicks Pries.

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Classification:
lcc - LD1780 2012
System ID:
UFE0044935:00001


This item is only available as the following downloads:


Full Text

PAGE 1

1 CARBON CYCLE CHANGES IN A CHANGING CLIMAT E: EFFECTS OF PERMAFROS T THAW ON SOIL C ARBON ACCUMULATION ECOSYSTEM RESPIRATIO N, AND DECOMPOSITION By CAITLIN E HICKS PRIES A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2012

PAGE 2

2 2012 Caitlin E Hicks Pries

PAGE 3

3 To Alex, for your unconditional support throughout this arduous process, and to Dale, for your companionship and protection in the field

PAGE 4

4 ACKNOWLEDGMENTS This dissertation would not have been possible without the help of many colleagues First, I would like to thank the Healy Alaska field crew; t ogether we shared field site s and a tiny cabin with only two rooms and no running water. Specifically, thanks to Christian Trucco for teaching me numerous field methods, Andres Baron Lopez and Derek Deraps for a ssistance in the field and Sue Natali for her hard work on CiPEHR and ge neral guidance. A special thanks to Fay Belshe, whose friendship kept me sane while living in rural Alaska and whose statistics advice has been invaluable I would also like to thank my original office mates at University of Florida: Silvia Alvarez, Jennie DeMarco, and Jenny Schafer. They taught me how to navigate graduate school and together we supported each other through our dissertations Th e stable carbon isotope work was made possible with assistance from Jason Curtis and Kathryn Venz Curtis at UF Thanks to Xiaomei Xu for running my radiocarbon samples on the accelerator mass spectrometer at UC Irvine. Many thanks to Grace Crummer who processed the hundreds of isotope samples I mailed her each summer ran my CN samples, and patiently taught me to process radiocarbon samples Thanks to all the undergraduate students who assisted me with my lab work: Elaine Pegoraro, Desirai Rogan, Chase Mason, Nancy Davison, Elizabeth Wells, Andrew Lee, Lauren Michaels, and Kalindhi Lavios They made sample process ing go a lot faster and were great company for long hours in the lab. Thanks to my committee members, Michelle Mack, Jeremy Lichstein, Jason Curtis, and Ellen Martin, for their comments and guidance. Lastly, thanks to my advisor Ted Schuur, for giving me t his amazing opportunity to study permafrost thaw in Alaska and for allowing me freedom to

PAGE 5

5 pursue my intellectual interests while also making sure my ambitious plans stayed feasible This work was funded by an NSF Doctoral Diss ertation Improve ment Grant, a Denali National Park Murie Science and Learning Center grant and DOE and NSF grants to Dr. Ted Schuur

PAGE 6

6 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 9 LIST OF FIGURES ................................ ................................ ................................ ........ 10 ABSTRACT ................................ ................................ ................................ ................... 12 CHAPTER 1 BACKGROUND AND OBJECTIVES ................................ ................................ ...... 14 2 HOLOCENE CARBON STOC KS AND CARBON ACCUMU LATION RATES ALTERED IN SOILS UND ERGOING PERMAFROST T HAW ................................ 20 Abstract ................................ ................................ ................................ ................... 20 Introduction ................................ ................................ ................................ ............. 21 Methods ................................ ................................ ................................ .................. 24 Study Area ................................ ................................ ................................ ........ 24 Co re Collection and Processing ................................ ................................ ....... 25 Carbon Accumulation Rates ................................ ................................ ............. 27 Result s ................................ ................................ ................................ .................... 30 Soil Carbon Inventories ................................ ................................ .................... 30 Soil Profiles ................................ ................................ ................................ ...... 31 Carbon Accumulation Rates ................................ ................................ ............. 32 Discussion ................................ ................................ ................................ .............. 33 Carbon Inventories and Depth Profiles ................................ ............................. 33 C arbon Accumulation Rates ................................ ................................ ............. 34 Net Ecosystem Production ................................ ................................ ............... 36 3 THAWING PERMAFROST I NCREASES OLD SOIL AN D AUTOTROPHIC RESPIRATION IN TUNDR A: PARTITIONING ECOS YSTEM RESPIRATION USING ................................ ................................ ......................... 50 Abstract ................................ ................................ ................................ ................... 50 Introduction ................................ ................................ ................................ ............. 51 Materials and Methods ................................ ................................ ............................ 54 Site Description ................................ ................................ ................................ 54 Ecosystem Respiration ................................ ................................ ..................... 55 Autotrophic Respiration ................................ ................................ .................... 57 Heterotrophic Respiration ................................ ................................ ................. 57 Data A nalysis and Partitioning Model ................................ ............................... 60 Respiration Fluxes ................................ ................................ ............................ 61 Results ................................ ................................ ................................ .................... 62

PAGE 7

7 Ecosystem Respiration ................................ ................................ ..................... 62 Source Res piration ................................ ................................ ........................... 63 Partitioning Ecosystem Respiration ................................ ................................ .. 64 Respiration fluxes ................................ ................................ ............................. 66 Discussion ................................ ................................ ................................ .............. 66 Variability of autotrophic source isotopes ................................ ......................... 67 Variability of heterotrophic source isotopes ................................ ...................... 68 Partitioning ecosystem respiration ................................ ................................ .... 69 Challenges of the isotopic partitioning approach ................................ .............. 73 Implications for net ecosystem carbon balance ................................ ................ 74 4 EXPERIMENTAL WARMING CHANGES THE RESPONSE OF RESPIRATION TO SOIL TEMPERATURE ................................ ................................ ...................... 85 Abstract ................................ ................................ ................................ ................... 85 Introduction ................................ ................................ ................................ ............. 86 Methods ................................ ................................ ................................ .................. 90 Site Description ................................ ................................ ................................ 90 Soil Enivronment ................................ ................................ .............................. 91 Ecosystem Respiration ................................ ................................ ..................... 91 Source Res piration ................................ ................................ ........................... 92 Partitioning Model ................................ ................................ ............................. 95 Data Analysis ................................ ................................ ................................ ... 96 Results ................................ ................................ ................................ .................... 98 Soil Environment ................................ ................................ .............................. 98 Ecosystem Respiration ................................ ................................ ..................... 99 Sources of Ecosystem Respiration ................................ ................................ 100 Respiration Source Contributions ................................ ................................ ... 101 Discussion ................................ ................................ ................................ ............ 102 Partitioning Ecosystem Respiration ................................ ................................ 102 Old Soil Carbon Losses ................................ ................................ .................. 105 Variability of Ecosystem Respiration Sources ................................ ................ 105 Implications for Net Ecosystem Carbon Balance ................................ ............ 107 5 MOISTURE CONTROLS DE COMPOSITION IN THAWI NG AND WARMING TUNDRA ................................ ................................ ................................ ............... 121 Abstract ................................ ................................ ................................ ................. 121 Introduction ................................ ................................ ................................ ........... 122 Methods ................................ ................................ ................................ ................ 125 Study Site ................................ ................................ ................................ ....... 125 Common Substrate Decomposition ................................ ................................ 127 Common Garden and Plant Community Decomposition ................................ 129 Data Analysis ................................ ................................ ................................ 132 Results ................................ ................................ ................................ .................. 134 Common Substrate Decomposition ................................ ................................ 134 Common Garden and Plant Community Decomposition ................................ 136

PAGE 8

8 Discussion ................................ ................................ ................................ ............ 137 Moisture Controls Decomposition ................................ ................................ ... 137 Community Decomposition ................................ ................................ ............. 141 Relative Effects of Environment and Substrate Changes ............................... 143 6 CONCLUSION ................................ ................................ ................................ ...... 152 APPENDIX 1 SUPPLEMENTAL TABLES FOR CHAPTER 3 ................................ ..................... 157 2 SUPPLEMENTAL TABLES FOR CHAPTER 4 ................................ ..................... 165 3 SUPPLEMENTAL TABLES FOR CHAPTER 5 ................................ ..................... 175 LIST OF REFERENCES ................................ ................................ ............................. 191 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 2 04

PAGE 9

9 LIST OF TABLES Table page 2 1 Mean (standard error) organic soil depth, active layer depth, and C inventory in the top 1 m, in the organic horizons, and in the active layer for minimal, moderate, and extensive perma frost thaw sites. ................................ ................ 39 2 2 Mean (standard error) bulk density, % C, and % N for three soil depths at minimal, moderate, and extensive permafrost thaw sites ................................ ... 40 2 3 I (C inputs), k (decomposition constant), and t i (x intercept) parameters fitted to Eq. 2.3 (decadal) or Eq. 2.2 (millennial) with their standard errors. ................ 41 2 4 Decadal Net Ecosys tem Production (NEP). ................................ ....................... 42 3 1 13 C of ecosystem respiration within the active layer categories throughout the growing season in 2008 and 2009. ................... 76 3 2 13 14 C of AG and BG plant respiration during the 2008 and 2009 growing season. ................................ ................................ ................. 77 3 3 13 14 C of young and old soil respiration ............................. 78 4 1 Mean (SE) environmental variables by sampling date and treatment ............. 109 4 2 Multiple regression results ................................ ................................ ................ 110 4 3 13 14 C of ecosystem respiration by treatment and sampling date ................................ ................................ ................................ ... 111 4 4 13 14 C of aboveground (AG) and belowground (BG) plant respiration sampled in 2009, 2010, and 2011 ................................ .......... 112 4 5 13 14 C of young and old soil respiration. .......................... 113 5 1 Multiple regression results for annual and growing season decomposition of a common substrate at the permafrost thaw gradient and warming experiment (CiPEHR ) ................................ ................................ ....................... 145 5 2 Decomposition constants (k, n=5) and initial litter quality (n=3) from the common garden experiment near the permafrost th aw gradient in Healy, Alaska ................................ ................................ ................................ .............. 146

PAGE 10

10 LIST OF FIGURES Figure page 1 1 A schematic of permafrost thaw ................................ ................................ ......... 18 1 2 Sources of ecosystem respiration (R eco ) in tundra underlain by permafrost ....... 19 2 1 Depth profiles for each core taken at minimal, moderate, and extensive permafrost thaw sites ................................ ................................ ......................... 43 2 2 13 15 N of organic (top) and mineral (bottom) soil layers at the minimal, moderate, and extensive thaw sites ................................ ............... 44 2 3 Depth profiles of radiocarbon values used to calculate ages for decadal (A) and millennial (B) C accumulation models ................................ .......................... 45 2 4 Surface cumulative C inventories versus the age of C in that soil layer.. ........... 46 2 5 Cumulative C inventories versus the age of C in that soil layer .......................... 47 2 6 Decadal Net Ecosystem Production (NEP) for two cores at each site in the thaw gradient and annual NEP calculated from recent flux measurements ........ 48 2 7 The relationship between C input ( I Table 3) and decomposition rates ( kC Tables 3 and 4) for the six surface soils (decadal rates) and the two deep soils (millennial rates, bottommost points). ................................ ......................... 49 3 1 Mean R eco 14 C for all active layer (AL) categories in all months sampled. ........ 79 3 2 Mean isotopic values (error bars=SE) of R h by depth .. ................................ ....... 80 3 3 Percent contributions of sources to R eco by month for each active layer category ................................ ................................ ................................ .............. 81 3 4 SIAR estimates for contributions of aboveground R auto (AG, black symbols) and belowground R auto (BG, open symbols) averaged over active layer category and year for each month sampled ................................ ....................... 82 3 5 Autotrophic respiration contributions to R eco (A), old soil respiration contributions to R eco (B), and the autotrophic to heterotrophic respiration ratio (C) with thaw depth for each active layer (AL) category and month ................... 83 3 6 Estimates of growing season fluxes from each respiration source (aboveground (AG), belowground (BG), young soil (YS), and old soil (OS)) for active layer (AL) categories in 2008 and 2009. ................................ ............. 84 4 1 R eco 14 C decreased as ecosystem respiration fluxes increased ...................... 114

PAGE 11

11 4 2 R eco 14 C decreased with soil temperature at 20 cm ................................ ........ 115 4 3 Mean proportional contributions of aboveground (AG) and belowground (BG) autotrophic respiration (top graphs) and young soil (YS) and deep soil (DS) heterotrophic respiration by tre atment for each sampling period ...................... 116 4 4 Old soil contributions to R eco generally increased with the soil temperature at 20 cm ................................ ................................ ................................ ................ 117 4 5 The mean ratio of heterotrophic to autotrophic respiration across all sampling dates ................................ ................................ ................................ ................. 118 4 6 Old soil contributions to R eco increased w ith increasing respiration flux ........... 119 4 7 Estimated growing season old C losses from the control and warming treatments in 2010 and 2011. ................................ ................................ ........... 120 5 1 Annual percent mass loss from the common substrate decomposition bags at the three thaw gradient sites from Septe mber 2004 through September 2011 147 5 2 Growing season (GS, top panel) and annual (bottom two panels) percent mass loss from the common substrate decomposition bags at CiPEHR .......... 148 5 3 Annual decomposition increased with more growing season precipitation at the thaw gradient ................................ ................................ .............................. 149 5 4 Annual decomposition increased with shallower water tables at CiPEHR. ....... 150 5 5 Community weighted d ecomposition constants (k) ................................ .......... 151 6 1 Permafrost thaw and warming increases the ratio of autotrophic to heterotrophic respiration duri ng the growing season ................................ ........ 156

PAGE 12

12 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy CARBON CYCLE CHANGES IN A CHANGING CLIMAT E: EFFECTS OF PERMAFROS T THAW ON SOIL C ARBON ACCUMULATION ECOSYSTEM RESPIRATIO N AND DECOMPOSITION By Caitlin E Hicks Pries December 2012 Chair: Edward A.G. Schuur Major: Botany Permafrost ecosystems have historically been carbon (C) sinks, but have the potential to become C sources to the atmosphere as climate change causes permafrost to thaw. Over thousands of years, generally freezing temperatu res have allowed permafrost ecosystems to accumulate a large soil C pool. In a warming climate, this C pool is at risk of being respired into CO 2 causing more warming via the greenhouse effect The magnitude of this permafrost thaw climate change feedback is currently unknown. I n tundra soils near Healy, Alaska, s oil C stocks are substantial, > 50 kg C m 2 in the top meter (e.g., m 3 ) and the average ecosystem carbon balance over the past 50 years has been 14.4 g C m 2 y 1 indicating this tundra was a C si nk However, measurable vertical soil accretion ceased within the past two decades indicating this tundra may now be becoming a C source. As permafrost thaws, ecosystem respiration C losses increase If old soil respiration drives this increase, a large positive feedback is likely. To investigate drivers of this increase ecosystem respiration was partitioned into autotrophic, young soil, and old soil sources using C isotopes in a permafrost thaw gradient and a warming

PAGE 13

13 experiment in Healy, Alaska Permafrost thaw and warmer soils increased old soil C loss by 34 to 150%, but autotrophic respiration also increased because thaw enhanced primary production. O ld soil respiration losses are currently balanced by production increases However, future incre ases in old soil respiration will result in a positive feedback because the soil C pool has greater potential for C loss than the plant C pool has potential to gain. Permafrost thaw and climate change also affect C loss at the soil surface (top 20 cm) by c hanging soil moisture. Surface decomposition rates increased when water tables were shallower i n the warming experiment and when growing season precipitation increased at the thaw gradient. O verall, w armer soils increase heterotrophic respiration which wi ll lead to a positive feedback to climate change However, this effect will be modified by soil moisture, which can both decrease and increase as a result of permafrost thaw.

PAGE 14

14 CHAPTER 1 BACKGROUND AND OBJECTIVES Climate change has increased air temperatures in high latitudes by 2 C over the past 60 years with 7 to 8 C increases predicted over the next century (IPCC, 2007). This warming is greatly changing arctic and boreal ecosystems causing shrub expansion (Strurm et al ., 2001), increases in plant productivity (Jia et al ., 2003), increases in fire frequency (Kasischke & Turetsky 2006), and permafrost thaw (Schuur et al ., 2008). Permafrost is soil or rock that is frozen for two or more consecutive years. Permafrost underlies many arctic and boreal ecosystems, but its extent is projected to decrease 20 to 89% by 2100 ( Lawrence et al ., 2008 ; Schaefer et al ., 2011) Soils in the permafrost zone store 1 672 Pg carbon (C), twice the amount of C as currently resides in our atmosphere (Tarnocai et al ., 2009). Freezing temperatures have allowed this large C pool to accumulate over hundreds to thousands of years by slowing decomposition. As permafrost thaws, t his carbon is exposed to microbial degradation and lost to the atmosphere as carbon dioxide (CO 2 ) or methane (Goulden et al ., 1998). Permafrost thaw may thus function as a large positive feedback to climate change as warming causes soil to thaw, releasing CO 2 which in turn causes more warming via the greenhouse effect Coupled carbon cycle climate models predict that 33 to 114 Pg C will be lost fr om permafrost soils by 2100 (Kov en et al ., 2011; von Deimling et al ., 2012). In an independent estimate based on field observations, C flux due to permafrost thaw in 2100 may be equal to the current flux of land use change, about 1.5 Pg C y 1 (Schuur et al ., 2009) The objective of this dissertation is to investigate how permafrost thaw changes th e carbon cycle of tundra ecosystems, with emphasis on how C outputs like respiration

PAGE 15

15 and decomposition are affected. Tundra ecosystems undergo numerous physical changes that can affect their C balance as permafrost thaw s (Fig. 1 1) T he amount of soil that thaws seasonally, known as the active layer, increases, soil temperatures rise, and the ground subsides, causing the water table to be closer to the surface. Microtopography becomes increasingly heterogeneous as the ground subsides into empty space left b y thawed ground ice, creating thermokarsts (Osterkamp et al ., 2009). These physical changes in turn affect plant communities shrubs and mosses replace sedges as the dominant vegetation in upland tundra (Schuur et al ., 2007) and carpets of sphagnum moss for m in collapsed peatland bogs (Turetsky et al ., 2007) In terms of C balance, thaw can result in increased C uptake by plants (Natali et al ., 2012 ) and increased soil C losses from microbial decomposition (Goulden et al ., 1998) Whether thaw causes permafrost underlain ecosystems to become a net C source, releasing CO 2 to th e atmosphere, or a net C sink, storing C in biomass and soils, ultimately depends on the balance between these inputs and outputs. In Chapter 2 of this dissertation, I quantify s urface and deep soil C stocks and C accumulation rates in a natural permafrost thaw gradient in Healy, Alaska. I inventory the amount of C stored in these soils to 1 meter and compare their relative degrees of decomposition based on their C:Nitrogen (N), 13 C, and 15 N profiles. Using soil radiocarbon profiles, I model millennial (over the Holocene) and decadal (past 50 years) C accumulation rates of this permafrost ecosystem and calculate its average net ecosystem production over the past 50 years. This re search finds the baseline C balance of this ecosystem that can then be compared with current functioning of the ecosystem as it undergoes permafrost thaw.

PAGE 16

16 One result of thaw is an increase in ecosystem respiration rates (Vogel et al ., 2009; Trucco et al ., 2012). Ecosystem respiration is made up of fluxes from autotrophic (e.g., plant) and heterotrophic (e.g., microbial) sources, either of which could cause the increase in ecosystem respiration, with different implications for the permafrost thaw climate ch ange feedback (Fig. 1 2). In Chapters 3 and 4 I use carbon isotopes, 13 C 14 C, to partition ecosystem respiration into four sources two autotrophic (above and belowground plant structures) and two heterotrophic ( young and old soil ) In Chapter 3 I partition ecosystem respiration in the aforementioned permafrost thaw gradient where thermokarsts have formed as a result of several decades of thaw. I investigate how the depth of thawed soil affects the proportional contributions and total growing season flux of autotrophic and heterotrophic respiration. In Chapter 4 I partition ecosystem respiration in CiPEHR (Carbon in Permafrost, Experimental Heating Research). This experiment is located several kilometers east of the thaw gradient and consists of two treatments in a factorial design. Summer warming warms the surface air during the growing season, and winter warming insulates the soil during winter, causing warmer deep soil temperatures year round and permafrost thaw. In CiPEHR, I investigate how soil temperature and treatment interact to affect autotrophic and heterotrophic respiration. The fate of old soil, which has accumulated over thousands of years is of particular concern because its mineralization adds CO 2 to the atmosphere and is likely irreversible over human timescales. Furthermore, the potential C loss from heterotrophic soil respiration is much larger than the potential C loss from autotrophic respiration or C gain through autotrophic production. The t undra soil C pool is 100 times larger than the tundra plant C pool (Hicks Pries et al. 2012; Natali et al ., 2012) and 10

PAGE 17

17 times larger than the boreal forest C pool that could potentially replace tundra vegetation as the climate warms ( Gower et al. 2001; Goulden et al. 2011). In Chapter 5 the focus narrows to how permafrost thaw affects a component of ecosystem respiration, the heterotrophic flux from the surface soil (0 20 cm). This decomposition flux is controlled by climate (both macro and micro scal e), substrate quality (i.e., nutrients and types of C molecules), and biota ( Levalle et al 1993 ) At the thaw gradient and CiPEHR, I use a common substrate to test how changes to the soil climate caused by permafrost thaw affect decomposition. I also inv estigate how interannual climate variability affects decompostion over seven years at the thaw gradient. To test the effects of substrate quality, I compare decomposition rates of 12 common plant substrates incubated at a single location near the thaw grad ient. I then combine those data with species level aboveground net primary productivity data to calculate plant community weighted decomposition constants for the thaw gradient and CiPEHR. The relative magnitude of the climate and plant substrate effects a re then compared. This dissertation quantifies the long term C balance of a subarctic tundra ecosystem and then explores how permafrost thaw is changing C outputs, a component of ecosystem C balance. I hypothesize that this ecosystem has been a C sink thr ough the Holocene and has continued to be a sink through the past 50 years. However, I also hypothesize that permafrost thaw and warming are turning this ecosystem into a C source by increasing C ouputs like old soil respiration and surface decomposition.

PAGE 18

18 Figure 1 1. A schematic of permafrost thaw. The brown area represents permafrost, the green area unfrozen soil, and the gray area ground water. In tundra with a lot of ground ice in the form of ice lenses and ice wedges, the ground subsides into the sp ace left by the melted ice as the permafrost thaws. As the ice thaws, active layer (AL) depths deepen exposing more soil carbon to above freezing decomposition. Due to ground subsidence, the water table (WT) is closer to the soil surface. Areas with deep s ubsidence (far right) are called thermokarsts and may have standing water. Areas adjacent to thermokarsts are often drier because ground water is funneled into the thermokarst.

PAGE 19

19 Figure 1 2. Sources of ecosystem respiration (R eco ) in tundra underlain by permafrost. In C hapters 3 and 4 carbon isotopes are used to partition ecosystem respiration into two autotrophic and two heterotrophic sources. Which source dominates ecosystem respiration partially determines the strength and direction of the permafrost thaw climate change feedback. Increases in old soil respiration with permafrost thaw indicate the loss of soil carbon that had previously been stored for hundreds to thousands of years, a large positive feedback to climate change. Increases in autotrophic respiration indicate a negative feedback, if the increases are driven by increased primary productivity, or a constrained positive feedback that is limited by the size of the plant carbon pool.

PAGE 20

20 CHAPTER 2 HOLOCENE CARBON STOC KS AND CARBON ACCUMULATION RATES A LTERED IN SOILS UNDERGOING PERMAFROST THAW 1 Abstract Permafrost soils are a significant global store of carbon (C) with the potential to become a large C source to the atmosphere. Climate change is causing permafrost to thaw, which can affect primary production and decomposition, therefore affecting ecosystem C balance To understand future responses of permafrost soils to climate 14 13 15 N depth profiles, mo deled soil C accumulation rates, and calculated decadal net ecosystem production in subarctic tundra soils undergoing minimal, moderate, and extensive permafrost thaw near Eight Mile Lake (EML) in Healy, AK. We modeled decadal and millennial soil C inputs, decomposition constants, and C accumulation rates by plotting cumulative C inventories against C ages based on radiocarbon dating of surface and deep soils respectively Soil C stocks at EML were substantial, over 50 kg C m 2 in the top meter, and did no t differ much among sites. Carbon to nitrogen ratio, 13 15 N depth profiles indicated most of the decomposition occurred within the organic soil horizon and practically ceased in deeper, frozen horizons. The average C accumulation rate for EML surfa ce soils was 25.8 g C m 2 y 1 and the rate for the deep soil accumulation was 2.3 g C m 2 y 1 indicating these systems have been C sinks throughout the Holocene. Decadal net ecosystem production averaged 14.4 g C m 2 y 1 However, the shape of decadal C a ccumulation curves combined with recent annual 1 This chapter has been published as Hicks Pries CE, Schuur EAG, Crummer KG (2012) Holocene c arbon s tocks and c arbon a ccumulation r ates a ltered in s oils u ndergoing p ermafrost t haw. Ecosystems, 12 162 173.

PAGE 21

21 NEP measurements, indicate soil C accumulation has halted and the eco system may be becoming a C source. Thus, the net impact of climate warming on tundra ecosystem C balance includes not only becoming a C sou rce, but also the loss of C uptake capacity these systems have provided over the past ten thousand years. Introduction One of the most significant global stores of carbon (C) lies within the soils of the an estimated 1672 Pg C is stored twice the current atmospheric C pool (Tarnocai et al ., 2009) Eighty eight percent of this C is frozen within actual permafrost (Tarnocai et al ., 2009) Over hundreds to thousands of years, permafrost has protected organic C from microbial degradation and allowed C to accumulate in soils. This vast store of C is now vulnerable due to accelerated warming at high latitudes (IPCC 2007). Permafrost area is projected to decrease substantially by 2100 with twenty (Schaefer et al. 2011) to eighty nine (Lawrence et al ., 2008) percent reductions predicted. As these permafrost soils thaw, organic C is exposed to microbial degradation and released as CO 2 or methane, both greenhouse gases (Goulden et al ., 1998) Permafrost thaw may thu s serve as a positive feedback to global climate change wherein warming causes soil thaw, releasing CO 2 which causes more warming (Schaefer et al ., 2011) In the first estimate based on field observations, effects of widespread permafrost thaw on raising atmospheric CO 2 levels by the end of this century may be equal to the current effect of deforestation (Schuur et al ., 2009) High latitude ecosystems undergo numerous changes that affect their C balance when the permafrost underneath them begins to thaw. Generally, active layer (the amount of soil that thaws seasonally) depths deepen and soil temperatures rise, which

PAGE 22

22 can cause gross primary production and winter respiration to increase (Vogel et al ., 2009) In tundra, shrubs and mosses have been found to r eplace sedges as the dominant vegetation (Schuur et al ., 2007) while in peatlands, carpets of sphagnum moss have been found to form where thaw has caused bog collapse (Turetsky et al ., 2007) While thaw can result in the mineralization of deep soil C (Gou lden et al ., 1998; Dorrepaal et al ., 2009; Schuur et al ., 2009) which can be thousands of years old (Goulden et al ., 1998; Dutta et al ., 2006; Zimov et al ., 2006; Schuur et al ., 2009) warming associated with thaw can also increase net primary productivit y (Natali et al ., 2012 ). Overall thaw induced changes in ecosystems affect C inputs to the soil from plant litter and moss growth and C losses from the soils due to decomposition. Whether thaw causes permafrost underlain ecosystems to become a net C source releasing CO 2 to the atmosphere, or a net C sink, storing C in biomass and soils, ultimately depends on the balance between these inputs and outputs. Scientists often use instantaneous CO 2 flux measurements of ecosystem respiration and gross primary pro duction to determine whether permafrost ecosystems are a C source or sink. These flux measurements can be chamber based (Goulden et al ., 1998; Wickland et al ., 2006) or from eddy covariance towers (Goulden et al ., 1998; Lund et al ., 2010) One chamber based study demonstrated that tundra ecosystems increase their C sink capacity at the start of permafrost thaw but become C sources as thaw progresses (Vogel et al ., 2009) Direct measurements of C fluxes give estimates of lance (e.g., net ecosystem production, NEP) that can change annually (Goulden et al ., 1998) due to variable weather conditions. For robust estimates of ecosystem C balance using flux measurements, multi year studies are needed.

PAGE 23

23 Furthermore, it can be diffi cult in cold systems to take the winter respiration measurements needed to calculate the annual net C balance using either chambers or eddy covariance (Lund et al ., 2010) Chamber based and eddy covariance studies of permafrost systems often only measure w inter respiration during a single month (Wickland et al ., 2006) or only measure growing season C fluxes (Turetsky et al ., 2007; Huemmrich et al ., 2010) Neglecting or under sampling winter respiration in high latitude ecosystems can lead to an incomplete u nderstanding of C balance because wintertime respiration fluxes, though low, occur for a much longer duration than growing season respiration (Vogel et al ., 2009) Alternatively, soil C accumulation rate measurements have been used to evaluate longer term ecosystem C balance. Several studies have investigated the response of C accumulation rates to recent climatic changes (Trumbore & Harden, 1997; Turetsky et al et al ., 2011) Turetsky et al. ( 2007) found collapsed bogs that had lost surfa ce permafrost had increased organic matter accumulation rates and therefore increased surface C storage. Such measurements are advantageous in addition to direct C flux measurements because they integrate C balance over many years and thus give a decadal C balance (Trumbore & Harden, 1997) But surface C accumulation rates do not account for deep soil decomposition and therefore do not measure NEP (Turetsky et al ., 2007) More C can be lost from deep soil than is accumulating in surface soil leading to a ne gative C balance but a positive C accumulation rate. However, decadal net C accumulation rates that include deep soil decomposition, and therefore estimate NEP, can be calculated when both recent and long term C accumulation rates from a soil are known (Tr umbore & Harden, 1997) Decadal NEP

PAGE 24

24 more accurately measure s net ecosystem C balance than surface C accumulation rates but cannot take into account recent changes to deep soil decomposition. In order to predict future responses of soil organic carbon to cl imate change, we must know the quantity of soil C currently stored and its response to recent climatic changes (Schuur et al ., 2008; Trumbore & Czimczik, 2008) To accomplish this, we chose to quantify surface and deep soil C stocks, inputs, and accumulati on rates in a permafrost thaw gradient near Eight Mile Lake in Healy, Alaska. Within this gradient, some areas have been undergoing thaw for at least two decades while others have only begun to thaw recently (Osterkamp et al ., 2009) We inventoried the amount of C stored in these soils to 1 meter and compared how much relative diagenesis has already occurred in these soils based on their C:N and stable isotope depth profiles. We used radiocarbon dates and C inventories to model how p ermafrost thaw has affected soil C accumulation and compared decadal C accumulation rates to millennial C accumulation rates. We hypothesized that this permafrost from accumulation models will show these soils have remained a C sink during the past five decades of climatic change. Finally, we compare decadal C balance calculated from C accumulation models to recent annual C balance calculated from flux measurements. Methods Study Area Our study area is a permafrost thaw gradient located near Eight Mile Lake (EML, named minimal moderate and extensive for the amount of vegetation change, active layer thickening, and thermokarst forma tion they have undergone due to different durations of permafrost thaw (Vogel et al ., 2009) At the extensive thaw site, permafrost

PAGE 25

25 thaw has been documented for the past two decades but likely began earlier (Osterkamp et al ., 2009) The vegetation is moist acidic tussock tundra underlain by soils that have permafrost within a meter of the surface (Gelisols). The soils consist of about 0.5 m of organic soil on top of mineral soil that is a mixture of loess deposits and glacial till (Vogel et al ., 2009) Perm afrost temperatures in this region are around 1C and therefore susceptible to thaw (Osterkamp & Romanovsky, 1999) The permafrost thaw gradient has ongoing monitoring of soil temperature, permafrost thaw depth, water table depth, and CO 2 fluxes that have been measured regularly from May through September for the past six years (Schuur et al ., 2009; Vogel et al ., 2009) Core Collection and Processing We collected six soil cores each at extensive, moderate, and minimal thaw sites in May 2004. In areas with out thermokarst, thawed soil was cut out in rectangular chunks using a bread knife. Chunk dimensions were measured from the hole left in the ground. Inside thermokarsts, a Reeburgur soil corer with a 36.8 cm diameter was used to core thawed soil. Frozen so il was cored using a Tanaka drill with a 7.6 cm diameter hollow bit. The lengths cored ranged from 50 to 90 cm because we stopped coring when the corer hit small rocks within the loess, and the depth where rocks stopped the corer varied. Cores were wrapped in aluminum foil and kept frozen until lab processing. In the lab, cores were weighed, thawed to be split into depth sections, and subsampled. Cores were split into the following depths: 0 5 cm, 5 15 cm, 15 25 cm, 25 35 cm, 35 cm to the end of the organic horizon, and 10 cm increments from the start of mineral soil until the end of the core. The organic/mineral horizon demarcation was determined visually and confirmed by %C analysis when the %C of the soil decreased to less than 20%. Lengthwise slices of t he core were refrozen for further analyses such

PAGE 26

26 as radiocarbon dating, while a slice was taken for moisture content and bulk %C, %N, 13 15 N analyses. To determine moisture content, we weighed a subsample of soil before and after drying in a 60C ov en for three days (organic soils) or a 100C oven for 24 hours (mineral soils). Dried organic soil was ground on a Wiley Mill while dried mineral soil was ground with a mortar and pestle. Ground soils were run on a Costech Analytical ECS 4010 elemental ana lyzer coupled to a Finnigan Delta Plus XL 13 15 N values. To determine bulk density, the dry mass of depth sections were divided by the field dimensions of rectangular chunks or corer dimensi ons then multiplied by section lengths. To determine C pools on a kg C m 2 basis, the bulk density was multiplied by the %C and divided by the height of each section. We calculated %C and %N in organic and mineral layers using bulk density weighted average s of the depth sections and 13 15 N using weighted averages based on the amount C or N in a depth section. To calculate C pools to a constant depth of 1 m, we extrapolated mineral pools in cores that were less than 1 m long. We divided the sum of the C pools in the mineral horizon by the depth of the mineral measured and multiplied by the depth needed to obtain a 1 m core. We used an average of all 10 cm mineral layers in the core to extrapolate because pool sizes did not significantly chan ge with depth in the mineral horizon. For comparisons of C pools, C:N ratios, and isotopic signatures among depth as main effects and a site x depth interaction in JMP 7 (SAS I nstitute, Cary, North Carolina).

PAGE 27

27 Carbon A ccumulation R ates To calculate millennial and decadal soil C accumulation rates, we modeled cumulative C inventories versus age using data from deep and surface soils, respectively. We then used parameters from both the millennial and decadal models to calculate decadal NEP. Radiocarbon values were used to age soil segments in each core. For 14 C analyses of decadal C accumulation rates, we used two cores from each site cores one and two were from minimal thaw, cores three and four were from moderate thaw, and cores five and six were from extensive thaw. The cost of radiocarbon dating limited the number of replicates we could have. We sectioned cores into 1 cm segments from 0 5 cm and into 2 cm segments thereafter. Fro m each section, we sorted out recognizable moss pieces. Holocellulose was extracted from a subset of the moss segments (five to six per core) using a modified Jayme Wise method wherein the moss is extracted with toluene and ethanol then bleached with sodiu m hypochlorite (Gaudinski et al ., 2005) Moss was used to date surface soil segments because moss accretes vertically in situ unlike bulk organic matter, which has vascular plant inputs. We dated moss cellulose specifically because the atoms in cellulose do not undergo exchange with newer photosynthate once fixed. Because moss cellulose was used to date layers while the total C inventory included vascular plant detritus, this method assumes moss a nd vascular plant inputs decompose at the same rate (Trumbore & Harden, 1997) Four to five additional samples were chosen along the whole profiles of core two (minimal thaw) and core three (moderate thaw) to calculate millennial accumulation rates. For m illennial rates, we plucked moss from 2 cm segments until we could no longer find enough recognizable moss (around 15 20 cm), then used 2 cm bulk samples

PAGE 28

28 for the rest of the organic layer, and 10 cm bulk samples for the mineral layer. We used bulk samples for millennial rates because the several year difference between when vascular plant litter was fixed and when the litter entered the soil is inconsequential on millennial timescales. All moss cellulose and bulk soil samples were combusted inside pre combu sted, evacuated quartz tubes with cupric oxide (Vogel et al ., 1987; Dutta et al ., 2006) The resulting air mixture was purified into CO 2 using liquid N 2 on a vacuum line before reducing the CO 2 into graphite by Fe reduction in He (Vogel et al ., 1987) Grap hite samples were sent to the UC Irvine W.M. Keck Carbon Cycle Accelerator 14 C analysis (precision 14 C values were referenced against the atmospheric radiocarbon record from Hua & Barbe tti (2004) to obtain calendar ages for the moss cellulose samples and subsequently each soil segment. Pre 1950 ages of moss cellulose and bulk samples were obtained using the InCal09 calibration dataset in Calib 6.0 (Stuiver & Reimer, 2010) We added 3 yea rs to modern segment ages to correct for time it takes newly fixed C in live green moss to become brown moss detritus (Osterkamp et al ., 2009) For moss and bulk samples, we assigned the age to the midpoint of each segment. To calculate C accumulation rat es, we fitted the 14 C age model described in Trumbore & Harden (1997). This model assumes a net change in C storage ( dC/dt ) is a balance between annual C inputs ( I) and decomposition ( kC ) : ( 2. 1) When solved, this equation becomes: ( 2. 2)

PAGE 29

29 where C 0 is the initial C pool (g C m 2 ), C t is the inventory of C in a given core as of year t (g C m 2 ; the year sampled (2004) minus the calendar age), I is the annual C inputs (g C m 2 y 1 ), and k is the first order decomposition rate constant ( y 1 ). The model was fitted to plots of cumulative C inventory ( C t ) versus time for each core using the nls function in R (R Core Development Team 2011) to obtain estimates for I and k which were used to calculate turnover time and accumulation rates. The millennial C accumulation curves crossed the origin so the initial C pool was zero and the C 0 e kt term in Eq. 2. 2 dropped out. However, the decadal C accumulation curves did not cross the origin and needed a negative C 0 for the model to converge. While correct mathematically, conceptually this result was nonsensical because a negative initial C pool is not possible. To fit the decadal data, the model needed to be shifted to the ri ght on the x axis, so we dropped the C 0 e k t term and added an x intercept, t i to the model: ( 2. 3) This modified equation gave the same parameter estimates as Eq. 2 .2 without needing a negative initial C pool. The parameter t i can be interpreted as the number of years before sampling (2004) that measurable C accumulation stopped. For cores one, two, five, and six, the oldest data points were dropped from the decadal analyses because they prevented model convergence. These data points (from 5 7 or 7 9 cm segments) were likely sampled too coarsely at depth increments that included the increasing slope 14 C could dominate 14 C signature leading to erroneously young e stimated ages.

PAGE 30

30 Estimates of decadal NEP were calculated by subtracting decomposition of deep soil from decadal C accumulation rates (Trumbore et al., 1999) : ( 2. 4 ) w here I decadal and k decadal are the C input rate and decomposition constant from the decadal C accumulation model, C shallow is the C pool in the surface soil, k millennial is the decomposition constant from the millennial C accumulation model, and C deep is t he C pool in the deep soil. For Eq. 2. 4 and decadal C accumulation rate calculations, C shallow is the amount of C from the soil surface to the end of the depths used to fit parameters in Eq. 2. 2, which varies by core (4 cm in core 1, 5 cm in cores 2, 5, an d 6, and 7cm in cores 3 and 4). For Eq. 2. 4 and millennial C accumulation rate calculations C deep is the amount of C from the end of the modeled shallow depths until the start of permafrost because we assumed minimal microbial activity within frozen soil. In order to calculate NEP for cores one, four, five, and six, an average of the millennial decomposition constants was used. Results Soil Carbon Invent ories Carbon pools at all sites were large, exceeding 50 kg C m 2 in the first meter. Carbon pools in both the top 1 m and the active layer of soils were greatest in moderate thaw but the effect was only marginally significant for the top meter (Table 2 1, p=0.062 and p=0.011, respectively). Minimal thaw had the least amount of C stored in the organic soil horizons, but again the effect was only marginally significant (Table 2 1, p=0.059), and minimal thaw had a shallower organic horizon than extensive thaw (Table 1 2 p=0.016). Active layer depths ranged from 62 cm in minimal thaw to 73 cm in extensive thaw (Table 2 1).

PAGE 31

31 Soil Profiles All cores showed a significant increase in bulk density with depth, but bulk density did not differ among sites (Table 2 2, t wo way ANOVA, depth effect, p<0.0001, site effect p=0.69). Similarly, %C decreased significantly with depth but did not differ among sites (Table 2 2, two way ANOVA, depth effect p<0.0001, site effect, p=0.53). Carbon percentages increased in most cores wi thin the mineral soil, sometimes above 20%, the threshold for classification as an organic soil, indicating buried organic layers (Fig. 2 1). At all sites, %N was greatest in the bottom of the organic layer and least in the mineral layer (Table 2 2, two wa y ANOVA, depth effect p<0.0001, site effect, p=0.94). Site by depth interactions were not significant. For C:N ratios, there were no differences among sites (Fig. 2 2, two way ANOVA, site effect, p=0.48) but there were significant differences among depths (depth effect, p<0.0001). Within organic layers, C:N ratios were highest i n the 0 5 cm and 5 15 cm layers, lowest in the 25 35 cm and 35+ cm layers, and constant within mineral layers. 13 C w ith minimal thaw being more enriched than extensive thaw (Fig. 2 2, two way ANOVA, site effect, 13 C was most enriched in the bottom organic layers and most depleted at the top of the organic and in the mineral layers (Fig. 2 2, two way 15 N did not differ among sites (Fig. 2 2, two way ANOVA, site effect, p=0.77) but did differ with depth (depth effect, p<0.0001) becoming more enriched from the 0 5 cm to 15 25 cm organic layer, after which it re mained constant through the mineral layers.

PAGE 32

32 C arbon Accumulation Rates 14 C in surface and deep soils were used to date soil core segments for decadal and millennial C accumulation models, respectively (Fig. 2 3). Radiocarbon values peak within the first 5 cm of each core, which generally corresponds to the bomb peak of 1963 the year of increased atmospheric nuclear weapons testing before the test ban went into effect (Levin & Hesshaimer, 2000) 14 C peak in each surface profi le are similar between cores sampled from the same site but differ 14 C peaks closest to the surface, extensive 14 14 C peaks of moderate thaw cores fall in between (Fig. 2 3A). 14 C profiles (Fig. 2 3B) 14 C values at the bottom of the core correspond to calendar ages of 8,000 to 10,000 years ago. Core two has a 14 C that is likely due to cryoturbation. Modeled decadal C accumulation rates for EML soils ranged from 6.84 to 41.6 g C m 2 y 1 (Table 2 3, Fig. 2 4). Carbon inputs ranged from 42.3 to 140 g C m 2 y 1 The decomposition rate consta nt ( k ) ranged from 0.031 to 0.064, excluding core two. Turnover times were decadal ranging from 16 to 32 years, excluding core two. The x intercept, t i ranged from 7.5 to 18 years before sampling. The average decadal C accumulation rate, C inputs, and tur nover times for EML soils were 25.8 g C m 2 y 1 87 g C m 2 y 1 and 25 years (excluding core two), respectively. Millennial scale C accumulation rates were an order of magnitude less than decadal rates, ranging from 2.2 to 2.4 g C m 2 y 1 (Table 2 3, Fig. 2 5). Carbon input rates were similar in both cores 12 to 13 g C m 2 y 1 less than, but within the same order of magnitude as, recent C input rates. The decomposition constants for these

PAGE 33

33 millennial accumulation rates were two orders of magnitude less tha n decadal accumulation rates leading to turnover times on the order of three thousand years (Table 2 3). Decadal NEP at EML averaged 14.4 g C m 2 yr 1 but ranged over an order of magnitude (Table 2 4, Fig. 2 6). Discussion C arbon Inventories and Depth Profiles Carbon inventories in the top meter at EML ranged from 55 to 69 kg C m 2 within the upper half of the 16 to 94 kg C m 2 range reported for similar tundra soils across Alaska (Michaelson et al ., 1996) Active layer carbon po ols and the percentage of the 1m C inventory exposed by the active layer (63 to 71%) were also within ranges for Alaskan and Canadian arctic soils (Michaelson et al ., 1996; Hugelius et al ., 2010) The permafrost at EML held a lot of C due to high C concent rations (20 30%) within the affected soils because organic material is mixed into deeper horizons by cryoturbation where it becomes protected from further decomposition (Ping et al ., 2010) The active layer C pool was larger in moderate thaw than in minimal thaw, but did not increase further in extensive thaw because extensive thaw had a lower C density (C per gram of soil). the 1 m C pool contained in the active layer at EML increased from 63% in minimal thaw to 71% in extensive thaw active layer that was a similar depth to minimal thaw, an addi tional 6.4 kg C m 2 has been made available to above freezing decomposition due to permafrost thaw. 13 15 N became more enriched with depth, indicating increasing degrees of decomposition ( Fig 2; Kuhry

PAGE 34

34 & Vitt, 1996; Bostrom et al ., 2007; Hobbie & Ouimette, 2009) Carbon to nitrogen ratios decrease over the decomposition process because C is lost from the soil as it is mineralized, while N, though it may change form, generally stays within the soil (Malmer & Holm, 1984; Sterner & Elser, 2002) Loss pathways of C and N from soil typically favor lighter isotopes, so more heavy isotopes relative to light isotopes remain over time as organic matter is decomposed (Bostrom et al ., 2007; Hobbie & Ouime tte, 2009) In the organic horizon, minimal thaw was more 13 C enriched than extensive thaw indicating they have different soil organic matter sources, not that decomposition is 15 N did not differ. Minimal thaw is dom inated by 13 C of 13 C around (Schuur et al ., 2007 ; unpublished data ) Within the 13 15 N remained constant, indicating little additional decomposition was occurring in this cryo protected layer all but the top of which is within permafrost. This lack of decomposition in the permafrost leads to the buildup of deep soil C that is 8,000 to 10,0 00 years old. C arbon Accumulation Rates Eight Mile Lake tundra soils have substantial decadal and millennial rates of C accumulation indicating they have been active C sinks from the early Holocene through the most of the past fifty years (Table 2 3). Considering the six cores across the gradient together, EML soils had decadal C inputs within ranges reported for nutrient poor bogs and black spruce forests (Trumbore & et al ., 2011) Decadal decomposition rate constants ( k ) we re generally within previously reported values for boreal forest sites with moss cover, which range from 0.0001 to 0.045 y 1 (Trumbore & et al ., 2011) Carbon accumulation rates in EML

PAGE 35

35 soils also fell within values for boreal forest s and wetlands, which range from 3 to 260 g C m 2 yr 1 (Trumbore & et al ., 2011) The similarities among C input rates, decomposition constants, and C accumulation rates across these subarctic ecosystems likely result from their sim ilar poorly drained surface soils dominated by recalcitrant mosses and underlain by shallow permafrost. Millennial C accumulation rates in EML tundra soils were similar to deep soil C accumulation rates in boreal wetland and forest soils but deep soil C inputs and decomposition constants differed ( Trumbore & et al ., 201 1 ) EML had lower C inputs to deep soils than Canadian boreal forest s and wetlands ( Trumbore & Harden, 1997) Lower C inputs to deep soils may be due to the shallowe r active layer at EML than at boreal forests and wetland s which cause s cryoturbation, not root growth, to be the main mechanism transporting C from the surface to deep soil. The shallower active layer also explain s the lower deep soil decomposition constants at EML because more deep soil is cryoprotected. In contrast, EML had deep soil C inputs an order of magnitude higher than mineral and permafrost soils of an Alaskan black spruce forest but had similar deep soil decomposition const ants et al ., 201 1 ) EML soils had a larger deep C pool than this Alaskan black spruce forest ( about 34 kg C versus 6 to 10 kg C) and therefore more deep soil decomposition overall, which is how EML could have similar millennial C accumulation ra tes (about 2 g C m 2 yr 1 ) despite having much greater C inputs. Carbon input and decomposition rates are related wherein greater C inputs coincide with larger decomposition losses at individual sampling points across this landscape (Fig. 2 7). Carbon inp uts in this model are analogous to net primary

PAGE 36

36 production (NPP) of the tundra ecosystem. Environmental conditions that favor greater plant growth such as warm soil and air temperatures, moderate soil moisture, and high soil nutrient availability also favor greater microbial activity and therefore faster soil decomposition (Shaver et al ., 1992; Chapin et al ., 2002; Raich et al ., 2006) Conversely, low temperatures, dry or excessively wet soils, and low nutrient availability cause both decreased plant growth and microbial activity. This positive relationship appears to constrain variation in C accumulation rates across the landscape because high inputs are likely to be balanced by faster decomposition rates. Net Ecosystem Production Decadal NEP estimates calc ulated with parameters from decadal and millennial C accumulation models (Eq. 2. 4 ) can be compared to annual NEP estimates calculated from ecosystem gas exchange measurements as way to independently evaluate our model estimates. Decadal NEP calculated with this dataset averaged 14.4 g C m 2 yr 1 (Table 2 4) These fluxes were within the range of decadal NEP at boreal forest sites, though much less than the maximum rate of 180 g C m 2 yr 1 calculated for fen sites (Trumbore, 1997; Trumbore et al ., 1999) Decadal estimates of NEP were similar to current annual estimates of NEP made with gas exchange measurements at EML, which ranged from 58 to 57 g C m 2 yr 1 from 2004 to 2006 (Vogel et al ., 2009) Furthermore, estimates of the percentage of heterotrophic respiration coming from deep soil derived from the decadal NEP equation (10 to 30%) match previous estimates (7 to 23%) derived from partitioning respiration using radiocarbon measurements of ecosystem gas exchange (Schuur et al ., 2009) While the mean of decadal NEP estimates was positive indicating over the past five decades these soils were a C sink, it appears that surface C accumulation stopped

PAGE 37

37 8 to 18 years ago. This estimate of 8 to 18 years is from the t i parameter needed to fit the decadal C accum ulation data. Th is deviation from conditions of steady accumulation may be interpreted as either a halting of recent soil C accumulation within surface soils (inputs balanced by outputs) or a shift of C inputs from vertical accumulation at the soil surface to deeper in the soil. The latter explanation would be consistent with surface moss growth being replaced by belowground vascular plant growth. In fact, a shift to belowground NPP could also explain the outlier points in our decadal C accumulation data. T he outlier layers were either younger than expected or had more C than expected by the average accumulation rate over the entire profile. Root biomass peaks within the 5 15 cm depth from which these outlier segments came (unpublished data), but outliers we re found in only four of the six cores, while surface accumulation stopped in all cores. The former explanation is supported by observed negative flux derived annual NEP at all EML sites in 2005 and at extensive and minimal thaw in 2004 (Vogel et al ., 2009 ) If there were enough negative NEP years to counteract positive NEP years over the past two decades, then C inputs would roughly balance C outputs and C accumulation would be negligible. T he estimate that vertical C accumulation halted within the past tw o decades coincides with an era of permafrost warming at EML (Osterkamp et al ., 2009 ). The cessation of vertical accretion may also be due to methodological constraints whereby we did not have fine enough resolution at the top of the cores to measure recen t moss growth. Overall, the deviation from steady vertical C accumulation after millenia indicates the C cycle in this tundra ecosystem is undergoing major change in response to permafrost warming and thaw Tundra soils that were once a significant C sink are likely

PAGE 38

38 becoming a C source. Soils from minimal, moderate, and extensive thaw all had positive decadal and millennial C accumulation rates demonstrating they have been accumulating C since the start of the Holocene (Fig. 2 5). However, according to ann ual flux derived estimates of NEP, all sites have been observed to be C sources during at least one year in the past decade, and based on a three year average, extensive and minimal thaw are C sources (Vogel et al ., 2009) Even one of our decadal NEP estim ates (core 1) was negative indicating that decomposition in the deep soil has been outpacing C i nputs from the surface soil Furthermore, our decadal accumulation models indicate vertical soil C accumulation halted within the past 18 years. Thus, this shif t in the C balance of these tundra soils is leading to increased atmospheric CO 2 concentrations through the microbial release of soil C to the atmosphere (Schuur et al ., 2009) and through the loss C uptake these soils used to provide.

PAGE 39

39 Table 2 1. Mean (st andard error) organic soil depth active layer depth and C inventory in the top 1 m, in the organic horizons, and in the active layer for minimal, moderate, and e xtensive permafrost thaw sites (n=6) Letters not shared among sites indicate significant dif ferences (one Site Org d epth Active l ayer 1 C pool to 1m Org C pool Active l ayer C pool cm cm kg C m 2 kg C m 2 kg C m 2 Minimal 37 (3.1) a 62.2 (1) 55.7 (4.1) a 18.6 (2.3) a 35.0 (1.5) a Moderate 49 (5.9) ab 65.8 (1) 68.5 (5.3) a 33.0 (5.8) a 45.8 (2.5) b Extensive 54 (4.8) b 72.6 (0.8) 54.6 (3.0) a 28.4 (4.4) a 40.1 (2.1) ab 1 2009 Data from Trucco et al. ( 2012)

PAGE 40

40 Table 2 2. Mean (standard error) bulk density, % C, and % N for three soil depths at m inimal, moderate, and e xtensive permafrost thaw sites (n=6). The deeper organic layer was labeled 35+ becaus e the depth of the organic horizon varied by core. Letters not shared among depths indicate significant differences (two differences among sites. Site Depth Bulk d ensity g cm 3 Carbon % Nitrogen % Minimal 0 15 cm Org 0.070 (0.007) a 42.2 (0.69) a 0.93 (0.04) a 15 35+ cm Org 0.17 (0.02) b 37.4 (0.60) b 1.59 (0.06) b Mineral 0.53 (0.07) c 12.6 (2.7) c 0.49 (0.1) c Moderate 0 15 cm Org 0.071 (0.005) a 42.1 (0.30) a 0.80 (0.07) a 15 35+ cm Org 0.22 (0.02) b 37.6 (1.2) b 1.68 (0.06) b Mineral 0.46 (0.05) c 13.6 (2.4) c 0.52 (0.09) c Extensive 0 15 cm Org 0.056 (0.006) a 41.6 (0.26) a 0.88 (0.08) a 15 35+ cm Org 0.18 (0.02) b 35.2 (1.4) b 1.49 (0.04) b Mineral 0.46 (0.05) c 12.4 (2.2) c 0.57 (0.1) c

PAGE 41

41 Table 2 3. I (C inputs), k (decomposition constant), and t i (x intercept) parameters fitted to Eq. 2. 3 (decadal) or Eq. 2. 2 (millennial) with their standard errors. Turnover times (TT) and C accumulation rates (C acc) were calculated from the fitted parameters. I and k have upper and lower standard errors that are not symmetrical around the estimate because they were fitted as natural logs to avoid placing non zero bounds on the parameters. Cores 1 and 2 are from minimal thaw, cores 3 and 4 are from moderate thaw, and cores 5 and 6 are from extensive thaw. Decadal C accumulation rates are calculated from 0 to end of the mode led depth (4 cm in core 1, 5cm in cores 2, 5, and 6, and 7cm in cores 3 and 4). Millennial rates are calculated from the end of depths modeled for decadal accumulation to the bottom of the active layer (Table 2 1). Core I Upper Lower K Upper Lower t i TT C acc g C m 2 yr 1 SE SE yr 1 SE SE year year g C m 2 yr 1 Decadal 1 83.0 110 63 0.050 0.07 0.03 7.48 (1.7) 19.9 6.84 2 42.3 140 13 0.0015 1 0.00 17.0 (7.3) 669 39.9 3 140 150 130 0.064 0.07 0.05 8.40 (0.55) 15.7 14.3 4 109 150 80 0.032 0.06 0.02 12.3 (2.6) 31.5 41.6 5 43.1 56 33 0.031 0.06 0.02 10.2 (1.5) 32.2 14.3 6 105 512 21 0.041 0.65 0.001 18.0 (9.0) 24.5 37.9 Millennial 2 13 16 11 0.00032 0.00041 0.00025 3090 2.19 3 12 19 8.0 0.00030 0.00057 0.00016 3330 2.39

PAGE 42

42 Table 2 4. Decadal Net Ecosystem Production (NEP) and the amount of shallow soil C (C shallow ) and deep soil C (C deep ) used to calculate decadal NEP with I and k values from Table 2 3 (Eq. 2. 4) for each soil core. The shallow and deep C inven millennial C accumulation rates, respectively. Core C shallow kg m 2 C deep kg m 2 NEP g C m 2 yr 1 1 1.52 28.0 1.98 2 1.66 34.0 28.9 3 1.97 33.4 4.08 4 2.13 41.9 28.4 5 0.93 38.6 2.12 6 1.64 40.9 25.0

PAGE 43

43 Figure 2 1. Depth profiles for each core taken at minimal, moderate, and extensive permafrost thaw sites. The depth is the middle of the layer in which %C was measured. Note the high %C at depth, which often exceeds 20 % (black vertical lines, the threshold used to categorize organic from mineral soil) and indicates cryoturbation of organic soil into mineral horizons Not all profiles are the same length because of variance in the depth where small rocks inhibited further coring

PAGE 44

44 Figure 2 2. 13 15 N of organic (top) and mineral (bottom) soil layers at the minimal, moderate, and extensive thaw sites. The depths start over again for the mineral layer because the organic layer does not have a uniform thick ness. Error bars represent the standard error. n=6 in organic layers but varies from two to six in mineral layers due to varying core lengths. Depths that do not share a letter are significantly different (two The only significant diffe rence among sites was that minimal thaw was more 13 C enriched than extensive thaw.

PAGE 45

45 Figure 2 3. Depth profiles of radiocarbon values used to calculate ages for decadal (A) and millennial (B) C 14 C found in the surfac e profiles show where radiocarbon from the nuclear weapons testing bomb peak (circa 1963) was fixed. Negative radiocarbon values found in deep profiles indicate the radiocarbon has undergone radioactive decay since 14 C at the bottom of the cores (B) correspond to calendar ages of 8 000 to 10,000 years ago.

PAGE 46

46 Figure 2 4. Surface c umulative C inventories versus the age of C in that soil layer. All points were used to fit the accumulation model (Eq. 2. 3 ) except for the oldest point(s) (ope n circles) in cores 1, 2, 5, and 6, which were excluded from the analyses (see text). Curves are the predicted values calculated from estimated I k and t i values (Table 2 3). Cores 1 and 2 are from minimal thaw, cores 3 and 4 are from moderate thaw, and cores 5 and 6 are from extensive thaw.

PAGE 47

47 Figure 2 5. Cumulative C inventories versus the age of C in that soil layer. All points were used to fit the accumulation model (Eq. 2. 2) except for one point from core 2 (as indicated by the open circle) which was excluded from the analysis as an outlier due to cryoturbation. Curves are the predicted values calculated from estimated I and k values (Table 2 3).

PAGE 48

48 Figure 2 6. Decadal Net Ecosystem Production (NEP) for two cores at each site in the thaw gradient and annual NEP calculated from recent flux measurements at the thaw gradient averaged over three years (Vogel et al., 2009). Positive NEP indicates the ecosystem is a C sink while negative NEP indicates the ecosystem is a C source.

PAGE 49

49 Figure 2 7. The relationship between C input ( I Table 3) and decomposition rates ( kC Tables 3 and 4) for the six surface soils (decadal rates) and the two deep soils (millennial rates, bottommost points).

PAGE 50

50 CHAPTER 3 THAWING PERMAFROST I NCREASES OLD SOIL AN D AUTOTROPHIC RESPIRATION IN TUNDRA : PARTITIONING ECOSYSTEM RESPIRATIO N USING 1 Abstract Ecosystem respiration (R eco ) is one of the largest terrestrial carbon (C) fluxes. The effect of climate change on R eco depends upon the responses of its au totrophic and heterotrophic components. How autotrophic and heterotrophic respiration sources respond to climate change is especially important in ecosystems underlain by permafrost. Permafrost ecosystems contain vast stores of soil C (1672 Pg) and are loc ated in northern latitudes where climate change is accelerated. Warming will cause a positive feedback to climate change if heterotrophic respiration increases without corresponding increases in primary production. We quantified the response of autotrophic and heterotrophic respiration to permafrost thaw across the 2008 and 2009 growing seasons. We partitioned R eco using 14 C and 13 C into four sources two autotrophic (above and belowground plant structures) and two heterotrophic ( young and old soil ) We sampled the 14 C and 13 C of sources using incubations and the 14 C and 13 C of R eco using field measurements. We then used a Bayesian mixing model to solve for the most likely contributions of each source to R eco Autotrophic respiration ranged from 40 to 70% of R eco and was greatest at the height of the growing season. Old soil heterotrophic respiration ranged from 6 to 18% of R eco and was greatest where permafrost thaw was deepest. Overall, growing season fluxes of autotrophic and old soil heterotroph ic respiration increased as permafrost thaw deepened. Areas with greater 1 This chapter is currently available on the Global Change Biology website as : Hicks Pries CE, Schuur EAG, Crummer KG ( in press ) Thawing permafrost increases old soil and autotrophic respiration in tundra: 13 14 C Global Change Biology doi: 10.1111/gcb.12058.

PAGE 51

51 thaw also had the greatest primary production. Warming in permafrost ecosystems therefore leads to increased plant and old soil respiration that is initially compensated by increased net primary productivity. However, barring large shifts in plant community composition, future increases in old soil respiration will likely outpace productivity, resulting in a positive feedback to climate change. Introduction Ecosystem respiration (R eco ) is the largest carbon (C) flux from the terrestrial biosphere to the atmosphere (Raich & Schlesinger, 1992) Thus, understanding the response of R eco to climatic changes is critically important to making predictions about the C cycle on local, regional, a nd glob al scales. However, measured responses of R eco to temperature increases have been highly variable (Davidson & Janssens, 2006) because R eco is a combination of respiration by autotrophs and heterotrophs, which often respond differently to changes in climate (Borken et al. 2006; Muhr & Borken, 2009; Gomez Casanovas et al., 2012) The relative responses of autotrophic (R a ) and heterotrophic respiration (R h ) to climatic change s affect the C balance of ecosystems: on short timescales, R a is ge nerally balanced by current production but R h does not have to be. Their relative responses are particularly important in permafrost ecosystems, which have historically been C sinks (Hicks Pries et al. 2012) and have the potential to cause a large positiv e feedback to climate change (Schuur et al. 2008) Soils in the permafrost zone store 1672 Pg C, over twice as much C as the atmosphere currently holds (Schuur et al. 2008) because frozen soil has protected organic C from decomposition. These p ermafros t soils are found mainly in high latitudes where up to 7 C temperature increases are predicted over the next century ( IPCC, 2007 ). In Alaska, some p ermafrost is already thawing downward at a rate of 0.1 to 0.9

PAGE 52

52 m year 1 (Osterkamp & Romanovsky, 1999; Osterkamp, 2007) As permafrost thaw s soil organic C is e xposed to microbial degradation, increasing R h (Goulden et al. 1998; Schuur et al. 2009) However, warmer, thawed soils can also increase biomass C storage and R a by increasing nutrient availabil ity and causing plant communities to shift to larger growth forms ( e.g. shrubs ; Schuur et al ., 2007 ). With CO 2 flux measurements, researchers have shown permafrost thaw increases R eco in tundra (Vogel et al. 2009) and peatlands (Dorrepaal et al. 2009) However, measuring fluxes alone does not reveal whether the increase is being driven by R a or R h I ncreases in R eco as a result of thaw may indicate different outcomes of an respiration source drives the change. For e xample, if R h of old soil C is driving the increase, the system is losing C that had been stored for hundreds to thousands of years to the atmosphere, likely resulting in a positive feedback to climate change. If R a is driving the increase, the system is either turning over newly photosynthesized C faster, a neutral or constrained positive feedback to climate change, or is fixing more C, a negative feedback to climate change. To help predict the strength of the perma frost thaw feedback, autotrophic and heterotrophic contributions to R eco must be known. Partitioning R eco into its sources is necessary for a mechanistic understanding of how respiration responds to climat e change, and many partitioning methods have been developed. Natural abundance 13 C or 14 C are used for partitioning R eco when their values differ among respiration sources. Isotope partitioning is less destructive than methods like trenching or girdling, which can change environmental conditions (e.g.,L uan et al. 2011; Subke et al. 2011), and may cause less sampling artifacts than

PAGE 53

53 methods like excising roots from soil to measure fluxes separately (Kuzyakov, 2006; Yi et al. 2007) M any studies have used either 13 C or 14 C to estimate source contributi ons to soil or ecosystem respiration (Ehleringer et al. 2000; Gaudinski et al. 2000; Trumbore, 2000; Ngao et al. 2005; Schuur et al. 2009) In permafrost ecosystems, deep soil contributions to R eco 13 C (Dorrepaal et al. 2009) 14 C (Schuur et al. 2009) However, source contributions can be determined with increased accuracy using both carbon isotopes simultaneously (Phillips & Gregg, 2003) Combining C isotopes is powerful because 13 C and 14 C separate sources based on different principles 13 C via biological fractionation and water relations (Bowling et al. 2002) and 14 C via age (Trumbore, 2000) 13 C differs among sources because many enzymatic processes, like C fixation by Rubisco, discriminate against the heavy i sotope. 13 C also varies among autotrophs due to different photosynthetic strategies, water relations, and CO 2 concentrations because the less C limited plants are, the more they discriminate against 13 C (Dawson et al. 2002) 14 C acts as a timestamp once isotopic fract ionation effects have been corrected for b ecause 14 C undergoes radioactive decay. Further separating sources based on age is the 1963 14 C bomb peak caused by atmospheric nuclear weapons testing which causes C fixed in the past 50 years to have 14 C (Levin & Hesshaimer, 2000) Despite the potential of dual isotope partitioning, both C isotopes have rarely been used together in natural abundance studies (Mayorga et al. 2005; Billett et al. 2007; Hardie et al. 2009) The main obje ctive of this study was to quantify the response of plant and microbial respiration to permafrost thaw and seasonality 13 14 C to

PAGE 54

54 partition R eco into four sources : aboveground plant structures (AG) belowground plant structur es (BG), you ng soil (YS, 0 15 cm), and old soil (OS, 15 80 cm) Specific objectives were to: 1) Determine spatial and temporal variability in source and R eco 13 C 14 C ; 2) D etermine whether permafrost thaw increases the contribution of R h from old soil C; 3) Determine how the contributions of R a and its components, AG and BG plant respiration, change seasonal ly ( May through September ) and with thaw. This mechanistic approach to understanding how the components of R eco respond to seasonality and permafrost thaw will increase our understanding of permafrost can be used to parameterize C cycling models, increasing our capacity to predict the future state of the Earth system. Materi als and Methods Site Description Our site is tundra in Healy, Alaska. The vegetation is moist acidic tussock tundra underlain by soils that have permafrost within a meter of the surface (Geliso ls). The soils consist of about 0.5 m of organic soil on top of mineral soil that is a mixture of loess deposits and glacial till (Vogel et al. 2009). The water table is usually 15 25 cm below the soil surface (Trucco et al. 2012), and so methane product ion out of the ecosystem is negligible (unpublished data). Permafrost temperatures in this region are around 1C and therefore the permafrost is susceptible to thaw (Osterkamp & Romanovsky, 1999) Within t he study site, some areas have undergone active l ayer thickening and thermokarst formation due to permafrost thaw (Vogel et al. 2009) T haw has been documented for the past two decades at this site but likely began earlier (Osterkamp et al. 2009) This site has had

PAGE 55

55 ongoing monitoring of soil temperatur e, active layer depth, water table depth, and CO 2 fluxes since 2004 (Schuur et al. 2009; Vogel et al. 2009; Trucco et al., 2012) Ecosystem Respiration To measure the 13 C 14 C of R eco we installed 12 permanent PVC collars (25.4 cm diameter x 10 cm deep) 8 cm into the soil across 1 km of the study site For sampling R eco 10L dark chambers (13 cm high) were fit onto the collars over the soil and encompassing the aboveground plant biomass. Ecosystem respiration 1 3 C and 14 C was sampled over a week in June, July, and August 2008 and May, July, and September 2009. Sampling occurred from 6 to 11 am to limit diurnal variation and ensure calm conditions. After each set of R eco samples were taken, thaw depth was measure d twice adjacent to and once within each collar using a metal probe (2 mm in diameter) pushed into the ground until it hit resistance. Active layer depth (AL), the thaw depth at the end of the growing season and a measure of the permafrost thaw extent, was measured in September and used to stratify collars into three categories for the partitioning model. We used Keeling plots to measure the 13 C of R eco wherein we took air samples from the chamber every 2 to 3 minutes while pCO 2 increased for a total of se ven samples (Keeling, 1958) The air samples were collected into exetainers (septa topped vials, Labco Limited, Lampeter, UK) in line with the chamber, a pump and an infrared gas analyzer (IRGA; LI 820, LI COR, Lincoln, Nebraska). We recorded the pCO 2 from the IRGA when each exeta iner was removed from the line. The exetainers were sent back to the University of Florida to be run on a GasBench II coupled to a Finnigan Delta Plus XL stable isotope ratio mass spectrometer (precision Their holding time was no longer than 10 days, and the majority of the samples were run

PAGE 56

56 within seven days (Mortazavi & Chanton, 2004; Midwood et al. 2006) Standard samples of similar pCO 2 and a 13 C of Oztech Trading Corporation, Safford, AZ, USA ) wer e sent with each batch of exetainers to correct for changes in 13 C due to travel and storage. The 13 C and 1/pCO 2 were fit with a linear 13 C of R eco 14 C, we first removed as much atmospheric CO 2 from the chamber as possible by pumping chamber air through soda lime for 45 minutes while maintaining ambient pCO 2 (Schuur & Trumbore, 2006) After scrubbing, we pumped chamber air through a zeolite molecular sieve (A lltech 13X Alltech Associates, Deerfield, IL, USA ) trap that quantitatively adsorbs CO 2 for 15 minutes (Hardie et al. 2005) By maintaining pCO 2 around ambient levels, we removed CO 2 from the chamber at roughly the same rate it was fluxing out of the eco system and avoided an unnatural CO 2 concentration gradient. Sampling only occurred under calm wind conditions to minimize atmospheric CO 2 entering chambers through the soil. The molecular sieve traps were baked at 625 C to desorb CO 2 (Bauer et al. 1992) which was purified using liquid N 2 on a vacuum line and reduced to graphite by Fe reduction in H 2 (Vogel et al. 1987) T he graphite was sent to the UC Irvine W.M. Keck Carbon Cycle Accelerator Mass Spectrometry (AMS) 14 C analysis (precision 14 C data are reported at the same 13 C value to correct for mass dependent fractionation effects. 14 C data were corrected for the atmospheric CO 2 remaining in the chambers 13 C data from each chamber in a 2 pool (atmosphe ric and R eco ) mixing model as in Schuur & Trumbore (2006)

PAGE 57

57 Autotrophic Respiration Short term incubations were used to measure the 13 C 14 C of AG and BG R a Autotrophic respiration 13 C was measured from nine randomly chosen sites in June and August 2008 and in May, July, and August 2009. Aboveground and BG R a 14 C was measured in July 2008 and in May and July 2009 from three sites. To measure the 13 C 14 C of R a we collected plants from a randomly placed 20 cm 2 quadrat; we clipped all the aboveground material (including lichens and mosses) to the soil surface, and collected all live roots and rhizomes (>2 mm in diameter) from th e thawed soil. Aboveground samples were immediately plac ed into foil covered mason jars (0.24 L) while belowground samples were rinsed twice in water to remove soil particles and shaken dry before being put into foil covered mason jars. We incubated plants as soon as possible after clipping (within 5 minutes fo r AG and 30 for BG), since the 13 C of excised root respiration can change slightly after 40 minutes (Midwood et al. 2006) Air from the sealed mason jars was then pumped through soda lime for 5 minutes at 1 liter per minute to remove CO 2 from the headspa ce before starting a 5 1 0 minute or four hour incubation (until pCO 2 13 C 14 C analys e s, respectively). At the end of the incubation, headspace air was pumped into a Helium flushed exetainer for 13 C analysis or a molecular sieve trap for 14 C analysis. Heterotrophic Respiration To measure the 13 C 14 C of R h nine surface soil cores (0 25 cm) were randomly sampled in May, July, and August 2009, and 12 deep soil cores (25+ cm) were sampled in May 2009. We samp led surface soil more often to test if respired 13 C 14 C change throughout the growing season due to rhizosphere processes. We manually cored surface soils using a serrated knife to the depth of thaw or 25 cm,

PAGE 58

58 whichever came first, and sectioned them into 0 5 cm, 5 15 cm, and 15 25 cm depths. For deeper soils (25+ cm) we used a Tanaka TIA 340 permafrost drill with carbide bits to drill through frozen soil down to about 80 cm below which gravel impeded coring Surface soil incubations were started the day of sampling, while deep soils were kept frozen until February 2010 when they were thawed, cut into 10 cm sections, and incubated. R oots (>1 mm in diameter) were removed from all soil sections before soils were put into mason jars (0.95 L) for incubati ons Care was taken to minimize disturbance to the soil structure, which was a minor problem for mineral soils as they had very few roots. Surface s oils sat at room temperature for five days before 13 C and 14 C sampling to ensure the majority of labile C from small roots and root exudates decomposed and would not affect the R h isotope ratio As in previous studies (e.g., Schuur & Trumbore, 2006) we assumed the CO 2 respired from soils after 5 days was dominated by the heterotrophic flux and therefore inclu ded fast cycling rhizosphere C in the autotrophic flux. Soils were incubated at field moisture under aerobic conditions. Deep soils sat at room temperature for 10 days to allow microbial populations to stabilize after thaw During the two days preceding isotopic sample collection, three short term (3 hour) incubations were performed to measure rates of R h flux. Samples were run on an IRGA (LI 820) connected to an injection loop to measure pCO 2 The 13 C and 14 C sample colle ction was performed as for R a except that incubation times were based on the time it took for 1.5 mg C to accumulate in the headspace, which ranged from 12 to 72 hours Heterotrophic respiration was split into two sources, YS (0 15 cm) and OS (15 80 cm), based on soil 14 C values; YS included the top of the soil profile that contained

PAGE 59

59 post 14 C (Hicks Pries et al. 2012). To calculate 13 C 14 C of YS and OS, we weighted the isotopic signatures of each incubated depth by its CO 2 flux per g C field temperature using the Q 10 (Table A 1; Schuur & Trumbore, 2006). We calculated this weighted average for each replicate core separately and then averaged all cores to ob tain monthly mean 13 C 14 C for YS and OS. Before averaging, 13 C values were corrected for the incubation temperature shift as in Dorrepaal et al. (2009) This 13 C is depleted by 0.12 temperature rise (Andrews et al. 2000; Biasi et al. 2005) and our soils were incubated at temperatures warmer than field conditions. The same 14 C values from the 2009 cores were used to calculate 2008 R h end members. The effect of radioactive decay (<1 /year) and the addition of new OC with slightly depleted values (due to the Suess effect and biosphere mixing) on R h 14 C are likely below the 2 AMS (Schuur et al. 2009) A subset of 12 soil samples from various soil cores and d epths were used to calculate Q 10 and the 13 C temperature correction. We removed roots, homogenized, and split each sample into two jars one was incubated for 10 days at 2.5C and the other at 12.5C. The average CO 2 flux from three short term incubations were used to 10. For 13 C 12mL samples from the headspace were injected into vacuumed exetainers after a 6 hour incubation preceded by scrubbing CO 2 from the headspace. The average Q 10 of these soils was 2.5 and the average 13 C s hift was 13 C we used the following equation: (3.1)

PAGE 60

60 where the 13 C inc is the 13 C from incubating a soil depth section, T inc is the temperature of the incubation, and T field is the in situ field temperature of that soil depth from soil temperature sensors ( Table A 1; see Trucco et al., 2012 for sensor details). Data Analysis and Partitioning Model Ecosystem respiration was partitioned into AG, BG, YS, and OS using SIAR (Stabl e Isotope Analysis in R; Parnell et al. 2010) For partitioning, the 12 R eco collars were stratified into three categories based AL depth because SIAR gives more robust estimates when fitting parameters to a group of R eco values than to a single R eco valu e (Inger et al ., 2010). Six collars were classified as shallow AL (46 55 cm), three as intermediate AL (62 69 cm), and three as deep AL (80 103 cm). Active layer depth is a good indicator of permafrost thaw extent; ALs are deeper where the soil surface has subsided due to ground ice thaw (Osterkamp et al., 2009). Partitioning was performed separately for each AL category and each month sampled. The SIAR method uses Markov chain Monte Carlo to find possible solutions to this set of three equations: (3.2) w here the unknowns are f each source s proportional contribution to R eco 13 C 14 C of each source and R eco have known distributions The input data for this model include the mean and standard deviation of all source isotopic values (Tables 3 2 and 3 3) and the individual isotopic values of R eco collars measured for each AL category (Table A 2 ). For July 2008 partitioning, R a 13 C was an average of June and August 2008 values because July R a 13 C was not sampled. The results of the model

PAGE 61

61 are f While SIAR uses a Bayesian framework, we used uninformative priors because previous partitioning results are limited. For statistical analyses of source isotopes, we used one and two way analyses of va a only) and month as main effects, and core as a random effect (YS and OS only). For R eco isotopes and main effec ts and collar as a random effect. Analyses were done for 2008 and 2009 separately. To analyze how isotopes varied in the soil profile, we used a one way ANOVA with depth as the main effect and soil core as a random effect for deep soil cores and a two way ANOVA with month as an additional main effect for surface soil cores. One AL categories. We investigated the relationships between mean thaw depth of the AL categories and the mean sourc e contributions ( f ) using linear regressions with category as a random effect in R (R Core Development Team, 2012). Source contributions were logit transformed before analyses. All residuals were checked for normality and homogeneity of variances to ensure the assumptions of ANOVA and regressions were met. Respiration Fluxes Ecosystem respiration fluxes were sampled with static and auto chambers throughout the 2008 and 2009 growing seasons from plots adjacent to the isotopic sampling collars. The respiratio n measurements were part of a C balance study and are described in Vogel et al. (2009) and Trucco et al. (2012). We used the same data as presented in Trucco et al. (2012), except that we averaged the respiration plots into the

PAGE 62

62 three AL depth categories in stead of by site, to pair with our partitioning estimates (Table A 3). To estimate growing season respiration from each source, we combined each month of the growing seaso n (May through September; we averaged 2008 and 2009 July values) and multiplied the proportions by their corresponding mean flux (for each month and AL category). We then summed the flux of each source across all growing season months by AL category. This first approximation of a growing season flux includes some uncertainties because our isotopic sampling reflected only the big changes over the growing season. However, sampling respiration at low frequencies has been shown to accurately capture seasonal va riation (Savage & Davidson 2003). Logistical constraints prevented frequent isotopic sampling, so we did not focus on shorter timescales. Results Ecosystem Respiration Ecosystem respiration collars were stratified into three categories based on their AL d epth. Among these AL categories, thaw depths differed significantly throughout the growing season in 2008 and 2009 (Table 3 1; repeated measures ANOVAs, p<0.005). Thaw depths significantly increased throughout each growing season (Table 3 1; repeated measu res ANOVAs, p<0.0001) and differences among categories became larger throughout the growing season in both years (category x month interaction, p<0.028). In both 2008 and 2009, R eco 13 C did not differ among AL categories (Table 3 1; repeated measures ANOVA, p>0.60) but did differ among months (p<0.056). There were significant AL category differences in R eco 14 C. Shallow AL R eco was significantly

PAGE 63

63 14 C than deep AL R eco in 2009 ( Fig. 3 1, repeated measures ANOVA, p=0.035), but the effec t was marginally significant in 2008 (p=0.082). Early growing season R eco 14 C than mid and late season R eco a difference which was significant in 2008 ( Fig. 3 1, repeated measures ANOVA, p=0.0007) and marginally significant in 2009 (p =0.064). Across all categories and years, there is a negative linear relationship between thaw depth and R eco 14 C (R 2 =0.37, p<0.0001), but no relationship between thaw depth and R eco 13 C (R 2 =0.01, p=0.44). Source Respiration Aboveground R a wa 13 C than BG R a (Table 3 2; two way ANOVAs, p<0.0001). Early growing season R a 13 C was more enriched than later growing season R a in both 2008 and 2009 (Table 3 2; two way ANOVAs, p<0.0001). In 2009, AG R a was 2 13 C in July than in May or September ( Table 3 2; two way ANOVA, type x month interaction, p=0.043). 14 differences among AG and BG structures or month sampled ( Tab le 3 2; two way ANOVAs, p>0.4). 13 C became more enriched with depth 14 C became more depleted with soil depth, 13 C and 14 C increased with depth ( Fig. 3 2 ). Statistical analyses of R h were split into surface cores ( 0 25 cm ) and deep cores ( 25+cm ) due to different sampling frequencies (see methods) B oth 13 C and 14 C differed significantly with depth in surface ( Fig. 3 2 ; two way ANOVA, p<0.021) and deep cores ( Fig. 3 2; one way ANOVA, p<0.0073). Corre cting R h 13 C from each depth section for in situ soil temperatures caused the top 15 cm of soil to be

PAGE 64

64 13 C depleted in July than in other months due to warmer soil temperatures ( Fig. 3 2b). On average, the R h 13 enrichments relative to measured values of YS and OS, respectively. Heterotrophic respiration sources, YS and OS (calculated from averaged core 14 13 C ( Table 3 3). Young soil was more depleted 13 C ( 25.7 to 14 13 C around 14 C around 13 C differed by month in 2009 but not 2008 ( Table 3 3; one way ANOVAs, p>0.0001 and p=0.41, 14 C di d not differ by month in either 2008 or 2009 ( Table 3 3; one 14 13 C of the other R h source, OS, did not differ by month in either year ( Table 3 3; one way ANOVAs, p>0.69). Partitioning Ecosystem Respiration The greatest contributions to R eco came from AG R a whose contributions ranged from 16 to 48% in 2008 and 26 to 43% in 2009 ( Fig. 3 3). Belowground R a contributions ranged from 17 to 34% in 2008 and 15 to 32% in 2009 ( Fig. 3 3). Results reported in this section are the means of the posterior probability density distributions for each eco Combining mean contributions across years and AL categories, AG R a had the greatest contributions at the height of the growing season, July and A ugust, while BG R a did not significantly differ throughout the growing season ( Fig. 3 4; two way ANOVA, month x type interaction, p<0.0001). We did not use July 2008 partitioning results in this analysis because of the uncertainty associated with estimatin g the July 2008 AG and BG source isotopes. The greatest proportional contributions of AG R a at the height of the growing season corresponded to the lowest proportional contributions of BG R a while AG and BG R a were similar during months of

PAGE 65

65 seasonal transit ion, May and September. Combining all AL categories, months, and years, mean total R a (AG plus BG) contributions to R eco ranged from 40 70% and increased with increasing thaw depth (Fig. 3 5A ; regression, n=18, R 2 =0.25, p=0.039). One rmed to tease apart the effects of AL category and month, since both contribute to increasing thaw depths. Autotrophic respiration did not differ significantly among AL categories (one way ANOVA, p=0.37) but did vary significantly among months (one way ANO VA, p=0.0047) with the greatest contributions occurring in July and August, implying the relationship between R a and thaw depth was driven by seasonal differences. For heterotrophic contributions to R eco YS had greater contributions than OS ( Fig. 3 3). Y oung soil contributions ranged from 20 to 53% in 2008 and 20 to 41% in 2009. Young soil generally contributed more to R eco where the AL was shallow. Old soil contributions ranged from 6 to 17% in 2008 and 8 to 18% in 2009. The greatest contributions from AL categories, months, and years (except May 2009 outliers), OS contributions increased with increasing thaw depth (Fig. 3 5B ; regression, n=16, R 2 =0.52, p=0.025). Old soil respiration did not differ significantly among months (one way ANOVA, p=0.43) but did differ significantly among AL categories (one way ANOVA, p=0.0022) with shallow AL having the smallest OS contributions. These results imply the relationship between OS and thaw depth w as driven by AL category and not seasonality. The relative contributions of autotrophic and heterotrophic respiration to R eco can be compared using a ratio. Combining all AL categories, months, and years, R a :R h increased with increasing thaw depth (Fig 5c; regression, n=18, R 2 =0.23, p=0.047).

PAGE 66

66 We performed sensitivity analyses to test how model estimates responded to uncertainty in source isotopic values (Table A 14 C by one standard ons up to five percentage points and all other source contributions less than three percentage points. Using uncorrected 13 C soil values (see methods) resulted in three to 10 percentage point changes in 13 in zero to six percentage point changes. Relationships between source contributions and month or thaw depth only changed slightly as a result of these changes and maintained their significance. Our results are therefore qualitatively robust to source isot opes uncertainty. Respiration fluxes Table A 3 ). Growing season C flux from AG and BG R a and OS R h were greatest in the deep AL category ( Fig. 3 6). Th is large r growing season C flux from R a and OS reflect both the larger contributions of R a and OS to R eco and the greater R eco flux rates where the AL was deepest. Discussion Both autotrophic and old soil heterotrophic respiration increased with permafrost thaw. We were able to measure the crucial loss of old soil C because we used a dual isotope approach and explicitly measured the isotopic value of all sources. This method allowed us to more accurately partition R eco into more sources than past studies. We even detected seasonal differences in aboveground and belowground R a contributions to R eco Taken together the thaw induced increases in R a and old soil R h can have different C balance outcomes depending on the response of primary produc tion. If the magnitude of C input (net primary productivity) response to permafrost thaw is not

PAGE 67

67 greater than the C output (respiration) response, thaw will have caused the system to become a C source. Variability of a utotrophic s ource is otopes Our first ob jective was to sample source isotopes temporally and spatially, which 13 14 C in future partitioning studies. Previous studies have made untested assumptions, such as assuming R a has the same 14 C as the atmos phere (e.g., Subke et al. 2011) or assuming R h has the same isotopic values as solid organic C (e.g., Dutta et al. 2006). Isotopic variation in R a was driven by 13 C from BG plant structures w structures as observed in previous studies (e.g. Badeck et al. 2005; Klumpp et al. 2005) Belowground respiration is likely depleted relative to AG respiration because root metabolism is f ueled by depleted substrates, like lipids, while leaf metabolism is fueled by enriched substrates, like sugars (Bowling et al. 2008) In contrast, AG and BG 14 C did not differ indicating they were respiring substrates of similar ages. In 14 C relative to the atmosphere indicating plants were using a mixture of stored carbohydrates and recently fixed photosynthates for respiration (Czimczik et al. 2006; Schuur & Trumbore, 2006) Since R a 14 C enrichment has been previously seen in other perennial plants (Schuur & Trumbore, 2006; Czimczik et al ., 2006), only annual plants should be assumed to respire CO 2 14 C as the atmosphere. Autotrophic respira 13 C varied temporally across the growing season in both years, but R a 14 13 C was likely due to plant/water relations wherein C 3 13 C enriched under dry conditions due to

PAGE 68

68 limitations on stomatal con ductance (McDowell et al. 2004; Bowling et al. 2008) In July 2009, when the study site experienced below 13 C of AG R a was 2 August 2008, after a wet Jul y, both AG and BG R a 13 C, R a 14 C did not vary temporally, indicating atmospheric 14 C was well mixed and that plants used a similar amount of stored C during the growing season, reinforci ng results in forests (Cisneros Dozal et al. 2006; Czimczik et al. 2006) Due to temporal variability, when partitioning R a 13 C should be measured at the same time as R eco 13 C and potential differences in isotopic values before and after precipitation events should be considered. In contrast, sampling R a 14 C once a growing season is sufficient. Variability of heterotrophic source isotopes Carbon isotopes of R h 14 C became depleted with depth, so the deeper the soil, the older the organic C respired. 14 C became measured in black spruce forests (Schuur & Trumbore, 2006). While R h 14 C became depleted with depth, R h 13 as seen in numerous other studies (Ehleringer et al. 2000; Hgberg et al. 2005; Bostrom et al. 2007) This enrichment may be due, in part, to the Sues s effec t in which fossil fuel burning has 13 C. This effect causes surface soils (Ehleringer et al. 2000; Hogberg et al ., 2005). Th explained by increasing proportions of microbial derived enriched C with depth relative

PAGE 69

69 to depleted plant derived C (Ehleringer et al. 2000; Hogberg et al. 2005; Bostrom et al. 13 14 C of R h were more variable in deep soils than in surface soils. Respired CO 2 from the top 5 cm was all less than 20 years old while respired CO 2 from soil 80 cm deep ranged from 1500 to 7000 years old. This variability could be a result of cryoturbation (i.e., soil mixing caused by freeze/thaw cycles; Hicks Pries et al. 2012) or variable C accumulation rates. Respired CO 2 from most depth sections was 13 C enriched than bulk organic C (Hicks Pries et al. 2012) indicating isotopic values of bul k soil organic C should not be used to partition R eco. 13 14 C did vary spatially but did not show consistent temporal variation, supporting previous results from a pine forest (Carbone et al. 2011) Partitioning e cosystem r espiration 14 C became more depleted throughout the growing 14 C was also more enriched than the atmosphere and more enriched than R a across all categories 14 C were equal to the atmosphere. The depletion in R eco 14 C could therefore be due to: 1) increasing contributions of R a which 14 C only slightly more enriched than the atmosphere; 2) increasing contributions of 14 C (e.g., Schuur et al. 2009); 3) decreasing 14 C (e.g., Borken et al. 2006) ; or 4) a combination of the above. Areas with deeper active layers at our study site al so have been shown to have greater R eco C losses (Vogel et al ., 2009). Only partitioning can decipher the mechanism behind the depleting R eco 14 C and increasing R eco losses.

PAGE 70

70 Partitioning revealed that the seasonal and thaw induced decreases in R eco 14 C were driven by increases in both R a and heterotrophic respiration of old soil C. Autotrophic contributions to R eco increased throughout the growing season peaking in August, likely a result of increased plant biomass, increased net primary production, and warmer soils. We were able to detect seasonal changes in R a by sampling several times throughout the growing season. Our results differed from previous studies, which found R a contributions peaked at the start of the growing season (Chiti et al. 2011) or did not change during the growing season (Nowinski et al. 2010) Outside the growing season, R a contributions have predictably decreased when plants are dormant (Ruehr & Buchmann, 2010; Subke et al. 2011) Across all AL depths and times, R a contributio ns ranged from 40 70% of R eco Similar wide ranges of R a contributions, 41 54% in a peatland (Hardie et al. 2009) and 40 80% in tundra (Nowinski et al. 2010), were found in previous R eco partitioning studies. In terms of total growing season respiration flux, autotrophs respired 58% and 28% more C in areas with deep permafrost thaw than where thaw was shallow in 2008 and 2009, respectively. Similarly, permafrost thaw induced by a 1C warming doubled autotrophic contributions to R eco in a Swedish permafros t peatland (Dorrepaal et al. 2009) In contrast, autotrophic contributions to R eco decreased with snow fence induced permafrost thaw in Toolik, AK; though the snow fence shortened the growing season, likely eliciting the negative plant response (Nowinski et al. 2010) Plants have generally responded to warming by increasing primary production (Rustad et al. 2001) At our site, plant biomass is 35% greater where permafrost thaw is extensive relative to where thaw is minimal (Schuur et al. 2007). Nearby i n Healy, AK, warming soil by 2.3C

PAGE 71

71 caused a 20% increase in aboveground productivity (Natali et al. 2012) Ecosystem respiration is positively correlated with aboveground net primary productivity at our study site (Vogel et al. 2009). The maintenance of higher productivity where active layers are deepest may therefore necessitate higher rates of R a Autotrophic respiration is made up of respiration from AG and BG plant structures, and respiration from AG plant structures drove the R a increases discussed a bove. Aboveground R a increased as the unfrozen layer deepened throughout the growing season and among AL categories, but BG R a did not vary temporally as in a previous study (Cisneros Dozal et al. 2006) Relative contributions of AG and BG R a shifted seas onally indicating changes in plant C allocation. At the shoulders of the growing season, respiration contributions from AG and BG plant structures were similar, while at the height of the growing season (July and August), AG R a contributions were roughly t wice that of BG. In absolute magnitude, however, BG R a remained steady indicating root and rhizome respiration does not increase even when plants reallocate C from and to stored reserves at the beginning and end of the growing season. The difference betwee n AG and BG R a contributions in July and August were therefore driven by an increase in AG R a, which temporally corresponds with the greatest rates of primary production at and near our study site (Vogel et al. 2009; Natali et al. 2011) The observed dep letion in R eco 14 C with deepening thaw (seasonally and across AL categories) was not only caused by increased R a but also by increased OS R h 14 C value around likely a mix of younger and much older C). With deepening thaw, more soil C is available to decomposition. In September, there is 21.1 kg more C m 2 available to

PAGE 72

72 above freezing respiration where thaw is deep than where thaw is shallow, a 90% increase in thawed soil C (soil data from Hicks Pries et al. 2012). However, thawed soil below 50 cm is still very cold with temperatures less than 1C and is often saturated. Thaw may therefore increase OS R h in other ways, by causing soils to be warmer farther up in the soil profile or via prim ing as plant productivity increases and roots grow deeper. Our estimates for the contributions of OS to R eco across the study site (6 to 18%) fall within the ranges of previous peatland studies (Hardie et al. 2009; Dorrepaal et al. 2009; Schuur et al. 2 009). The increase in OS respiration is particularly of note since this soil has been stored away from the atmosphere for hundreds of years, making its release a potentially huge positive feedback to climate change ( Schuur & Abbott 2011 ): as permafrost thaws, more old C is respired, which increases atmospheric CO 2 causing more warming, which in turn thaws additional permafrost. This potential for a positive feedback is illustrated by the 67 103% increase in growing season old C flux where the permafrost thaw is deep relative to where thaw is shallow, similar to the 78% increase measured previously at our study site (Schuur et al. 2009). The increase in OS flux with deepening thaw was 25 50% less than the R a growing season increase (21 30 g C m 2 versus 4 0 81 g C m 2 ), demonstrating OS losses can be obscured by changes in plant respiration, except when revealed by isotopes. While OS R h generally increased as thaw deepened throughout the growing season, there were two exceptions. In May 2009, the highest O S contributions reported in this study occurred in places with deep active layers. These high contributions early in the growing season are indicative of a burst of CO 2 from deeper in the soil profile during thaw (Lee et al. 2010) Due to the downward mov ement of soil freeze in the

PAGE 73

73 autumn, microbial respiration continues to occur at above freezing temperatures while the distance between the frozen surface soil and the permafrost closes. Decomposition can also occur within unfrozen soil micro sites during w inter. When the surface soil thaws in the spring, some of the old C decomposed during the autumn and winter is released. Challenges of the i sotopic p artitioning a pproach The isotope partitioning method assumes isolat ed incubations of soil, roots, and above ground plant structures minimally affect the 13 C or 14 C of respiration. For 14 C, the assumption is supported by several studies (Dioumaeva et al. 2002; Czimczik & Trumbore, 2007) 13 13 C can ch measurements are made directly after excision (Midwood et al. 2006) as we did in this study 13 C from incubations have also been shown to change over times from a few hours (Millard et al., 2008) to months (Blagodatskaya et al. 2011) to years (Follett et al. 13 C. One 13 C sh ift is the respiration and loss of labile root exudates et al. (2008) was a shift away from the depleted root respiration value. In the present study, we waited five and 10 d 13 C from surface and deep soil cores, respectively. The five day wait was based on a field study wherein soil respiration decreased by 50% in the five days after trees were girdled (Hogberg et al. 2001). By waiting, we ex cluded root exudates from our YS source and included them in the BG source. Root exudates turnover quickly and are tied to the autotrophic response, so including them in the BG source is appropriate for this study, which was principally

PAGE 74

74 designed to measure how OS respiration responded to permafrost thaw. During 13 C also shifts due to changes in microbial substrate preference from labile to more recalcitrant C as the labile C pool is exhausted. Our incubation length was specifi 13 C while the labile C pool was respired. Incubation studies demonstrate the labile C pool takes at least five to 20 days to be exhausted in tundra soils, the longer time applying to deeper soils (Lee et al., 2012; Lavoie et al., 2011). Lastly, soil oxygen gradients in incubations may differ from in situ conditions, which may also affect isotopic values. 13 C based off the change measured during soil incubations in Foll ett et al. caused the AG proportion to increase and the YS and OS proportions to decrease, and 2.7% of R eco All AL categories responded similarly, s 13 C would not change the relationships between thaw depth and source contributions. Our results are 13 C shift during short incubations and how the shift affects part itioning results warrant further study as the isotope partitioning method becomes more widely used. Implications for n et e cosystem c arbon b alance Flux measurements indicate areas with deeper permafrost thaw at our study site had the greatest gross primary production and were likely a net C sink in 2008 and 2009 (Trucco et al. in press ). Therefore the increased growing season losses from R a and old soil R h with increasing permafrost thaw are currently more than compensated for by increased net primary production. However, the loss of old soil C is concerning in the long term because sustained losses of old soil C are likely given the climate change

PAGE 75

75 trajectory. Because the size of the soil C pool (55 70 kg m 2 ; Hicks Pries et al. 2012) is much larger than the tundra plant C pool (0.35 kg m 2; Shaver & Chapin, 1991) future losses of old soil C will likely outpace increases in plant production, barrin g a major shift in plant community from tundra to boreal forest (e.g., Callaghan et al. 2004) Even then, a mature boreal spruce forest stores about 6 kg C m 2 in plant biomass (Gower et al. 2001; Goulden et al. 2011) only enough to compensate for a 10% loss of soil C. The questions that remain are what proportion of permafrost soil C is vulnerable to permafrost thaw and to what extent primary production increases can compensate soil C losses.

PAGE 76

76 Table 3 1. Mean (SE) thaw depth and 13 C of eco system respiration within the active layer categories throughout the growing season in 2008 and 2009. Significance was tested separately for each year with two Categories that do not share a capital letter, months that do not share a lowercase letter, and individual means that do not share a number are significantly different. 13 C was significantly different among months but not among active layer categories. Thaw (cm) 13 2008 June a July b August c June a July b August ab Shallow A 24.4 1.2 1 36.1 1.1 2,3 49.5 1.1 4 24.4 0.8 22.8 0.3 23.3 0.3 Intermediate B 29.2 1.2 1,2 46.0 3.9 3,4 67.8 2.8 5 23.8 0.4 22.6 0.4 23.3 0.5 Deep C 43.5 2.4 2,3,4 68.0 6.0 5 86.0 7.6 6 23.9 0.5 23.1 0.4 23.1 0.2 2009 May a July b September c May ab July a September b Shallow A 9.82 .36 1 32.9 .71 2 50.8 1.4 3 23.1 0.6 22.6 0.2 23.4 0.2 Intermediate AB 10.7 .60 1 41.9 2.1 2,3 65.1 2 .0 4 22.6 0.6 22.3 0.2 24.8 1.6 Deep B 17.9 5.7 1 56.9 4.8 3,4 88.3 7.2 5 23.8 0.2 22.8 1.0 24.0 0.5

PAGE 77

77 Table 3 2 13 14 C of AG and BG plant respiration during the 2008 and 2009 growing season. Significance was tested separately for each year with two Asterisks represent a significant difference between aboveground and belowground R a 13 C only Months that do not share a lowercase letter and individual means that do not share a number are significantly differen t in 13 C. There were 14 C significant differences. N S stands for not sampled. 13 14 13 14 13 14 2008 June a July August b Aboveground* 20.5 0.3 NS NS 44.2 1.7 22.7 0.3 NS Belowground 23.1 0.6 NS NS 48.5 1.8 25.8 0.3 NS 2009 May a July a September b Aboveground* 22.2 0.3 1 48.1 2.0 20.3 0.4 2 48.8 2.8 23.2 0.5 1,3 NS Belowground 24.4 0.5 3,4 49.1 0.4 24.5 0.2 3,4 49.7 2.4 25.8 0.4 4 NS

PAGE 78

78 Table 3 3 13 14 C of young and old soil respiration. Significance was tested separately for each year and each source with one 13 C in 2009 ( letters not shared). 13 14 13 14 13 14 2008 June July August Young Soil 24.7 0.2 76.4 4.2 25.0 0.2 76.4 4.2 25.0 0.2 76.4 4.2 Old Soil 22.7 0.3 30.5 38 22.8 0.2 26.8 35 23.0 0.2 26.3 35 2009 May July September Young Soil 24.6 0.2 ab 76.4 4.0 25.7 0.2 a 76.5 4.1 24.4 0.2 b 76.5 4.2 Old Soil 22.6 0.3 30.7 37 22.9 0.3 26.3 34 22.8 0.3 30.3 37

PAGE 79

79 Figure 3 1. Mean R eco 14 C for all active layer (AL) categories in all months sampled (error bars=SE). In both years, early growing season R eco was more enriched 14 C than later growing season R eco Letters not shared indicate significant differences among AL categories which 14 C of the at mosphere in each year.

PAGE 80

80 Figure 3 2. Mean isotopic values (error bars=SE) of R h by depth. 14 13 C. Statistical analyses were split into surface soil cores ( capital letters, sampled in May, July, and September 2009) and deep soil cores (lowerca se letters, sampled in May 2009 only). Depths that do not share a letter are significantly 13 C, laboratory soil respiration values had to be corrected for in situ soil temperatures for each month we partitioned (see methods) The 2 009 corrected data are shown here

PAGE 81

81 Figure 3 3. Percent contributions of sources to R eco by month for each active layer category. Autotrophic sources are on the left (AG is for aboveground and BG is for belowground plant structures) and heterotrophic sources are on the right (YS is for young soil and OS is for old soil). These contributions are the mean f estimates from SIAR, and th e error bars are the 25 and 75 percentiles of the estimates. Both 2008 (June, July, and August) and 2009 (May, July, and September) data are shown. Please note that July 2008 data (gray symbols) 13 C (see me thods).

PAGE 82

82 Figure 3 4. SIAR estimates for contributions of aboveground R auto (AG, black symbols) and belowground R auto (BG, open symbols) averaged over active layer category and year for each month sampled. The asterisks show significant 5). AG contributions were greater than BG contributions, and AG contributions were greatest in July and August. BG contributions did not vary by month.

PAGE 83

83 Figure 3 5. Autotrophic respiration contributions to R eco ( A ), old soil respiration contributions to R eco ( B ), and the autotrophic to heterotrophic respiration ratio ( C ) with thaw depth for each active layer (AL) cat egory and month. Note that for A and B the contributions are the means of the probability density distributions that were a result of runnin g the SIAR model for each AL category. Ratios greater than one indicate autotrophic respiration dominates. The r 2 B the regression was performed without the May 2009 outliers (see text for details).

PAGE 84

84 Figure 3 6. Estimates of growing season fluxes from each respiration source (aboveground plant structures (AG), belowground plant structures (BG), young soil (YS), and old soil (OS)) for active layer (AL) categories in 2008 and 2009. The error bars are the spatial er ror of the respiration fluxes. The mean estimates of source proportional contributions from SIAR for each month and AL category were multiplied by the fluxes for each month and AL category in each year. The results were then summed over May, June, July, Au gust, and September.

PAGE 85

85 CHAPTER 4 EXPERIMENTAL WARMING CHANGES THE RESPONSE OF RESPIRATION TO SOIL TEMPERATURE Abstract Ecosystem respiration (R eco ) generally increases as a result of warmer temperatures. In tundra ecosystems undergoing permafrost thaw, understanding w hether autotroph ic or heterotroph ic source s drive R eco increases is necessary to arge positive feedbacks to climate change. A positive feedback is likely if the respiration of old soil, previously stored for hundreds to thousands of years, increases without a concurrent increase in primary productivity. Here, we appl ied a dual natural abundance carbon isotope 13 14 C) to partition R eco into four sources: aboveground plant structures, belowground plant structures, young soil (0 15 cm), and old soil (15 75 cm). We measured the isotopes of ecosystem and source respiration a t CiPEHR (Carbon in Permafrost Experimental Heating Research), a warming experiment in subarctic tundra that passively warms the air with open top chambers and warms the deep soil using snow fences, causing permafrost thaw. After three years of experimenta l warming at CiPEHR, R eco flux in the warming treatments had increased up to 57% relative to the control. We found that proportional contributions of old soil respiration to R eco increased as soil temperatures increased, and the slope of this relationship was dependent upon treatment. Old soil proportional contributions increased slower in the warming treatments than in the control because warming treatments also increased the proportional contributions of autotrophic respiration, a result of greater prima ry productivity. In contrast, total growing season C losses from old soil were up to 150% greater in the warming treatments than in the control, because absolute R eco fluxes were

PAGE 86

86 smaller in the control. Overall, growing season losses of old soil C were mor e than compensated for by increases in primary production in the warming treatments. These results complement an isotope partitioning study in a natural gradient of permafrost thaw, which found deepening thaw caused increases in old soil and autotrophic re spiration. However, increases in plant productivity will likely not continue to outpace future increases in old soil respiration, especially if warming and thaw increase old soil respiration beyond the growing season. I ntroduction The largest carbon (C) fl ux from the terrestrial biosphere to the atmosphere is respiration. Globally, respiration increases with greater mean annual temperatures (Raich & Schlesinger 1992). Thus, respiration increases due to air and soil warming have been documented in many ecos ystems (Wu et al 2011), including tundra ( Hobbie & Chapin, 1998; Oberbauer et al 2007 ), boreal and northern forests ( Peterjohn et al .,1993; Goulden et al ., 1998; Rustad & Fernandez 1998 ), and grasslands (Zhou et al ., 2006 ) There are exceptions though, where warmer temperatures do not increase respiration, likely a result of conditions that are too dry (Liu et al ., 2009) or too wet (Oberbauer et al ., 2007), or once labile soil carbon pools have been metabolized (Mellilo et al ., 2002). Despite the plethora of studies on the response of respiration to temperature increases, the mechanisms behind the increase remain unclear. Ecosystem respiration (R eco ) is a result of both autotrophic and heterotrophic respiration, both of whi ch can increase with warming. Numerous laboratory soil incubations demonstrate increasing temperatures increase rates of heterotrophic microbial respiration ( Conant et al ., 2011). On a physiological level, plant respiration increases with warmer temperatur es but eventually levels off due to acclimation

PAGE 87

87 (Lambers et al ., 2008). However, plant respiration can also increase as an indirect effect of warming, if warming increases primary productivity and biomass (Hobbie & Chapin 1998). The relative responses of autotrophic (R a ) and heterotrophic respiration (R h ) to climatic change s can ultimately affect ecosystem C balance because R a is generally balanced by production, but R h may not be Therefore i ncreases in R eco can indicate different C cycle outcomes depending on which respiration source drives the change. For example, if respiration of decadal or older soil C is driving the increase, the eco system is losing C that had previously been stored away from atmosphere, resulting in a positive feedback to cli mate change. However, i f R a is driving the increase, the eco system is either turning over recent photosynthate faster or is fixing more C, leading to different feedback outcomes than increased soil respiration. Turning over recent photosynthate faster is a constrained positive feedback to climate change because this C loss is limited by the size of the plant biomass C pool. F ixing more C in plant biomass is a negative feedback to climate change. A mechanistic understanding of R eco is especially important i n ecosystems underlain by permafrost (perennially frozen ground). Permafrost ecosystems have the potential to become a huge positive feedback to climate change if permafrost thaw and warming increase heterotrophic respiration (Schuur et al ., in review). So ils in the permafrost zone currently store over 1600 Pg C (Tarnocai et al ., 2009) because the frozen soil has protected organic C from decomposition for hundreds to thousands of years. In a subarctic tundra, organic C deep in the soil ( 85 cm) dates back t o the start of the Holocene, 10,000 years ago (Hicks Pries et al ., 2012). This permafrost C is vulnerable to being lost via increased heterotrophic respiration because much of it is

PAGE 88

88 located in high latitudes, where up to 7 C temperature increases are predi cted over the next century ( IPCC, 2007 ). Permafrost in some areas is currently thawing downward at a rate of 0.1 to 0.9 m year 1 continually exposing more organic C to increased microbial activity (Osterkamp & Romanovsky, 1999; Osterkamp, 2007) There have been numerous warming experiments in arctic and subarctic tundra including ITEX (International Tundra Experiments; Oberbauer et al ., 2007), experiments at Toolik Lake, AK (Hobbie and Chapin 1998; Oberbauer et al ., 1998; Jones et al ., 1998), Bar row, AK (Huemmrich et al ., 2010), Abisko, Sweden (Dorrepaal et al ., 2009), and most recently, in Healy, AK (Natali et al ., 2011). The majority of these studies measured the response of the total R eco flux to warming and found R eco increased. However, measu ring fluxes alone cannot indicate whether R h of the critical old soil C pool is increasing (Dioumaeva et al ., 200 2 Hicks Pries et al ., in press ). To our knowledge, only one tundra warming experiment has investigated whether R eco increases were a result of autotrophs or heterotrophs. Dorrepaal et al (2009) found warming increased respiration by 52% relative to the control and 69% of that increase was due to respiration of older, deep (25 50 cm) soil C. Numerous methods for partitioning R eco exist, includi ng trenching roots, girdling trees, measuring CO 2 fluxes of excised ecosystem components, or using natural abundance or enriched carbon isotopes as tracers (Kuzyakov 2006). Isotopic methods are advantageous because they do not alter environmental conditio ns (e.g., Subke et al ., 2011) or depend on accurate, separate respiration fluxes (Yi et al ., 2007). M any studies including Dorrepaal et al (2009), have used either 13 C or 14 C separately to estimate source contributions to soil or ecosystem respiration ( Ehleringer et al ., 2000;

PAGE 89

89 Gaudinski et al ., 2000; Trumbore, 2000; Ngao et al., 2005; Schuur et al., 2009 ) However, combining C isotopes in the same partitioning analysis can be a more powerful method because 13 C and 14 C separate sources based on differe nt principles 13 C separates respiration sources via biological fractionation when enzymes discriminate against the heavy isotope (Bowling et al 2002), 14 C separates respiration sources by age on millennial timescales due to radioactive decay, and deca dal 14 C caused by atmospheric nuclear weapons testing (Levin & Hesshaimer, 2000, Trumbore, 2000) With dual isotope partitioning, R eco source contributions can be determined with greater accuracy compared t o a single isotope approach (Phillips & Gregg, 2003) Terrestrial studies using dual natural abundance C isotopes are rare (Hardie et al ., 2009; Hicks Pries et al ., in press ). Here, we apply this dual isotope approach at CiPEHR (Carbon in Permafrost Experimental Heating Research), a warming experiment in subarctic tundra (Natali et al ., 2011). CiPEHR is unique among warming experiments because it warms not only air temperature and surface soil, but also deep soil, causing permafrost thaw. After three y ears of experimental warming at CiPEHR, R eco increased up to 57% relative to ambient (Natali et al ., in review ). To investigate whether autotrophic or heterotrophic sources are driving this increase, we used 13 14 C to partition R eco at CiPEHR into f our sources : aboveground plant structures (AG) belowground plant structur es (BG), young soil (YS, 0 15 cm), and old soil (OS, 15 75 cm).

PAGE 90

90 Methods Site Description CiPEHR is located near Eight Mile Lake ( EML, 63 52' 59"N, 149 13' 32"W ) in the foothills o f the Alaska mountain range in Healy, Alaska. The vegetation consists of moist acidic tussock tundra dominated by Eriophorum vaginatum The vegetation also includes the graminoid Carex bigelowii dwarf shrubs Vacinnium uliginosum V. vitis idaea Betula nana Rhodendron subarticum Rubus chamaemorus and Empetrum nigrum and various mosses and lichens. The soils are Gelisols (permafrost is within 70 cm of the surface) and consist of 0.3 to 0.5 m of organic soil atop a mixture of mineral loess deposits an d glacial till In the EML area, permafrost temperatures have been monitored via a borehole and have been increasing over the past several decades (Osterkamp & Romanovsky, 1999). The CiPEHR experiment consisted of winter warming (WW) and summer warming (SW ) treatments set up in a factorial design: control, SW, WW, and annual warming (SW+WW). The summer warming was achieved passively with open top chambers (OTCs, 60 x 60 cm), and the winter warming was achieved with snowfences. The snowfences created snowdri fts over one meter deep that insulated soils throughout the winter. At the end of each winter, the excess snow was shoveled off the WW treatment so as not to add aditional water to or delay snowmelt. There were six replicate snowfences distributed evenly a mong three blocks. The WW treatment and WW control plots were the north and south sides of each fence, respectively. Each WW treatment and control plot contained both SW treatment and SW control plots. The SW treatment raised growing season air temperature s by about 1C and the WW treatment raised winter soil temperatures by 2 7C (depending on the soil depth) and growing season

PAGE 91

91 soil temperatures up to 1.5C (Natali et al ., in review ). The WW treatment also increased the depth of thawed soil during the grow ing season by 10% (Natali et al ., in review environmental conditions, see Natali et al ., 2011, 2012, and in review Soil Environment soil environment were used to investigate relationships between R eco isotopes and the soil environment. In all plots, the soil temperature at three depths (5, 20, 40 cm) and the soil volumetric water content (VWC) integrated over the top 20 cm were record ed every half hour during the growing season (Natali et al., in prep). Soil temperature was measured using constantan copper thermocouples and VWC was measured using site calibrated Campbell CS616 water con tent reflectometer probes. Thaw depth (the depth fr om the soil surface to the frozen soil) was sampled at 3 points in each plot using a metal probe pushed into the ground until it hit frozen ground. Throughout the growing season, w ater table depth (WTD) was measured three times per week from water wells in stalled in each WW treatment and control plot (12 total; Natali et al ., in review ). Ecosystem Respiration To measure the 13 C 14 C of R eco we installed 24 PVC collars (25.4 cm diameter x 10 cm deep) 8 cm into the soil one per each of the four WW/SW treatment combinations at each of the six fences. We used the same methods as Hicks Pries et al. ( in review) to sample R eco 13 C 14 C. In brief, 10L dark chambers were fit onto the PVC collars encompassing the abovegrou nd plant biomass. Ecosystem respiration 13 C was sampled using the Keeling plot method wherein CO 2 was collected into septa capped glass vials (Exetainers, Labco Limited, Lapeter, UK) every two to three minutes

PAGE 92

92 for a total of seven samples per collar, whil e an Infrared Gas Analyzer (LI 820, LI COR, Lincoln, Nebraska USA ) simultaneously measured pCO 2 The sample exetainers and a set of field standards ( Trading Corporation Safford, AZ, USA ) with similar pCO 2 were sent back to the University of Florida to be run on a GasBench II coupled to a Finnigan Delta Plus XL stable isotope ratio mass spectrometer (precision 0.2 n=215) 13 C changes due to travel and holding time were corrected using the field standards 14 C was collected by pumping CO 2 through a molecular sieve (Alltech 13x, Alltech Associates, Deerfield, IL, USA ) for 15 minutes. Prior to the collection, the chamber headspace was scrubbed for 45 minutes with soda lime while maintaining ambient pCO 2 to remove atmospheric contamination The molecular sieves were baked at 625 C to desorb CO 2 (Bauer et al. 1992) which was purified using liquid N 2 on a vacuum line and reduced to graphite by Fe reduction in H 2 (Vogel et al. 1987) T he graphite was sent to the UC Irvine W.M. Keck Carbon Cycle Accelerator Mass Spectrometry (AMS) Laboratory for radiocarbon analysis (precision 2.3 n=102 ) 14 C data were reported at the same 13 C value to correct for mass dependent fractionation effects. 14 C data were c orrected for atmospheric contamination using 13 C data in a 2 pool (atmospheric and R eco ) mixing model (Schuur & Trumbore 2006) Ecosystem respiration isotopes were sampled at the beginning of July and in mid September 2009, and in mid Augu st 2010 and 2011. Sampling only occurred under calm conditions and in the morning to control for potential diurnal variation. Source Respiration Short term incubations were used to measure the 13 C 14 C of autotrophic and heterotrophic source respiration using methods outlined in Hicks Pries et al ( in

PAGE 93

93 press ). We collected AG and BG plant material to measure R a isotopes by cutting 5 x 10 t at each fence. Samples from the same treatment were combined by block, for a total of three AG and three BG replicates per treatment. We clipped all live AG plant material from each brownie and placed it in foil covered mason jars (0.24L). Belowground ro ots and rhizomes (>1 mm in diameter) were collected from the thawed soil, rinsed twice, and shaken dry before being put into their own mason jars. We incubated plants as soon as possible after collection (within 5 minutes for AG and 30 for BG), to avoid ch anges to 13 C that can occur 40 minutes to an hour after excision ( Midwood et al ., 2006) The jar headspace was scrubbed by pumping the air through soda lime for 5 minutes at >1 L min 1 before starting the incubations. 13 C incubations lasted for 5 1 0 minu te s after which headspace air was pumped into He flushed exetainer s. 14 C incubations lasted four hours after which headspace CO 2 was collected into molecular sieves. Autotrophic respiration 1 4 C and 13 C was measured at the beginning of July and in mid Se ptember 2009, and in mid August 2010 and 2011. Only WW treatment and control plots were sampled in July 2009 because the SW treatment had only been going on for six weeks, too soon to expect an SW treatment effect. Autotrophic respiration 1 4 C was sampled only from control plots in 2010 and 2011 because winter and control R a significantly in 2009 (p=0.33). To measure the 13 C 14 C of heterotrophic respiration we collected surface (0 25 cm) and deep (25 75 cm) soil cores. Surface soil cores were collected in WW treatment and control plots in July 2009 and from all treatments in August 2010. Because the isotopes of surface soil resp iration did not change significantly from 2009

PAGE 94

94 to 2010, we did not sample surface soils in August 2011. To core surface soils, we used a serrated knife to cut 5 x 5 cm blocks of soil 25 cm deep. We sectioned the blocks of soil into 0 5 cm, 5 15 cm, and 15 25 cm depths. As with the plant sampling, we combined samples from the same treatment within a block for a total of three replicates per treatment. We removed all roots >1 mm in diameter and let the surface s oils sit at room temperature in mason jars for 5 days before sampling 13 C and 14 C This wait ensured we would be sampling the labile soil C pool and not the recent root derived C (Hicks Pries et al., in review, Schuur and Trumbore 2006). Wait time was informed by a field study wherein soil respiration rates decreased by 50% 5 days after tree girdling (Hogberg et al., 2001) and from tundra soil incubations wherein the labile C pool was respired during the first 5 20 days (Lee et al ., 2012, Lavoie et al ., 2011). In May 2009, we sampled deep soil cores (25 75 cm) while the ground was still frozen using a Tanaka TIA 340 permafrost drill Cores were taken from the WW and winter control side of each fence for a total of 12 samples. Deep soil cores were kept frozen until spring 2011, when we cut them into 10 cm sections, removed roots >1 mm diameter, and let them sit in mason jars at room temperature for 10 days to allow microbial populations to stabilize after thaw. For both shallow and deep soils, we measured R h flux during three short (~3 hour) incubations pri or to sampling CO 2 for isotopic analysis. Soils were kept at field moisture and under aerobic conditions. For 13 C and 14 C sample collection, we scrubbed CO 2 from the jar headspace, incubated the soils for 12 to 72 hours (the time it took for 1.5 mg C to build up in the headspace), and pumped the headspace CO 2 into molecular sieves The 14 C sample preparation and analysis was carried out as

PAGE 95

95 described for R eco However, after purification, a small (0.1 0.2 ml) subsample of CO 2 was removed and put in a He f illed exetainer for 13 C analysis. Partitioning Model For partitioning R eco the R h isotopes from the soil core depth sections were combined into two heterotrophic sources, YS (0 15 cm) and OS (15 75 cm) based on the age of the respired CO 2 Note that the depths of these two sources do not correspond to how the surface (0 25 cm) and deep soils (25 75 cm) were sampled. The 14 C from the past 50 years, and the OS include d soil sections that respired older, depleted 14 C from the upswing of the bomb peak and earlier ( Table B 1). To calculate 13 C 14 C values of YS and OS, we weighted the isotopic signatures of each incubated depth by its CO 2 flux per g C, corrected for each average monthly field temperature (Schuur & Trumbore 2006; Hicks Pries et al ., in press ; Table B cores were averaged to obtain mean 13 C 14 C for YS and OS. Before averaging, 13 C values were corrected for the incubation temperature shift (Dorrepaal et al ., 2009; Andrews et al ., 2000; Biasi et al ., 2005 ) because our soils were incubated at temperatures warmer than field conditions. We used a Q 10 of 2.5 and a 13 C temperature correction of (Hicks Pries et al ., in review) in the calculations described above. The same 14 C values from the 2009 deep soil cores (incubated in 2011) were used to calculate the R h sources for partitioning in all years because changes in 14 C due to radioactive decay during that three year time span were smaller than the 2 (Schuur et al ., 2009)

PAGE 96

96 We used SIAR (Stable Isotope Analysis in R; Parnell et al ., 2010) to partition Reco into two autotrophic sources (respiration of AG and BG plant material) and two heterotrophic sources (YS and OS). Partitioning was performed separately for each co llar and sampling period. The SIAR method uses Markov chain Monte Carlo to find possible solutions to this set of three equations: ( 4.1 ) where the unknowns are f eco 13 C 14 C of each source and R eco have known distributions The input data for this model include the mean and standard deviation of all source isotopic values and the individual R eco isotope values An average AG and BG R a 13 C was used to partition all sampling periods and treatment s, because there were no significant differences within AG or BG R a 13 C among sampling periods or treatments. The July 2009 R a 14 C values were used for September 2009 partitioning. The results of the model are probability density distributions of each so f We used the default uninformative prior (Parnell et al ., 2010) because previous partitioning results are limited. Data Analysis To investigate treatment and sampling date differences in the soil environment, isotopic values, and R eco source cont ributions (mean f ), we performed analyses of date, the winter treatment, and the summer treatment nested within the winter treatment. For the soil environment variables, R eco i sotopes, and source contributions,

PAGE 97

97 fence and plot nested within fence were the random effects. For source isotopes, block was a random effect since samples from each pair of fences were combined (core was nested in block for soil isotopes), and type (AG/BG or YS/OS) was an additional main effect. For isotopes from individual soil sections, depth was an additional main effect. Source contributions were logit transformed before analyses. All residuals were checked for normality and homogeneity of variances to ensure the assumptions of ANOVA were met. The ANOVAs demonstrated that differences among treatments in isotopic values and source contributions to R eco were not straightforward possibly due to spatial and temporal heterogeneity with regards to the soil environment. To explore how the soil environment affected 14 C and source contributions to R eco we performed multiple regressions in R (R Development Team, 2012). First, 14 C was square root transformed and source contributions were logit transformed to en sure normality. We included soil temperatures at 20 cm (we chose only one soil depth to reduce collinearity among predictors), soil volumetric water content (VWC), depth to the water table, thaw depth, and treatment as explanatory variables for 14 C and so urce contributions ( Table B 2 ). We used the water table depth measured one to two days before sampling and the thaw depth measured immediately after sampling. The other variables were measured continuously, so they were averaged over the day R eco 14 C was collected for the 14 C model, and they were averaged over the entire sampling period for the source contribution models. We used the full model to optimize random effects and variance structures using AIC values following Zuur et al. ( 2009). Random effects (block, fence, and plot) did not improve the models, so they were not included. A power variance

PAGE 98

98 structure taking soil temperature at 20 cm and treatment into account improved the 14 C model, but not the source contribution models. Once random effects wer e optimized, we then performed a series of pair wise model comparisons using the F test; sequentially testing the models by dropping the least significant explanatory variable each time (highest p value) until only significant e xplanatory variables remaine d ( Zuur et al ., 2009). Ultimately, the 14 C model was fitted with the gls command in the nlme package (Pinheiro et al ., 2010) using restricted maximum likelihood, and the source contribution models were fitted with the base lm command because they had no s ignificant random effects. We ran separate regressions to investigate how R eco 14 C and OS proportional contributions were related to R eco fluxes measured at CiPEHR. The R eco flux had been sampled from each plot during the growing season via auto chambers (Natali et al ., 2011; Natali et al ., in review ; Table B 3 ). We used the average R eco flux over each sampling period. We used the regression of OS against R eco flux, with trea tment as additional factor, t o estimate total growing season losses from OS re spiration This regression relationship was used to predict the proportion of R eco coming from OS in each plot hourly throughout the growing season (May 1 st through September 30 t h ). The OS proportion was then multiplied by the corresponding R eco flux and summed over the growing season for each plot. Results Soil Environment In addition to warming soils during the snow covered period, winter warming affected the soil environment du ring sampling; however, summer warming did not (Table 4 1). Warming significantly increased thaw depth and soil temperature at 40 cm

PAGE 99

99 (p<0.026). Soil volumetric water content was also greatest in the winter warming treatments (p=0.037). Water table and soil temperatures at 5 and 20 cm did not differ significantly among treatments (p>0.17). All soil environment variables varied significantly by sampling date (p<0.0001). August 2010 had the wettest soils of all the sampling dates in terms of VWC and had the wa rmest soil temperatures. The water table was about 14 cm further below the soil surface in August 2011 than during other sampling periods when it was about 23 cm below the surface. Thaw was deepest in August 2010 and shallowest in June 2009. Ecosystem Resp iration 14 C became significantly more depleted as R eco flux increased (Fig. 4 1 Table 4 2 ) and, in a separate regression, became significantly more depleted as soil temperature at 20 cm increased (Fig. 4 2; Table 4 2 ). Treatment was a significant variable in the multiple regression with soil temperature and treatment 14 C was more pronounced in the control than in the warming treatments (Fig. 4 2; Table 4 2 ). Without soil temperature as a covariate, R eco 14 C did not differ significantly among treatments in 14 C (p=0.37; Table 4 3 ) whose treatment averages ranged from 22.5 to 37.7 across all sampling dates. S ampling date averages across treatments had a wide r range, from 2.5 to 50.9 so that August 2010 R eco 14 C than R eco from all other sampling dates (p<0.0001 ; Table 4 3 ) In contrast, R eco 13 C differed marginally among treatments (p=0.07 ; Table 4 2 ) because 13 C in the winter warming treatm ent ( 24 ) was more depleted than the winter control ( 23.5 ; Table 4 3 ) but R eco 13 C did not diffe r among sampling dates (p=0.31 ; Table 4 3 )

PAGE 100

100 Sources of Ecosystem Respiration 14 C did not differ among treatments (p > 0.33, WW and SW were tested separately because SW was not sampled in all years ; Table 4 4 ) but R a 14 C became depleted from 47.9 to 40.2 throughout the experiment (p=0.014 ; Table 4 4 14 C of the atmosphere 13 C of aboveground and belowground R a did not differ among treatments or sampling dates (p>0.47; Table 4 4 ). Belowground R a 14 C averaged 49.2 and was significantly more enriched than aboveground R a by about 7 (p=0.007; Table 4 4 ). Belowground R a 13 C averaged 25.3 and was always about 2.7 more depleted than aboveground R a (p<0.0001; Table 4 4 ). There were no significant treatment differences in R h isotopes from either surface (0 25 cm) or deep soil cores (25 75 cm; p>0.10), but R h differed significantly with depth (Table B 13 C became significantly more enriched with depth in shallow 14 C became significantly more depleted with d epth in both shallow and deep core depth sections were analyzed separately because they were sampled at different frequencies. Due to the lack of treatment differences amon g soil cores, data from all surface and deep cores were combined to calculate the depth 13 14 C of young and old soil R h sources. Young soil 13 C averaged 23.6 was significantly 13 C than old soil, which averaged 21 .6 (p<0.0001; Table 4 5 14 C averaged 88.2 was significantly more enriched in 14 C than old soil, which averaged 81.6 (p<0.0001; Table 4 5 ). 13 14 C differed significantly among sampling dates (p<0.0001; Table 4 5 ) due to di fferent in situ

PAGE 101

101 soil temperatures used to calculate depth integrated isotopic values for each date (see Methods). Respiration Source Contributions Using the isotopic signatures of R eco and its sources, we estimated proportional contributions of autotrophi c and heterotrophic sources to R eco W e only found marginally significant treatment effects on source contributions to R eco (Fig. 4 3 ). Surprisingly, despite deeper thaw, old soil contributions were smaller in the winter warmed treatments (annual and winter) compared to control (p=0.078). Aboveground R a contributions were larger in the annual warming treatment (p=0.076). The greatest treatment effect was in August 2010 when old soi l source contributions were 150 % greater in the control than in the annu al warming plots (p=0.0073). In contrast to treatment effects, there were strong temporal variation s in source contributions from old soil, young soil, and aboveground plants ( p < 0.0 079 for all, Fig.4 1 ). The contribution of old soil to R eco increased si gnificantly with the soil temperature at 20 cm and interacted significantly with treatment; old soil contributions in the control increased faster with temperature than those in the annual and summer warming treatments (Fig. 4 4 ; Table 4 2 ). Seemingly cont radicting the increase with soil temperature, old soil contributions actually decreased with increasing thaw depth (Table 4 2). Autotrophic (aboveground plus belowground) contributions did not change significantly with any of the soil environment variables (Table 4 2 ). Autotrophic contributions only differed among treatments with the winter treatment having significantly greater autotrophic contributions than the control and the annual and summer warming treatments falling in between (Table 4 2 ). The R a to R h ratio also only

PAGE 102

102 changed significantly among the treatments with winter and annual warming treatments having larger ratios than the control (Fig 4 5 ; Table 4 2 ). Lastly, old soil contributions to R eco significantly increased with the R eco flux (p<0.0000 1) at a rate of 2.4% more OS contributions to R eco per 0.01 g C m 2 hour 1 increase in flux, indicating old soil is a greater proportion of R eco when flux rates are faster. Treatment significantly affected the intercept of this relationship (Fig 4 6 ; Table 4 5). This relationship was used to estimate total growing season loss of old soil C. In 2010 and 2011, old soil C losses were larger in treatments that had winter warming than the winter control (p<0.0071; Fig, 4 7 ). Discussion Overall, warmer soil temp eratures increased old soil contributions to R eco a trend that was modified by treatment. Proportional contributions of old soil increased faster with soil temperature in the control than in the warming treatments because autotrophic contributions were a greater proportion of R eco in the warming treatments. When autotrophic and heterotrophic contributions were combined into a single number, the R a to R h ratio, the winter warming treatment emerged as the most significant effect. Winter warming increased the autotrophic proportion of R eco more than the heterotrophic proportion; therefore three years of warming did not cause a positive growing season feedback to climate change. On the contrary, net ecosystem productivity at our site indicates warming is inc reasing growing season C uptake (Natali et al ., in review ), of which the increase in autotrophic respiration in the warming treatments is a direct result. Partitioning Ecosystem Respiration 14 C generally became more depleted as soil temperatures increased with the fastest decrease occurring in the control treatment. Ecosystem

PAGE 103

103 14 C has previously been shown to become more depleted (Hicks Pries et al ., in press ) or more enriched with permafrost thaw (Nowinski et al ., 2010). Though thaw depth was not a significant predictor of R eco 14 C in this study, it is positively correlated with soil temperature. In the previous studies, both 14 C shifts were due to increases in the contribution of heterotrophic sources to R eco Similarly the depletion of R eco 14 C in this study was caused by an increase in heterotrophic respiration of the old soil C pool. As temperatures r i se, microbial activity increases, causing more old soil C 14 C value ( 82 ) to be respired. The response of R eco 14 C and old soil contributions to soil temperature was affected by treatment. Old soil contributed more proportionally to R eco in the control than in warming treatments T he old soil contribution is a proportion of R eco so its decrease can be driven by a concurrent increase in another respiration flux, such as autotrophic respiration. Autotrophic proportional contributions were significantly driven by treatment, with winter warming increasing autotrophic contributions. Thus, smaller old soil proportional contributions in the warming treatments were caused by increased contributions of plant respiration. The slight decrease in old soil contributions with deeper thaw was likely also caused by increased autotrophic respiration since the wint er warming treatment where autotrophic contributions are greater, have deeper thaw than the control. Plant respiration increased as a result of warming due to increased aboveground net primary production, especially of graminoids, which was greatest in t he winter warming treatments (Natali et al ., 2012; Natali et al ., in review ). The increased plant productivity and biomass necessitated an increase in growth and maintenance

PAGE 104

104 respiration within the warming treatments (Hobbie & Chapin 1998; Amthor 2000). T he mechanism for the plant respiration increase was not a direct result of soil temperature or a release from water limitation (the winter warming treatments were often wetter; Natali et al ., in review ) as these were not significant predictors in the regre ssion. Instead, it is likely that plants in the winter warming treatments responded to increased nitrogen availability caused by faster decomposition in the winter warming treatments. Canopy nitrogen was greatest in winter warming plots (Natali et al ., 201 2), and nitrogen mineralization often increases with experimental warming (e.g., Shaver et al ., 1998; Rustad et al ., 2001). The overall range of old soil contributions to R eco (3 to 73%) was larger in this study than in previous studies from northern peat lands (Dorrepaal et al ., 2009 ; Hardie et al ., 2009 ) and tundra (Schuur et al ., 2009; Hicks Pries et al ., in press ). In these studies, old soil contributions to R eco topped out at 45% (Schuur et al ., 2009). A tenth of our old soil estimates were greater than 45% and all of these occurred in August 2010. Our regression models were unable to fully explain this sampling date variability. Residuals of both old soil and R eco 14 C models varied significantly by date August 2010 differed from August 2011 and Sep tember 2009 but was the same as July 2009 Unlike the treatment effect on R eco 14 C and OS contributions, the larger OS proportional contributions in August 2010 were not due to a decrease in autotrophic respiration, because gross primary production was t he same during August 2010 and 2011 sampling (0.17 g C m 2 hour 1 ; data from Natali et al ., in review, averaged over sampling dates). Th e variability was also evident in the R eco flux, which was greatest during the August 2010 sampling: 0.10 g C m 2 hour 1 versus 0.08 and 0.05 g C m 2 hour 1 for

PAGE 105

105 August 2011 and both 2009 sampling dates, respectively. E nvironmental conditions that favor greater R eco also favor more old soil C loss as implied by the significant relationship between R eco flux and R eco 14 C The sampling date variability is likely due to a combination of interactions among soil environmental variables and seasonality (Hicks Pries et al ., in press) that we lacked the statistical power to detect. August 2010 had the optimal conditions for hete rotrophic decomposition: it had the deepest thaw, the warmest soils, and the moistest soils (Chapter 4) of all sampling dates. Old Soil Carbon Losses In terms of C flux (not proportional contributions), growing season C losses from old soil respiration we re greater in the warming treatments than in the control. Although the control generally had greater old soil proportions than the warming plots, the control had smaller R eco fluxes overall (Natali et al ., in review ), leading to less old soil C loss. In 20 10, 34 to 78% more old soil respiration came out of the warming treatments than the control, and in 2011, there was 59 to 150% more old soil respiration. The significant positive relationship between old soil respiration contributions and the total R eco f lux show that increases in respiration rates measured with warming were partially being driven by increases in old soil respiration. This increase in growing season old soil C loss due to warming is similar to the 67 103% increase in old soil C loss previo usly estimated in tundra with deep permafrost thaw relative to tundra with shallow thaw (Hicks Pries et al ., in press ). Variability of Ecosystem Respiration Sources The isotopes of autotrophic and heterotrophic respiration sources, while used here to part ition R eco also contain information about ecosystem s when viewed on their own A previous partitioning study found autotrophic 13 C varied monthly

PAGE 106

106 (Hicks Pries et al ., in press) due to changes in water availability and photosynthetic demand. 13 C did not differ among sampling dates, likely because our partitioning and source sampling occurred at an annual timescale. Similar to past studies, respiration of aboveground plant structures were around 2 more enriched than belowground plant structures, likely a result of different C substrates driving respiration in roots and rhizomes versus leaves and stems (Bowling et al ., 2008). Many tundra plants store C in rhizomes (Olsrud & Christensen, 2011), so belowgr ound respiration is likely supplied in part by older C from a storage pool. On average, 14 C in any given year while b elowground structures respired C that was enriched rel ative to the atmos phere by 5 one to two years old on average, but was likely a mix of contemporary and older C (S chuur & Trumbore, 2006) Older than atmospheric C has been found respired from roots in several studi es (Czimczik et al ., 2006; Schuur & Trumbore, 2006; Hicks Pries et al ., in press) and by stems and leaves in a previous study (Hicks Pries et al ., in press). Lastly, 14 C respired from both autotrophic sources became more depleted from 2009 through 201 14 C dead fossil fuel C to the atmosphere dilutes the enriched bomb peak (Levin & Hesshaimer, 2000), 14 C. 13 14 C changed down th e soil profile of both 13 14 C becoming 13 C enrichment has been measured in numerous studies and potential mechanisms include the Suess effect and increas es in microbially

PAGE 107

107 derived C with depth (Bowling et al 14 C is due to soil C becoming older with depth, especially in vertically accreting organic soils. Some cores 14 C from the 0 5 cm to the 5 15 cm depths due to respiration of enriched bomb 14 2 from the deepest core sections (65 75 cm) ranged from 327 to 5750 years old, a smaller age range than found in nearby deep soils (Hick s Pries et al ., in press). As a result of these depth differences, young and old soil respiration sources 13 14 differences made them strong end members in the partitioning model and lead to more accurate source contribution estimates. Implications for Net Ecosystem Carbon Balance Overall, warming in this moist acidic tussock tundra is causing an increase in R eco that is driven by both plant and old soil respiration. The increase is being driven more by autotrophs than heterotrophs because the ratio of autotrophic to heterotrophic respiration is greater in the warming treatments than in the control. A similar incr ease in this ratio has also been measured with deepening natural permafrost thaw (Hicks Pries et al ., in press), but in a northern peatland warming increased autotrophic and heterotrophic respiration equally (Dorrepaal et al ., 2009). The increase in autotr ophic respiration is because plants are fixing more C as a result of warming such that the warming treatments are currently a growing season sink of 102 g C m 2 (Natali et al ., in review). Thus, the growing season loss of old soil C is being compensated fo r by increased plant growth and biomass C storage. However, due to the relative sizes of the tundra plant C pool (0.45 0.63 kg m 2 ; Natali et al ., 2012) and soil C pool (60 kg m 2 ; Hicks Pries et al ., 2012), it is likely that old soil C losses will eventua lly surpass gains in

PAGE 108

108 plant biomass, even if the tussock tundra becomes a boreal forest (6 kg C m 2 in a mature spruce forest; Gower et al ., 2001; Goulden et al ., 2011 ) This prediction matches an ecosystem model of shrub tundra, which predicted net primary production increases would outpace heterotrophic respiration increases due to climate change over the next 100 years before eventually decreasing (Grant et al ., 2011). Furthermore, this study only examined growing season dynamics in R eco Warming likely i ncreases autumnal and wintertime soil respiration losses, which may offset the autotrophic growing season gains (Natali et al ., in review).

PAGE 109

109 Table 4 1 Mean (SE) environmental variables by sampling date and treatment. Note that while most variables were m easured continuously and were thus averaged over the sampling period, water table and thaw depth were not. Their values are the measurements that were taken closest to the sampling period (right before for water table and immediately after for thaw depth). Asterisks indicate there are differences between treatments with (winter and annual) and without (control and summer) winter warming. Control Summer Winter Annual Thaw Depth (cm) Jul 2009 a 26.9 1 27.8 1 37.0 1* 33.2 1* Sep 2009 b 51.4 2 52.0 2 59.0 3 53.3 1 Aug 2010 c 54.5 2 55.6 2 65.4 2 59.1 1 Aug 2011 b 45.5 2 46.5 2 62.1 2 56.7 2 Water Table Depth (cm) Jul 2009 a 23.4 3 21.9 1 21.3 2 25.2 3 Sep 2009 a 22.5 2 22.3 2 18.0 3 21.1 4 Aug 2010 a 23.3 3 22.8 3 17.2 3 21.8 4 Aug 2011 b 37.2 1 34.0 2 33.7 3 35.3 3 Temperature at 5 cm (C) Jul 2009 a 6.6 0.8 6.2 0.4 7.4 0.3 8.1 0.7 Sep 2009 b 6.1 0.4 6.6 0.2 6.4 0.3 6.4 0.4 Aug 2010 c 9.6 0.5 10.1 0.6 10.0 0.2 9.2 0.3 Aug 2011 a 7.0 0.3 6.8 0.3 7.3 0.3 6.9 0.1 Temperature at 20 cm (C) Jul 2009 a 1.5 0.3 1.2 0.4 1.8 0.3 1.5 0.4 Sep 2009 b 2.8 0.2 3.2 0.3 2.9 0.2 2.8 0.2 Aug 2010 c 4.6 0.4 5.2 0.3 5.5 0.5 5.4 0.4 Aug 2011 b 2.8 0.2 2.8 0.2 3.5 0.2 3.5 0.4 Temperature at 40 cm (C) Jul 2009 a 0.05 0.2 0.15 0.03 0.10 0.02* 0.13 0.02* Sep 2009 b 1.1 0.3 1.1 0.2 1.4 0.1 0.94 0.2 Aug 2010 c 2.5 0.5 2.1 0.2 3.7 0.7 3.5 0.7 Aug 2011 b 1.2 0.4 0.6 0.1 2.1 0.3 2.1 0.5 Volumetric Water Content (%) Jul 2009 a 38.3 3.0 45.2 3.6 46.7 3.1* 45.5 2.8* Sep 2009 b 46.1 0.8 47.9 1.9 50.9 5.2 49.0 2.2 Aug 2010 c 53.8 1.5 54.6 2.1 58.6 1.5 62.2 2.0 Aug 2011 b 50.8 1.5 48.2 2.2 53.9 1.6 56.3 1.9

PAGE 110

110 Table 4 2 Multiple regression results. These are the predictors that came out as significant after stepwise regressions involving the soil temperature at 20 cm (T20), soil volumetric water content, thaw depth, water table depth, and treatment ( n=69, see Methods). The transformation used for th e response variables are in parentheses. Response Coefficients Estimate SE t value p value R eco 14 C Intercept 0.026 1.42 0.018 0.9854 (reflected Summer 5.38 1.60 3.36 0.0013** square Winter 4.06 1.69 2.40 0.0194** root) Annual 3.65 1.50 2.43 0.0178** T20 2.00 0.49 4.11 0.0001*** T20:Summer 1.90 0.56 3.37 0.0013** T20:Winter 1.52 0.57 2.69 0.0092** T20:Annual 1.50 0.50 3.01 0.0038** R eco 14 C Intercept 2.74 0.63 4.29 <0.00001*** (reflected square root) R eco flux 34.61 7.4 4.69 <0.00001*** Old Soil Intercept 2.65 0.71 3.73 <0.000 43 *** (logit) Summer 1.48 0.84 1.77 0. 081* Winter 1.29 0.91 1.41 0. 16 Annual 2.15 0.86 2.50 0.0 15 ** Thaw.cm 0.04 0.014 2.93 0.00478** T 20 1.07 0.24 4.41 <0.00001*** T 20 :Summer 0. 62 0.25 2.48 0.0 16 T 20 :Winter 0.54 0.25 2.11 0. 039* T 20 :Annual 0.82 0.25 3.21 0.0 019 ** Old Soil Intercept 2.59 0.37 7.0 <0.00001*** (logit) Summer 0.56 0.31 1.8 0.0697* Winter 0.76 0.31 2.5 0.0151** Annual 0.68 0.30 2.2 0.0255** R eco flux 18.52 4.13 4.4 <0.00001*** Autotrophic Intercept 0.26 0.15 1.8 0.085* (logit) Summer 0.32 0.21 1.5 0.136 Winter 0.56 0.21 2.7 0.010** Annual 0.29 0.21 1.4 0.171 R a :R h Intercept 0.26 0.13 1.9 0.05921* (log) Summer 0.32 0.19 1.6 0.10127 Winter 0.57 0.19 2.9 0.00464** Annual 0.49 0.19 2.6 0.01144**

PAGE 111

111 Table 4 3 13 C 14 C of ecosystem respiration by treatment and sampling date T here were no significant differences among treatments, but 14 C did differ among sampling dates ( n=3 for 2009 and n=6 for 2010 and 2011, Control Summer Winter Annual 13 Jul 2009 22.9 1.2 23.7 1.3 24.8 0.3 22.1 0.7 Sep 2009 23.0 0.8 23.6 0.9 24.3 0.8 23.7 0.4 Aug 2010 23.9 0.7 24.1 0.6 24.7 0.5 24.2 0.3 Aug 2011 22.9 0.6 23.7 0.6 24.3 0.8 23.9 0.4 14 Jul 2009 a 41.2 1.3 27.6 6.1 36.9 1.8 46.2 1.4 Sep 2009 a 51.7 7.2 54.3 9.5 58.3 0.9 39.1 4.8 Aug 2010 b 34.1 12 4.87 13 10.8 9.8 24.7 2.3 Aug 2011 a 45.5 1.2 43.4 5.8 47.2 4.3 45.7 1.4

PAGE 112

112 Table 4 4 13 14 C of aboveground ( AG ) and belowground ( BG ) plant respiration sampled in 2009, 2010, and 2011. Asterisks indicate significant differences among AG and BG R auto and letters not shared indicate significant differences among sampling dates ( ). AG R auto was significantly more enriched in 13 C than BG R auto but was significantly more depleted in 14 C ( n=30 and n=9 ) 14 C differed significantly among sampling dates ( n=6 ) NS means there were no samples taken of the treatment during the sampling date Control Summer Winter Annual Control Jul 2009 13 Jul 2009 a 14 AG* 22.3 0.7 NS 21.1 0.1 NS AG* 44.1 1.0 BG 25.1 0.3 NS 25.1 0.6 NS BG 52.6 2.5 Aug 2010 Aug 2010 ab AG* 20.9 0.5 23.0 1 21.7 0.3 22.1 0.9 AG* 43.8 1.6 BG 25.9 0.1 25.9 0.5 24.6 0.9 25.3 0.1 BG 51.6 0.1 Aug 2011 Aug 2011 b AG* 22.3 0.6 21.4 0.5 22.9 0.7 22.8 1 AG* 37.5 1.2 BG 25.3 0.2 25.8 0.6 25.4 0.8 25.0 0.8 BG 42.9 9.3

PAGE 113

113 Table 4 5. 13 14 C of young and old soil respiration. There were significant differences between young and old soil and among sampling dates 13 14 values of these sources are based on the same soil samples, their isotopes vary by year because the values were correc ted based on soil temperature profiles, which did vary by sampling date (Table B 1 ). Young Soil* Old Soil Young Soil* Old Soil 13 14 Jul 2009 a 24.0 0.18 21.5 0.2 88.9 2.0 74.6 7.1 Sep 2009 b 23.0 0.17 21.3 0.3 87.2 2.0 90.1 8.3 Aug 2010 c 23.9 0.18 21.8 0.2 88.4 1.9 79.5 7.5 Aug 2011 d 23.6 0.18 21.6 0.2 88.1 1.9 82.0 7.7

PAGE 114

114 Figure 4 1. R eco 14 C decreased as ecosystem respiration fluxes increased. The points are the actual data and t he line is the model prediction (see Table 4 2 for model coefficients).

PAGE 115

115 Figure 4 2 R eco 14 C decreased with soil temperature at 20 cm Treatment changed the slope of the R eco 14 C and temperature relationship. The points are the actual data and the lines are the model predictions (see Table 4 2 for model coefficients) Open triangles and the dotted line represent the control treatment, open squares and the dashed line represent the summer treatment, closed diamonds and the dot dashed line represent the winter treatment, closed circles and the solid line represent the annual treatment.

PAGE 116

116 Figure 4 3 Mean proportional contributions of aboveground (AG) and belowground (BG) autotrophic respiration (top grap hs) and young soil (YS) and old soil ( OS ) heterotrophic respiration by treatment for each sampling period. The error bars represent the standard error of the mean source contributions averaged over all plots within a treatment (n=3 for Jul 2009 and Sep 200 9, and n=6 for Aug 2010 and Aug 2011). Letters not shared indicate significant differences among treatments, which only differed significantly ( =0.05) in August 2010.

PAGE 117

117 Figure 4 4 Old soil contributions to R eco generally increased with the soil temperatu re at 20 cm, a relationship that changed significantly with treatment. The points are the actual data and the lines are the predicted data for each treatment (see Table 4 2 for model coefficients) Open triangles and the dotted line represent the control t reatment, open squares and the dashed line represent the summer treatment, closed diamonds and the dot dashed line represent the winter treatment, closed circles and the solid line represent the annual treatment.

PAGE 118

118 Figure 4 5 The mean ratio of heterotrophic to autotrophic respiration across all sampling dates. The bars represent the standard error. In a mixed linear regression, the only significant predictor was treatment. The ratio is significantly greater in the plots with the winter warming t reatment than in black line is a ratio of one, below which heterotrophic respiration dominates.

PAGE 119

119 Figure 4 6 Old soil contributions to R eco increased with increasing respiration flux. The points are the actual data, the solid line is the prediction of a linear regression (in which the old soil proportion was logit transformed ; Table 4 2) ), and the dotted lines are the 95% confidence intervals of the prediction. There was a significant treatment effect on the intercept of this relationship that is not shown on this graph (see Table 4 2 ).

PAGE 120

120 Figure 4 7 Estimated growing season old C losses from the control and warming treatments in 2010 and 201 1. The relationship shown in Figure 4 6 was used to estimate the losses (see M ethods). Old soil C losses are significantly greater in the plots with the winter warming treatment than in plots with the

PAGE 121

121 CHAPTER 5 MOISTURE CONTROLS DE COMPOSITION IN THAWING AND WARMING TUNDRA Abstract As permafrost thaws, soils become warmer and thermokarsts form, causing some areas to become wetter and others to become drier; these environmental changes affect de composition rates. We used a common substrate (cellulose filter paper) to measure how permafrost thaw and warming affect decomposition rates in the top 20 cm of soil in a natural permafrost thaw gradient and a warming experiment in Healy, AK. Permafrost th aw also shifts the composition of plant communities, affecting decomposition by changing substrate quality. We used a common garden of 12 plant substrates (making up 90% of primary productivity at our site) to test how substrate affects decomposition rates We then combined those data with species level estimates of aboveground net primary productivity to calculate plant community weighted decomposition constants at both the thaw gradient and warming experiment. We found that soil moisture, as measured by g rowing season precipitation and water table depth, was an important control of decomposition. At the thaw gradient, a 100 mm increase in growing season precipitation increased mass loss of a common substrate by 100%. At the warming experiment, a 15 cm shal lower water table also increased mass loss by 100%. In contrast to soil moisture, community weighted decomposition did not change much with permafrost thaw and warming despite shifts in plant community composition. At the thaw gradient, community weighted decomposition was only 21% faster at the site with extensive permafrost thaw than at the site with minimal thaw, a difference that would likely be negated if moss production were included. Therefore climate change and ition are driven more by precipitation and

PAGE 122

122 soil environment than by changes to plant communities. Increases in soil moisture are then another mechanism by which permafrost thaw can become a positive feedback to climate change. Introduction Large amounts of soil carbon (C) have accumulated in eco systems underlain by permafrost over the past 100,000 years. Permafrost soils store 16 7 2 Pg C more than twice the amount of C that currently resides in our atmosphere (Schuur et al 2008; Tarnocai et al 2009; ). These ecosystems have accumulated l arge amounts of C because cold and often frozen soil temperatures reduce decomposition rates Accompanying these cold soil temperatures are extremes in soil moisture that can also limit decomposition. Permafrost soils c an be waterlogged as the permafrost is a barrier to drainage, limiting aerobic microbial respiration (Gebauer et al ., 1996) However, many permafrost soils are found in high latitude tundra ecosystems, where mean annual precipitation is lo w, similar to des erts, grasslands, or savannas (Chapin et al ., 2002) Therefore, while deeper soil in the permafrost zone is often waterlogged, the surface soil can dry out quickly, also limiting decomposition. Permafrost thaw changes the environmental conditions that affe ct decomposition by increasing soil temperatures and shifting local hydrology. Many permafrost soils are currently warming and even thawing as their temperatures rise above 0C (Osterkamp 2007; Osterkamp & Romanovsky 1999 ). This thaw is a result of clima te change, which has caused air temperatures at high latitudes to increase 2C over the past 60 years and will cause temperatures to increase 7 8C over the next century ( IPCC, 2007 ). Thaw can create thermokarsts, uneven ground due to soil subsidence follo wing the loss of ground ice (Osterkamp et al ., 2009) What type of thermokarst forms depends on the

PAGE 123

123 slope, aspect, and parent material of the ecosystem ( Jorgenson & Osterkamp, 2005 ) On flat ground, permafrost thaw can result in thermokarst ponds or wetlan ds where the water table is near or above the soil surface On slopes, permafrost thaw can create water tracks that funnel water and can dry out adjacent areas (Osterkamp et al. 2009) Aside from the creation of thermokarsts, permafrost thaw can also change the height of the water table in more subtle ways. Small amounts of ground subsidence can decrease the depth from the soil surface to the water table while deepening of the active l ayer without ground subsidence can increase the depth from the soil surface to the water table. Permafrost thaw and warming can also affect decomposition rates by altering biotic communities. Changes to the species composition of plant communities affect the quality of the substrate which supplies decomposition. Permafrost thaw has caused a shift in subarctic plant communities from graminoid and moss dominated to shrub dominated (Schuur et al 2007). In the arctic, climate warming is similarly causing a s hift to shrub dominated communities ( Sturm et al ., 2001 ). Shrub leaf litter is generally more easily decomposed than graminoid litter; however, the increase in woody litter, which graminoids do not have may decrease decomposition rates overall ( Hobbie et al ., 2000; Hobbie, 1996 ). Warming may also alter substrate quality by increasing nitrogen mineralization (Rustad et al ., 2001), which may cause plants to leave more nitrogen in their litter which can increase decomposition rates ( Hobbie, 2005 ). Lastly, pe rmafrost thaw can change decomposition rates by altering microbial communities like those involved with nitrogen and methane cycling (Mackelprang et al 2011).

PAGE 124

124 The relative importance of environment and substrate quality in affecting decomposition may b e a matter of spatial scale. Over large spatial scales, c limate is often considered the most important driver of decomposition, more so than substrate quality or soil organisms (Levalle et al 1993 ; Aerts et al ., 2007 ). In one meta analysis of cold biome decomposition, Cornellisen et al (2007) found that temperature and then species composition were the most important drivers of plant litter decomposition. In a similar meta analysis, Aerts (2006) found that warming increased litter decomposition only if t here was adequate soil moisture. Smaller scale studies in cold ecosystems found plant growth form, a determinant of substrate quality, was a more important driver of decomposition rates than climate (Doreepaal et al ., 2005; Baptist et al ., 2010). In a glob al meta analysis of 44 litter decomposition studies climate was the best predictor of decomposition rates, but within a region, substrate quality was the best predictor (Aerts, 1997). While many studies have investigated litter decomposition in cold ecosystems (e.g., Hobbie 1996 ; Hobbie & Gough 2004; Dorrepaal et al .; 2005), to our knowledge, none have directly investigated how permafrost thaw affects decomposition rates. Understanding how permafrost thaw affects decomposition will allow us to better predict the fate of permafrost soil C. If permafrost thaw increases decomposition rates, these ecosystems may shift from being a C sink, which they have been for thousands of years (Hicks P ries et a l ., 2012) to a C source (e. g., Schuur et al 2009; Vogel et al 2010). Previous research at these field sites addressed whether permafrost thaw increased the respiration of old soil C at depth in the soil profile (Schuur et al 2009 ; Hicks Prie s et al

PAGE 125

125 in press; Chapter 3). This study instead focused on how decomposition near the soil surface is affected by physical and climatic changes caused by permafrost thaw In this study, we used a common substrate to compare how environment affects decom position in surface soil across a natural gradient of permafrost thaw (Schuur et al 2009) and in CiPEHR (Carbon in Permafrost, Ex perimental Heating Research; Natali et al ., 2011), an experiment that warms the deep soil and thaws permafr ost. Additionally, seven years of data in the thaw gradient and three years of data in CiPEHR allowed us to investigate causes of interannual variability in decomposition rates and how they interact with permafrost thaw. We hypothesized decomposition rates would be faster w here the soils were warmer due to permafrost thaw and experimental heating and during warmer years; however, in areas with deep permafrost thaw, rates may slow because shallower water tables may limit oxygen availability We also use a common garden to com pare the decomposition rates of 12 common plant litters. We then used the data from the common garden to investigate how thaw (Schuur et al ., 2007) and warming induced (Natali et al ., 2012) changes to plant communities, and therefore substrate quality, affect decomposition rates. We then compare the relative effects of environment changes and substrate changes on decomposition rates to see which drives decomposition rates as permafrost thaws. Methods Study Site The permafrost thaw gradient and CiPEHR are respectively located adjacent and two miles east of Eight Mile Lake (EML, 63 52' 59"N, 149 13' 32"W) in the foothills of the Alaska mountain range in Healy, Alaska Their vegetation consists of moist acidic tussock tundra dominated by Eriophorum vaginat um The vegetation also includes the

PAGE 126

126 graminoid Carex bigelowii dwarf shrubs Vacinnium uliginosum V. vitis idaea Betula nana Rhodendron subarticum Rubus chamaemorus Empetrum nigr um and various mosses and lichens. The soils are Gelisols and consist of 0.25 to 0.5 m of organic soil atop a mixture of mineral loess deposits and glacial till (Vogel et al ., 2009). At the thaw gradient, permafrost temperatures have been monitored via a borehole over the past several decades and are currently around 1C, mak ing the permafrost susceptible to thaw (Osterkamp & Romanovsky, 1999). The thaw gradient has had ongoing monitoring of soil temperature s to 40 cm active layer depth, water table depth, and CO2 fluxes since 2004 (Schuur et al ., 2009; Vogel et al ., 2009; Tr ucco et al ., 2012). The thaw gradient consists of three sites: minimal, moderate, and extensive thaw. The main difference among the sites is the duration of permafrost thaw each has undergone. At the extensive thaw site, permafrost thaw has been documented for the past two decades, but likely began earlier (Osterkamp et al ., 2009). As a result the sites have different degrees of plant community changes (Schuur et al 2007), active layer thickening, and thermokarst formation (Vogel et al ., 2009 ; Ostrekamp et al ., 2009 ). The CiPEHR experiment consist s of summer warming (SW) and winter warming (WW) treatmen ts set up in a factorial design resulting in four treatments : control, SW, WW, and annual warming (SW+WW). Open top chambers (OTC s ; 60 by 60 cm ) acted as small greenhouses passively warming the air by about 1C during the growing season in the SW treatments (Natali et al ., 2011) Snow fences slowed prevailing winds, creating deep (>1 m) snowdrifts that insulated soils in the WW treatments (Natali et al ., 2011) Excess snow was shoveled off the WW treatments each April so additional water was not added to the soil and snowmelt was not delayed. The WW treatment raised

PAGE 127

127 winter soil temperatures by 2 7C (depending on depth), growing season soil temper atures up to 1.5C and increased the depth of thawed soil during the growing season by 10% (Natali et al in review ). There were six replicate snow fences. The WW treatment and WW control plots were the north and south sides of each fence, respectively. Summer warming plots were nested in WW treatment and control plot s perimental design and treatment effects, see Natali et al 201 1, Natali et al ., 2012, and Natali et al in review Common Substrate Decompositio n To investigate the effects of permafrost thaw and warming on the decomposition environment, we used a common substrate (cellulose filter paper) incubated in mesh bags. Decomposing cellulose is a widely used method (e.g., Clymo, 1965; Fox & van Cleve, 198 3; Wagner & Jones, 2006) for comparing relative rates of decomposition among environments because it eliminates litter quality variation (McClellan et al ., 1990). While cellulose decomposition is not a good estimator of absolute rates of litter decompositi on, trends in cellulose decomposition across environments follow the trends of actual plant litter (Clymo, 1965; Vitousek et al 1994). At the thaw gradient, 10 bags per site were incubated annually from September 2004 through September 2011. Additionally, a total of 16 bags were incubated in and adjacent to eight thermokarsts south of extensive thaw from September 2008 through September 2009. At CiPEHR, six bags per treatment were incubated annually from September 2008 through September 2011 and during the growing season from late May/early June through mid September in 2010 and 2011. To make the common substrate decomposition bags, we cut the cellulose filter paper ( Fisher brand P8 09 802 1B ) into 7.5 by 7.5 cm squares for the thaw gradient and

PAGE 128

128 7.5 x 5 cm rectangles for CiPEHR. We wore gloves to avoid contaminating the paper. Four filter paper pieces went into each mesh bag, which was made out of two pieces of 2 mm mesh fiberglass window screening separated into four compartments and sealed. The mesh bags m easured 21 by 21 cm for the thaw gradient and 21 by 13.5 cm or 10.5 by 13.5 cm for CiPEHR annual and growing season bags, respectively. The CiPEHR decomposition bags were smaller because space was limited in summer warming plots. Filter papers were individ ually weighed (Mettler Toledo AX204 balance, 0.1 mg precision) before being put into one of four compartments within each bag. The compartments were arranged two by two with the longest side of each filter paper piece orientated vertically. The bags were d esigned to measure decomposition at two depths, from 0 10 cm and from 10 20 cm. Growing season bags only had the 0 10 cm depth, so they could be placed into the ground earlier in the growing season, once the top 10 cm of soil thawed. Lastly, each bag was l abeled with a unique number pressed into an aluminum tag. Each decomposition bag was placed into the soil vertically. We first cut a slit into the organic layer using a breadknife and then used a flat bladed shovel to insert the decomposition bag into the slit so that the top of the bag was even with the soil surface. Decomposition bags were inserted in mid September each year, left to incubate in the soil for a year, and then collected the following September. The slits were sometimes re used, but new slit s were made nearby (<0.5 m) once gaping caused the bags to no longer be in full contact with the soil. Upon removal the bags were rinsed to remove soil and frozen for transport back to the lab.

PAGE 129

129 At the lab, we dried the decomposition bags at 60C for 24 hou rs. We used paintbrushes and fine pointed tweezers to carefully remove soil and roots from the filter papers before measuring their final weight. Since the filter papers quickly absorbed moisture from the air, we measured their initial and final mass at ro om conditions, which we corrected for the moisture absorption. To calculate the correction factor, several times during processing, we weighed a subset of 10 filters before and after being dried at 60C for 24 hours and left to come to room temperature in a desiccator. For each filter paper, we calculated a percent mass loss by subtracting the final from the initial weight and dividing by 100. Mass loss of horizontally adjacent filters was averaged to get one number per depth per bag. Common Garden and Plan t Community Decomposition To investigate how plant community changes caused by permafrost thaw and warming affect substrate quality, we first incubated 12 plant litter types for three years in a common garden (i.e., the same site) near the thaw gradient. The litter consisted of leaves from six shrub species ( Eriophorum vaginatum Betula nana, Vaccinium uliginosum, Rubus chamaemorus Rhododendron subarticum Vaccinium vitis idea), two moss types (Dicranum and Sphagnum spp.), and a 1:1 mixture of live stem a nd root tissue from four shrub species ( Betula nana, Vaccinium uliginosum, Rhododendron subarticum, Vaccinium vitis idea ). L eaves and live woody tissue were collected in September 200 7 across 1 km of the thaw gradient, so substrates were a mix of samples f rom minimal, moderate, and extensive thaw. For deciduous leaf litter, we only collected s enesced leaves that were easily taken off branches, indicating the petiole had begun to abscise For evergreen leaf litters, R. subarticum and V.vitis idea we were un able to collect enough senesced leaves so we also collected live leaves. For mosses we cut

PAGE 130

130 off the green tissue at the top of the moss a nd used the brown tissue from the next 2 cm. Mosses were also gamma radiated to prevent them from re sprouting. For E vaginatum tissue remaining on the leaf blades. All litter samples were air dried after collection and stored at room temperature. We tested the effect of using live woody materia l and evergreen leaves by comparing the percent mass loss of senesced and live litter during a laboratory leaching experiment. This leaching is a good proxy for decomposition differences since mass loss during leaching is significantly correlated with deco mposition rate ( R 2 =0.3 1 p=0.0027 n=27; unpublished data) After 24 hours in DI water, there were no significant differenc es in mass loss from leaching between senesced and live litter (p=0.63). To make the litter bags, we carefully homogenized the litter, and then placed 2 grams of substrate into 8 x 12 cm 0.5 mm mesh bags The bags were then sewn together with polyester thread and labeled with an aluminum tag. The litter bags were deployed in September 2008 into the common garden which had five b locks. At each block, five lines consisting of the 12 litter types strung together in random order radiated out from a central point. Each litter bag was placed into the moss layer at a slight angle so that the bag was 0 to 1.5 cm below the soil surface. W e did this to represent decomposition conditions in the tundra where fresh litter falls in between individual mosses. One line was randomly collected from each block in May 2009, September 2009, May 2010, September 2010, and August 2011. The collected litt er bags were frozen for transport to the laboratory. At the laboratory, the litter bags were thawed, rinsed in DI water, and carefully picked through to separate litter from roots and hyphae

PAGE 131

131 that had grown into the bags. The litter was then dried for three days at 30C before being weighed (Mettler Toledo PG403 S, 0.001 g precision). To calculate decomposition constants for each litter type, we fit nonlinear regressions of a single pool exponential decay model in R using nls (R Core Development Team, 2012; Adair et al ., 2010). Initial quality of the Healy litter including %N, C:N, and fiber content, was measured on three subsamples of each substrate. For %C and %N analysis, the litter was ground to a fine powder and then run on a n ECS 4010 elemental analyze r ( Costech Analytical Technologies, Valencia, CA ). To determine soluble cell contents, hemicelluloses plus bound proteins, cellulose, and lignin w e performed sequential extractions of each initial litter on an ANKOM fiber analyzer ( ANKOM Technology, Maced on, N Y ; R yan et al. 1990). Stem and root initial quality was determined separately for each species, but was then averaged by species into a single wood value as the litter bags were a 1:1 mixture of stem and root substrate We calculate d community weighte d decomposition constants using decomposition constants from the 12 litter types and ANPP data to investigate how changing plant communities affect substrate quality. We used the latest plot level ANPP data available, which was from 2009 for the thaw gradi ent (n=6 per site) and from 2011 for CiPEHR (n=8 per treatment; Table C 1). Aboveground NPP was determined using a method that combines nondestructive point framing with allometric equations; this method has been previously described for both the thaw gradient and CiPEHR (Schuur et al ., 2007; Natali et al ., 2012). Th is method estimated secondary stem growth using growth rates measured in tussock tundra at Toolik, AK (Shaver et al ., 2001; Bret Harte et al ., 2002).

PAGE 132

132 To calculate the community weighted decomposition constant ( k c ), we used the following equation: (5.1) where ANPP t is the total ANPP per plot, ANPP s is the ANPP of a given species and tissu e type (leaf or stem), k s is the decomposition constant of a given species and tissue type, and n is the number of unique species and tissue types. Our common garden did not incubate all species and tissue types found at our sites. Therefore, we used decom position constants from previous studies and from similar substrates of this study. For Carex bigelowii we used a k of 0.155 (Demarco, 2011). For the evergreen leaves Empetrum nigr um Andromeda polifolia and Oxycoccus microcarpus we used 0.208, an averag e of our evergreen leaf values following DeMarco (2011). For the stems of A. polifolia E. nigr um O. microcarpus and R chamaemorus we used 0.096, an average of our woody tissue values These estimates were used for only 9.5% of total gradient ANPP and 10.2% of total CiPEHR ANPP. Data Analysis To investigate site or treatment and sampling date differences in decomposition, we performed analyses of variance (ANOVAs) in JMP (SAS, North Carolina). For the gradient, the main effects were substrate depth, si te, and year, while for CiPEHR, the main effects were the winter treatment, the summer treatment nested within the winter treatment, substrate depth, and year. At the gradient, the same locations were used every year, so location nested in site was a rando m effect. Bag was also a random effect because each bag contained the substrate at 0 10 cm and 10 20 cm. For CiPEHR, the random effects were fence, plot nested within fence, and bag. Differences

PAGE 133

133 in community weighted decomposition constants were tested wit h ANOVAs using site or treatment as main effects. Mass loss and k were arcsine square root transformed before analyses. All residuals were checked for normality and homogeneity of variances to ensure the assumptions of ANOVA were met. The ANOVAs revealed that year was a significant effect in the thaw gradient, so we ran a mixed effect multiple regression in R (R Development Team, 2012) to explore interannual controls on decomposition. We included average growing season soil temperatures at 10 cm, active layer depth (the maximum depth of unfrozen soil in autumn), average water table depth, total growing season precipitation, substrate depth, and site as explanatory variables. Water table depths were measured weekly during the growing seas on at three wells per site. Soil temperatures were also monitored at three locations per site. Site level values of water table depth and soil temperature were used in the model. At the end of September, active layer depth was measured near six of the 10 b ags per site; a site level average was used for the remaining bags (Table C 2). More i nformation on soil sensors, rain gauge, an d active layer depth monitoring can be found in Trucco et al (2012). We included water table by precipitation and active layer depth by precipitation interactions because initial graphs of the data showed evidence for them. We used the full model to optimize random effects and variance structures using AIC values following Zuur et al (2009). We used a first order autoregressive c orrelation structure within location to account for the repeated measures and a fixed variance structure with active layer depth. Once random effects were optimized, we performed a series of pair wise model comparisons using the F test, dropping the least significant explanatory variable each time (highest p value) until only significant

PAGE 134

134 explanatory variables remained (Zuur et al ., 2009). The gradient model was fitted with the lme command in the nlme package in R (Pinheiro et al 2010) using restricted maxi mum likelihood. We ran similar mixed effect multiple regressions with the CiPEHR data. CiPEHR had finer scale monitoring of environmental parameters than the gradient so that water table depth, active layer depth, and soil temperature were measured at eac location (Table C 3) The CiPEHR regression included average growing season and winter soil temperature at 10 cm, active layer depth, growing season precipitation, water table depth, substrate depth, and treatment as the explanatory variables. Base d on graphical exploration, we included an interaction term between substrate depth and water table depth. Plot was included as a random effect and no variance structure was needed. The same model minus the winter soil temperatures and substrate depth was used for CiPEHR growing season decomposition. We followed the same model comparison procedure as for the gradient to choose the best explanatory models for CiPEHR annual and growing season decomposition. All explanatory variables were centered and standard ized to better interpret interactions and to compare effect sizes (Schielzeth, 2010). Results Common Substrate Decomposition Decomposition of the common substrate cellulose differed among sites in the permafrost thaw gradient but not among treatments in the warming experiment. Annual mass loss at the permafrost thaw gradient was greatest in the extensive thaw site (f 2, 31 =7.3, p=0.0025; Fig, 1). However, there was a site by depth interaction wherein mass loss at extensive thaw only differed significantly from minimal and moderate thaw at 10

PAGE 135

135 20 cm and was statistically similar to minimal and moderate thaw at 0 10 cm (f 2, 188 =7.2, p=0.0009). On average, annual mass loss at 0 10 cm was 26% while annual mass loss at 10 20 cm was only 12%, a difference that was significant (f 1, 188 =210, p<0.0001; Fig. 1). Annual mass loss also showed significant interannual variation over seven years at the thaw gradient (f 6, 158 =43.1, p<0.0001). The CiPEHR annual mass loss data was analyzed in two ways, from 2008 through 2011 f or winter warming effects and from 2009 through 2011 for all effects, since annual decomposition was not measured in the winter warming on annual decomposition at CiPEHR o ver the three years of the experiment (f 1, 25 =2.4, p=0.13; Fig. 2). There was also no significant summer warming effect from fall 2009 through fall 2011 (f 2,15 =1.6, p=0.24; Fig. 2). Growing season mass loss at CiPEHR also did not differ among winter warmin g (f 1, 15 =0.43, p=0.52) or summer warming treatments (f 2, 15 =0.54, p=0.59; Fig. 2). As at the thaw gradient, years significantly differed: annual mass loss was lowest during 2008 2009 (f 2,44 =6.3, p=0.0038) and growing season mass loss was greatest in 2010 (f 1,20 =21, p=0.0002). Again, significantly greater mass loss occurred at 0 10 cm (30%) than at 10 20 cm (16%; three year ANOVA, f 1, 68 =36, p<0.0001). In mixed effect multiple regressions, moisture related variables explained most of the variability in com mon substrate mass loss at both sites. At the thaw gradient, annual mass loss increased as growing season precipitation increased (Table 1, Fig. 3). While active layer depth and water table also significantly affected mass loss, their effect sizes, as judg ed by the coefficients of the standardized variables, were only 34% and

PAGE 136

136 significantly interacted with precipitation. Graphic exploration showed that when growing seas on precipitation was >230 mm, there was a positive relationship between mass loss and depth to the water table, but when precipitation was <230 mm, there was a slight negative relationship. Site and depth were significant fixed effects with extensive thaw having a greater intercept and 10 20 cm having a smaller intercept than minimal thaw at 0 10 cm, the baseline intercept (Table 1). At CiPEHR, annual mass loss decreased as the depth to the water table increased (Table 1, Fig.4). Depth was a significant fix ed effect and interacted significantly with water table (Table 1); the effect of water table was greater at 0 10 cm than at 10 20 cm as indicated by the shallower slope for mass loss at 10 20 cm (Fig. 4). Lastly, growing season mass loss at CiPEHR also inc reased significantly with precipitation (Table 1). Growing season average soil temperature and winter season average soil temperature (tested at CiPEHR only) variables were dropped from the models because they did not significantly affect mass loss. Commo n Garden and Plant Community Decomposition Over three years in the common garden, vascular plant leaves decomposed the fastest while mosses and woody material decomposed the slowest (Table 2). Out of the vascular plant leaves, Eriophorum vaginatum decompos ed the slowest and Betula nana initial percent nitrogen (R 2 =0.61, p=0.0026, n=12). At the thaw gradient, plant community weighted decomposition was significantly faster in extensive thaw than at minimal or moderate thaw (f 2 =6.97, p=0.0072; Fig. 5). In contrast, there were no winter warming (f 1, 39 =0.12, p=0.7272) or summer warming effects (f 2, 39 =1.00, p=0.3757) on community weighted decomposition at CiPEHR.

PAGE 137

137 Discussion M oisture Controls Decomposition Moisture was the most important control on decomposition at both the thaw gradient and CiPEHR. In all the regression models, soil moisture related variables were the most important predictors: precipitation for annual mass lo ss at the thaw gradient and growing season mass loss at CiPEHR and water table for annual mass loss at CiPEHR. While temperature was not a factor in terms of explaining interannual variability or environmental heterogeneity, soil profile temperature gradie nts likely controlled decomposition differences among depths. At the thaw gradient and CiPEHR, the heterotrophic respiration of old soil C (from 15 80 cm in the soil profile) was increased by thaw and warming, not soil moisture (Hicks Pries et al ., in pres s; Chapter 3). The contrasting results of this study suggest different mechanisms control decomposition at different depths in the soil profile. While decomposition at depth in permafrost ecosystems is controlled by temperature, decomposition at the surfac e appears more strongly controlled by moisture. Given that soil below 20 cm is usually waterlogged at the thaw gradient (Trucco et al ., 2012) and CiPEHR (Natali et al ., 2012), these results imply soil moisture may become a more important driver of deep soi l decomposition if the water table were to decrease. Investigating the controls on surface decomposition in more detail, we found differences among the study sites. Shallower water tables had a negative effect on decomposition at the gradient and a positi ve effect on decomposition at CiPEHR. The average growing season water table was closer to the surface at the gradient (18 cm) than at CiPEHR (22 cm). If this shallow water table was more consistent at gradient, it may have increase d decomposition retarding anaerobic conditions. Furthermore, the

PAGE 138

138 effect of water table on decomposition at the gradient was dependent on the amount of precipitation. Having the water table close to the surface sped up decomposition in dry years but slowed d ecomposition in wet years, possibly due to anaerobic conditions. At growing seasons with rainfall greater than 250 mm during its short, three year record. That depth wa s a significant effect in both models implies soil temperature had a role in controlling surface decomposition, but its role was limited to the effects caused by the temperature depth gradient within soils. Decomposition was likely faster at 0 10 cm becaus e growing season soil temperatures are 2.5 C warmer on average at 10 cm than at 20 cm. Moisture may have also play ed a role in the depth differences. The slower increase in 10 20 cm mass loss as the water table becomes shallower at CiPEHR could be due to i nundation slowing down decomposition. While the colder temperature of 10 20 cm depths could limit the ability of microbes to respond to water table increases, this effect was not seen with precipitation at the thaw gradient, implying it may be a result of inundation by the water table itself. While the water table by precipitation and water table by depth interactions indicate d inundation may begin to slow decomposition, overall it wa s surprising decomposition rates did not reach a threshold of soil moistu re that caused them to greatly decrease In fact, some of the greatest annual mass loss measured in this study was from the water track thermokarsts down slope from extensive thaw despite the water table being at or near the surface throughout much of the growing season there. In these thermokarsts, mass loss was 50% at 0 10 cm and 42% at 10 20 cm. In the dry areas adjacent to the karsts, mass loss was only 17% and 3% at 0 10 and 10 20 cm,

PAGE 139

139 respectively (f 1,14 =13.0, p=0.0029, n=8). Fast decomposition despite inundation may have been biological oxygen demand than soil water in warmer ecosystems (Gebauer et al ., 1996). Tundra inundation does not quickly result in anaerobic conditions. Furthe rmore, the water is not stagnant in water track thermokarsts; its slow flow may enhance oxygenation. Site was a significant effect in the thaw gradient but treatment was not a significant effect at CiPEHR, likely because thaw had been ongoing for only thre e years in CiPEHR compared to decades at the thaw gradient. Because physical properties of the soil environment such as thaw depth, temperature, and water table were included in the model, faster decomposition at extensive thaw was likely due to a factor w e did not measure. Extensive thaw has larger biomass nitrogen pools (Schuur et al., 2007), suggesting greater N availability, which can increase decomposition rates (McClaugherty et al ., 1985; Hobbie & Gough, 2004), especially of our nitrogen devoid common substrate. Extensive thaw also has greater plant biomass (Schuur et al ., 2007), which may prime decomposition through root exudates (Kuzyakov et al ., 2007; de Graff et al ., 2010). Soil moisture will be affected by climate change and permafrost thaw in se veral ways. Climate models predict that latitudes north of 60 will experience 10 20% more precipitation as a result of climate change (ACIA, 2005; IPCC, 2007) due to increased atmospheric transport of water vapor from low to high latitudes. Warmer air mas ses will also be able to hold more water vapor, and warmer temperatures will lead to increased evapotranspiration rates, adding to the available water vapor. However, increased

PAGE 140

140 evapotranspiration will also dry out the soil surface faster, so whether increa sed precipitation leads to more soil moisture depends partly on the timing and magnitude of precipitation events. Permafrost thaw alters microtopography and local hydrology in ways both obvious and subtle. When permafrost starts to thaw, there is an increa se in soil moisture near the surface. At CiPEHR, winter warming has increased the height of the water table (from the frozen soil to the top of the water table), which has led to a 5% increase in surface (0 20 cm) volumetric water content (Natali et al ., i n review). Thawing ice rich permafrost causes significant ground subsidence (Osterkamp et al ., 2009) that can lead to the formation of permafrost ponds on flat terrain or water tracks on gently sloping terrain (Jorgenson & Osterkamp, 2005). These thermokar st types increase water availability locally where the ground has subsided but also drain adjacent areas, decreasing surface soil moisture nearby. Permafrost thaw may therefore increase soil moisture heterogeneity on a landscape level, increasing surface d ecomposition in some areas, while reducing surface decomposition in others. Eventually, complete loss of permafrost could substantially lower water tables, greatly reducing soil moisture, and making the surface soil microbes solely dependent upon precipita tion. One model predicted that as permafrost degrades over the next century, there will be an initial increase in soil moisture followed by a decline in the areal extent of high latitude wetlands (Avis et al ., 2011). Surface soil respiration is an importan t component of ecosystem respiration. Annual mass loss in the top 10 cm of soil was positively related to total growing season ecosystem respiration at CiPEHR (regression with plot as a random effect, p=0.0036) and the thaw gradient (regression with repeat ed measures correlated within location,

PAGE 141

141 in determining ecosystem respiration in tundra, which may be driven, in part, by surface soil decomposition. Zona et al (2012) found that e xperimental flooding of tundra increased ecosystem respiration. In the Canadian low arctic, ecosystem respiration was greater in wet and mesic tundra than dry tundra (Dagg & Lafleur, 2011). However, a meta analysis of tundra warming experiments found that ecosystem respiration responses to warming were smaller where conditions were moist or wet (Oberbauer et al ., 2007). These negative responses to soil moisture may be driven by decreas ed autotrophic respiration and not heterotrophic surface soil decompositi on, which responds positively to moisture. Community Decomposition Community level decomposition constants are a way to estimate how the permafrost thaw and warming induced changes to plant communities affect ecosystem decomposition rates. In CiPEHR, there were no treatment differences in community weighted decomposition constant. Although, three years of winter warming increased graminoid productivity (Natali et al ., 2012), there has not been enough time to cause substantial differences in the plant commun ity as have occurred at the thaw gradient over decades. At the thaw gradient, decreased graminoid abundance and increased shrubs in extensive thaw (Schuur et al ., 2007) led to an overall increase in the community weighted decomposition rate constant. A rec ent natural gradient study also found increasing community weighted mass loss in sites with greater shrub abundance community decomposition rates would decrease with shr ub expansion due to the increased production of recalcitrant woody tissue.

PAGE 142

142 Less graminoid production at extensive thaw increased community decomposition rates because our most abundant graminoid, E. vaginatum had a decomposition constant of only 0.105, which was only slightly greater than woody tissue and was much smaller than shrub leaf litter. This constant was 37 to 50% less than previous estimates for E. vaginatum from Toolik, AK (Hobbie & Gough, 2004; De Marco, 2011). Our E. vaginatum litter decomposed slower because it had about half the nitrogen and double the C:N ratios as E. vaginatum litter from those studies. Collection methods of senescent tissue were similar, so differences in litter quality were l ikely caused by site. Our woody tissue had slightly faster decomposition constants (ranged from 0.07 to 0.011) than found in Toolik by Demarco (2011; ranged from 0.05 to 0.08), both of which were faster than the 0.025 constant measured by Hobbie & Gough (2 004). The increase in woody decomposition rates at our site may be due to warmer mean annual temperatures or our use of live stems and roots. If our wood decomposed too fast due to being collected live, the community decomposition constants may be overesti mated. Using average stem k values from the previous studies (0.07 and 0.025), community decomposition constants at extensive thaw decreased by 8 to 14% and minimal thaw decomposition constants decreased 4 to 7%. Despite our use of live litter, decompositi on of our evergreen leaf litter did not systematically differ from previous studies (Hobbie & Gough 2004; DeMarco, 2011): V. vitis idaea had faster decomposition rates, and R. subarticum had slower decomposition rates. Overall decomposition differences amo ng our substrates were smaller than the differences found among the same substrates in Hobbie & Gough (2004).

PAGE 143

143 To calculate community weighted decomposition, several assumptions were made. First, they ignore potential site by substrate interactions that may affect intraspecific litter quality. Unlike Hobbie & Gough (2004) and Demarco (2011), we did not correct for site differences in our estimates because mass loss of our common substrate did not differ among sites at 0 10 cm. One oversight of community deco mposition estimates is that they do not include moss productivity, which could greatly affect estimates because mosses decompose slowly compared to other functional groups (Lang et al ., 2009). The CiPEHR estimate would not be affected by this omission as m oss productivity was similar across all treatments (Natali et al ., 2012). The latest moss production estimate at the thaw gradient is from 2004 when extensive, moderate, and minimal thaw had 147, 55, and 25 g m 2 y 1 of moss production, respectively (Schuur et al ., 2007). If we include those ANPP numbers in the calculation of a site averaged community decomposition constant (using an average moss decomposition constant from Table 5 2), decomposition constants become similar among sites, ranging from 0.122 to 0.129. Taking into account that slightly faster decomposing Sphagnum mosses (relative to Dicranum ) are likely driving the moss decomposition constant only increases t o 0.137, still less than the vascular plant only estimate of 0.154. Relative Effects of Environment and Substrate Changes The relative effects of moisture on decomposition are greater than the effect of substrate shifts. Increasing precipitation by 50% fro m 200 to 300 mm a growing season increases mass loss in the top 10 cm of soil by almost 100%. Decreasing the depth to the water table by 50% from 30 to 15 cm increases mass loss in the top 10 cm of soil by

PAGE 144

144 over 100% In contrast, a 57% decrease in graminoi d productivity and a concurrent 47% increase in shrub productivity caused by permafrost thaw (Schuur et al. 2007) only increased the decomposition co nstant by 21%, and including the moss productivity increase in that estimate negates that increase. Theref ore permafrost thaw and climate soil environment than to plant communities. As climate change causes permafrost to thaw and precipitation to increase across tundra ecosy stems, decomposition rates will change as a result of shifting soil moisture availability. This research suggests that increased moisture will lead to increased decomposition and that moisture has a more significant effect on decomposition than soil temper ature at the top of the soil profile. Much of the current research on the permafrost thaw climate change feedback focuses on temperature effects because ultimately temperature controls whether the soil is frozen or not. Once thawed, soil temperature may be come a less important driver relative to soil moisture though more research is needed to determine the effects of moisture deeper in the soil profile Wetter tundra may increase all heterotrophic C O 2 losses, and is another mechanism by which the permafrost thaw climate change feedback can occur.

PAGE 145

145 Table 5 1. Multiple regression results for annual and growing season decomposition of a common substrate at the permafrost thaw gradient and warming experim ent (CiPEHR). Results are of the models run with only the significant predictors after model selection (see Methods for details). Model Coefficients Estimate SE t value p value Gradient Intercept 0.4 86 0.02 5 19.3 <0.00001*** (annual) Moderate 0.015 0.03 2 0. 46 0. 6456 Extensive 0. 162 0.03 3 4.94 <0.00001 ** Depth ( 10 20 ) 0.2 30 0.0 22 10.4 <0.00001*** Water Table 0.0 351 0.00 90 3.89 0.0001 ** Active Layer 0.0 454 0.001 2 3.81 0.0002 ** Precipitation 0.1 34 0.00 96 14.0 <0.00001*** WT:Precip 0.034 8 0.00 85 4.08 0.0001 ** CiPEHR Intercept 0.561 0.029 19.4 <0.00001*** (annual) Depth (10 20) 0.182 0.028 6.54 <0.00001*** Water Table 0.139 0.026 5.40 <0.00001*** Depth (10 20):WT 0.0846 0.028 3.02 0.032** CiPEHR Intercept 0.017 0.100 0.17 0.8647 ( growing season ) Precipitation 0.0021 0.0004 4.73 0.0001**

PAGE 146

146 Table 5 2. Decomposition constants (k, n=5) and initial litter quality (n=3) from the common garden experiment near the permafrost thaw gradient in Healy, Alaska. Wood refers to a 1:1 mixture of stem and root tissue. Species k (1/year) C:N N (%) Solubles (%) Hemicellulose + bound protein (%) Cellulose (%) Lignin (%) Eriophorum vaginatum 0.105 0.012 ABC 102 6 0.45 0.02 21.4 0.42 34.5 0.45 36.2 0.09 7.68 0.58 Betula nana 0.226 0.016 AB 37 1 1.36 0.05 43.3 1.89 15.6 0.37 13.8 0.61 27.0 1.4 Rubus chamaemorus 0.202 0.014 A 37 0.8 1.26 0.01 61.4 1.58 13.8 0.79 13.5 0.66 10.7 0.41 Vaccinium uliginosum 0.171 0.014 AB 78 2 0.64 0.01 49.2 0.32 9.5 0.56 14.9 0.18 26.0 0.77 Rhododendron subarcticum 0.201 0.015 ABC 42 0.3 1.27 0.004 64.1 0.79 8.1 0.46 12.8 0.40 14.5 0.28 Vaccinium vitis idaea 0.214 0.015 AB 59 1.3 0.84 0.01 64.3 1.76 11.2 0.64 13.4 0.70 10.5 0.28 Dicranum spp. 0.059 0.010 C 103 4 0.42 0.02 23.3 0.87 27.7 1.1 34.3 1.7 0.50 0.36 Sphagnum spp 0.108 0.022 ABC 68 2 0.63 0.02 23.0 0.76 28.1 1.0 48.0 1.2 23.2 1.38 B. nana wood 0.087 0.011 BC 56 4 0.93 0.07 34.7 1.9 16.9 1.4 24.8 1.8 21.2 0.52 V. uliginosum wood 0.108 0.011 ABC 76 6 0.67 0.07 29.2 0.22 17.7 0.29 31.6 0.61 21.5 0.15 R. subarcticum wood 0.071 0.010 C 87 6 0.56 0.02 32.3 1.2 18.8 0.48 26.9 0.71 19.8 0.65 V. vitis idaea wood 0.119 0.012 ABC 70 2 0.70 0.02 38.3 2.1 15.7 0.55 25.8 1.02 19.8 0.65

PAGE 147

147 Figure 5 1. Annual percent mass loss from the common substrate decomposition bags at the three thaw gradient sites from September 2004 through September 2011 (n=10 per site per year). The bags were split into two depths 0 10 cm and 10 20 cm and were incubated from Se ptember of one year until September of the following year. The year listed is the year in which the bag was picked up. Capital letters not shared indicate significant differences among sites, lowercase letters not shared indicate significant differences am ong years, and the asterisk indicate the depths were significantly different

PAGE 148

148 Figure 5 2. Growing season (GS, top panel) and annual (bottom two panels) percent mass loss from the common substrate decomposition bags at CiPEHR (n=6 per treatment per year). Annual percent mass loss was measured at two depths 0 10 cm and 10 20 cm. Capital letters not shared indicate significant differences among years for annual mass loss, and the asterisk represents significant differences among years for growing s There were no significant differences among treatments.

PAGE 149

149 Figure 5 3. Annual decomposition increased with more growing season precipitation at the thaw gradient. The points are the actual data and the lines show the average pre dicted relationship between precipitation and percent mass loss when the other significant predictors (water table depth and active layer depth; see Table 1) are held at their respective means. The solid and dashed lines are the relationships with decompos ition at 0 10 cm and 10 20 cm depths, respectively. The precipitation values of the 10 20 cm data were offset by 3 mm to show all data points.

PAGE 150

150 Figure 5 4. Annual decomposition increased with shallower water tables at CiPEHR. The points are the actual data and the lines show the average predicted relationship between the average growing season depth from the soil surface to the water table and annual percent mass loss. The solid and dashed lines are the relationships with decomposition at 0 10 cm and 10 20 cm depths, respectively.

PAGE 151

151 Figure 5 5. Community weighted decomposition constants (k) for the sites in the gradient (n=6) and the CiPEHR treatments (n=12). Capital letters not shared indicate significant differences among sites at the thaw gradient (

PAGE 152

152 CHAPTER 6 CONCLUSION The frozen soil conditions of permafrost have caused arctic and boreal ecosystems to be carbon (C) sinks throughout the Holocene and, if they were unglaciated, the Pleistocene. In Chapter 2 I found that the study site, a subarctic moist acidic tussock tundra ha d been accumulating C at a rate of 2.3 g C m 2 y 1 throughout the Holocene. Radiocarbon values revealed that soil C below 80 cm in the soil profile was 8,000 to 10,000 years old. Decadal net ecosys tem production (analagous to net ecosystem carbon balance; Chapin et al ., 2006) estimates were mostly positive, and the ecosystem ha d been accumulating an average of 14.4 g C m 2 y 1 over the past fifty years. However accumulation models indicate d measura ble vertical soil C accumulation may have ceased within the past 18 years potentially due to climate change and permafrost thaw Th is lack of vertical accumulation could be due to surface decomposition rates increasing so that C inputs were balanced by ou tputs or to plant derived C inputs shifting from the soil surface to the rooting zone as would occur when moss production is replaced by shrubs. The cessation may also have been a methodological artifact, although it was not seen in a similar study that u sed the same method (Trumbore & Harden, 1997). If the lack of vertical accretion within the past 20 years is not a methodological artifact, the cessation of vertical accumulation after millennia indicates climate change and permafrost thaw are changing thi carbon cycle. increasing ecosystem respiration rates (Trucco et al ., 2012; Natali et al ., in review). In Chapters 3 and 4 I found this increase was being driven by heterotrophic respiration of

PAGE 153

153 old soil C and autotrophic respiration. I was only able to measure th e crucial loss of old soil C using isotopes, particularly radiocarbon, because the increase is not evident from flux measurements alone. At the thaw gradient, the proportion of ecosystem respiration from old soil increased with deepening thaw, and in CiPEHR, the proportion increased with warmer soil temperatures. In the thaw gradient, total growing season old soil C loss was 67 103% greater in areas where perma frost thaw was deep relative to where thaw was shallow. At CiPEHR, growing season old C loss was 34 to 150% greater in summer and winter warming treatments than the control. The proportional contributions of autotrophic respiration increased with deepening thaw at the thaw gradient and were largest in the winter warming treatments at CiPEHR, where thaw depth was 10% deeper than in the winter warming control. Overall, permafrost thaw and warming in moist acidic tussock tundra is causing an increase in ecosy stem respiration that is being driven more by autotrophs than heterotrophs. At both the thaw gradient and CiPEHR, the ratio of autotrophic to heterotrophic respiration is greater where there is more permafrost thaw and warming (Fig. 6 1). The increase in a utotrophic respiration is caused by the demands of increased primary production, which is offsetting growing season old C losses at present (Trucco et al ., 2012; Natali et al ., in review). However, the loss of old soil C is concerning in the long term beca use sustained losses of old soil C are likely given the climate change trajectory of increasingly warm temperatures Due to the relative sizes of the tundra plant C pool (0.45 0.63 kg m 2 ; Natali et al ., 2012) and soil C pool (60 kg m 2 ; Hicks Pries et al ., 2012), it is likely old soil C losses will eventually surpass gains in plant biomass, even if climate change causes boreal forest with its greater plant

PAGE 154

154 biomass C pool (6 kg C m 2 ; Gower et al ., 2001; Goulden et al ., 2011 ) to replace tundra. Furthermo re, these studies only examined growing season dynamics of ecosystem respiration. It is likely warming also increases fall and winter old soil C losses, which would more than offset growing season C uptake (Natali et al ., in review). Another way permafrost is by altering soil moisture, which affects a component of heterotrophic respiration, decomposition in the In Chapter 5 I found that soil moisture was an important control of decomp osition at both the thaw gradient and CiPEHR. At the thaw gradient, a 50% increase in growing season precipitation increased mass loss of a common substrate by almost 1 00%. At CiPEHR, a 50% shallower water table increased mass loss by over 10 0%. In contras t to soil moisture, changes to plant litter due to permafrost thaw induced shifts in plant communities affected decomposition by less than 21% are driven more by changes to precipitat ion and soil moisture than by changes to plant communities. As climate change causes permafrost thaw (e.g., Osterkamp et al ., 2009) and precipitation increases (ACIA, 200 5 ) across t undra ecosystems, decomposition rates will change as soil moisture increase s in some areas and decreases in others Permafrost thaw is causing the loss of soil C that has been sequestered from the atmosphere for hundreds to hundreds of thousands of years. These increases may currently be compensated for by increased primary prod uctivity, but that balance is temporary. Increased soil temperatures will accelerate the loss of old soil C, at least where soils have adequate soil moisture While ultimately temperature determines whether or not permafrost thaws, after thaw, rates of soil C loss may be determined by

PAGE 155

155 soil mo isture availability Soil moisture is particularly important at the top of the soil profile, where tundra can dry out quickly However, complete loss of permafrost may enhance drainage making decomposition of deeper soil moisture dependent as well. Overall, the warming world is driving permafrost systems to become a C source a trend that will be moderated or exacerbated depending on changes to soil moisture availability.

PAGE 156

156 Figure 6 1. Permafrost thaw and warming increases the ratio of autotrophic to he terotrophic respiration during the growing season The increase was not significant across the thaw gradient but at CiPEHR, winter warming t reatments had significantly higher ratios ( =0.05). This increase coincides with increasing net primary productivity.

PAGE 157

157 157 APPENDIX A SUPPLEMENTAL TABLES FOR CHAPTER 3 Table A 13 14 C, flux rates, and estimated carbon pools and soil temperatures of individual soil sections used to calculate the depth 13 14 C of young soil and old soil. The average % carbon and bulk density for each depth section and location from Hicks Pries et al. ( 2012) were used to calculate the g carbon per m 2 and the flux on a per g carbon basis. The soil cores in that study and the current study were taken from similar locations. The shallow cores sampled in May 2009 were only used to test for monthly soil isotope variability and not for the youn g soil calculations because the full 15 cm were not fully thawed. For the old soil calculation, only the deepest cores were used (deep cores 1, 2, 3, and 8) so the calculations would be based on complete profiles. Date Sampled Core Type Location Rep Depth cm 14 C 13 C Carbon Flux Soil Temperature (C) g C m 2 g C hr 1 Jun 08 Jul 08 Aug 08 May 09 Jul 09 Sep 09 May 2009 Shallow 1 1 0 5 68.1 28.0 1508 0.11 5.2 7.3 7.0 6.0 13.0 5.0 May 2009 Shallow 1 1 5 8 84.8 27.2 838 0.09 4.7 6.8 6.5 5.7 11.9 4.1 May 2009 Shallow 1 2 0 5 56.6 27.7 1508 0.16 5.2 7.3 7.0 6.0 13.0 5.0 May 2009 Shallow 1 2 5 11 104.2 27.3 1675 0.06 4.7 6.8 6.5 5.7 11.9 4.1 May 2009 Shallow 1 3 0 5 78.9 27.3 1508 0.14 5.2 7.3 7.0 6.0 13.0 5.0 May 2009 Shallow 1 3 5 8 113.4 26.7 838 0.09 4.7 6.8 6.5 5.7 11.9 4.1 May 2009 Shallow 2 1 0 5 103.3 27.0 1645 0.24 5.2 7.3 7.0 6.0 13.0 5.0 May 2009 Shallow 2 1 5 8 120.0 26.7 858 0.07 4.7 6.8 6.5 5.7 11.9 4.1 May 2009 Shallow 2 2 0 5 54.2 28.3 1645 0.09 5.2 7.3 7.0 6.0 13.0 5.0 May 2009 Shallow 2 2 5 9 72.2 27.4 1145 0.06 4.7 6.8 6.5 5.7 11.9 4.1 May 2009 Shallow 2 3 0 5 78.6 27.8 1645 0.09 5.2 7.3 7.0 6.0 13.0 5.0 May 2009 Shallow 2 3 5 10 177.1 27.1 1431 0.08 4.7 6.8 6.5 5.7 11.9 4.1 May 2009 Shallow 3 1 0 5 NA 26.2 1142 0.18 5.2 7.3 7.0 6.0 13.0 5.0 May 2009 Shallow 3 2 0 5 70.8 27.6 1142 0.13 5.2 7.3 7.0 6.0 13.0 5.0 May 2009 Shallow 3 2 5 13 127.0 26.2 1869 0.10 4.7 6.8 6.5 5.7 11.9 4.1 May 2009 Shallow 3 3 0 5 83.8 27.7 1142 0.12 5.2 7.3 7.0 6.0 13.0 5.0 May 2009 Shallow 3 3 5 12 109.4 27.1 1635 0.05 4.7 6.8 6.5 5.7 11.9 4.1 May 2009 Shallow 3 4 0 5 54.0 29.3 1142 0.13 5.2 7.3 7.0 6.0 13.0 5.0 May 2009 Shallow 3 4 5 15 83.9 27.4 2336 0.16 4.7 6.8 6.5 5.7 11.9 4.1

PAGE 158

158 158 Table A 1. Continued Date Sampled Core Type Location Rep Depth cm 14 C 13 C Carbon Flux Soil Temperature (C) g C m 2 g C hr 1 Jun 08 Jul 08 Aug 08 May 09 Jul 09 Sep 09 May 2009 Shallow 3 4 15 22 113.9 25.8 2693 0.14 1.5 3.4 4.2 1.2 4.3 2.6 July 2009 Shallow 1 1 0 5 79.3 27.0 1508 0.29 5.2 7.3 7.0 6.0 13.0 5.0 July 2009 Shallow 1 1 5 15 104.6 26.6 2792 0.36 4.7 6.8 6.5 5.7 11.9 4.1 July 2009 Shallow 1 1 15 25 29.9 26.2 4878 0.47 1.3 3.0 3.8 0.9 3.8 2.4 July 2009 Shallow 1 2 0 5 62.7 28.1 1508 0.62 5.2 7.3 7.0 6.0 13.0 5.0 July 2009 Shallow 1 2 5 15 106.1 27.2 2792 0.24 4.7 6.8 6.5 5.7 11.9 4.1 July 2009 Shallow 1 2 15 25 52.8 26.5 4878 0.16 1.3 3.0 3.8 0.9 3.8 2.4 July 2009 Shallow 1 3 0 5 90.9 26.9 1508 0.31 5.2 7.3 7.0 6.0 13.0 5.0 July 2009 Shallow 1 3 5 15 39.2 28.4 2792 0.15 4.7 6.8 6.5 5.7 11.9 4.1 July 2009 Shallow 1 3 15 25 1.7 24.2 4878 0.18 1.3 3.0 3.8 0.9 3.8 2.4 July 2009 Shallow 2 1 0 5 113.3 27.1 1645 0.36 5.2 7.3 7.0 6.0 13.0 5.0 July 2009 Shallow 2 1 5 15 93.0 26.2 2862 0.14 4.7 6.8 6.5 5.7 11.9 4.1 July 2009 Shallow 2 1 15 25 30.3 25.7 5532 0.14 1.3 3.0 3.8 0.9 3.8 2.4 July 2009 Shallow 2 2 0 5 56.6 29.0 1645 0.22 5.2 7.3 7.0 6.0 13.0 5.0 July 2009 Shallow 2 2 5 15 105.5 27.2 2862 0.12 4.7 6.8 6.5 5.7 11.9 4.1 July 2009 Shallow 2 2 15 25 34.1 24.7 5532 0.21 1.3 3.0 3.8 0.9 3.8 2.4 July 2009 Shallow 2 3 0 5 80.7 27.9 1645 0.34 5.2 7.3 7.0 6.0 13.0 5.0 July 2009 Shallow 2 3 5 15 106.3 26.8 2862 0.26 4.7 6.8 6.5 5.7 11.9 4.1 July 2009 Shallow 2 3 15 25 18.6 24.1 5532 0.48 1.3 3.0 3.8 0.9 3.8 2.4 July 2009 Shallow 3 1 0 5 57.1 28.2 1142 0.29 5.2 7.3 7.0 6.0 13.0 5.0 July 2009 Shallow 3 1 5 15 88.5 26.1 2336 0.13 4.7 6.8 6.5 5.7 11.9 4.1 July 2009 Shallow 3 1 15 25 62.7 24.4 3847 0.20 1.3 3.0 3.8 0.9 3.8 2.4 July 2009 Shallow 3 2 0 5 103.8 26.2 1142 0.11 5.2 7.3 7.0 6.0 13.0 5.0 July 2009 Shallow 3 2 5 15 39.1 25.3 2336 0.11 4.7 6.8 6.5 5.7 11.9 4.1 July 2009 Shallow 3 2 15 25 10.6 24.9 3847 0.22 1.3 3.0 3.8 0.9 3.8 2.4 July 2009 Shallow 3 3 0 5 91.8 26.4 1142 0.28 5.2 7.3 7.0 6.0 13.0 5.0 July 2009 Shallow 3 3 5 15 35.9 25.0 2336 0.24 4.7 6.8 6.5 5.7 11.9 4.1

PAGE 159

159 159 Table A 1. Continued Date Sampled Core Type Location Rep Depth cm 14 C 13 C Carbon Flux Soil Temperature (C) g C m 2 g C hr 1 Jun 08 Jul 08 Aug 08 May 09 Jul 09 Sep 09 July 2009 Shallow 3 3 15 25 78.3 25.9 3847 0.48 1.3 3.0 3.8 0.9 3.8 2.4 Sept 2009 Shallow 1 1 0 5 112.0 26.9 1508 0.29 5.2 7.3 7.0 6.0 13.0 5.0 Sept 2009 Shallow 1 1 5 15 56.0 25.8 2792 0.12 4.7 6.8 6.5 5.7 11.9 4.1 Sept 2009 Shallow 1 1 15 25 25.8 25.6 4878 0.30 1.3 3.0 3.8 0.9 3.8 2.4 Sept 2009 Shallow 1 2 0 5 64.7 27.6 1508 0.51 5.2 7.3 7.0 6.0 13.0 5.0 Sept 2009 Shallow 1 2 5 15 119.6 26.3 2792 0.24 4.7 6.8 6.5 5.7 11.9 4.1 Sept 2009 Shallow 1 2 15 25 34.7 24.8 4878 0.23 1.3 3.0 3.8 0.9 3.8 2.4 Sept 2009 Shallow 1 3 0 5 NA 26.7 1508 0.30 5.2 7.3 7.0 6.0 13.0 5.0 Sept 2009 Shallow 1 3 5 15 NA 25.1 2792 0.10 4.7 6.8 6.5 5.7 11.9 4.1 Sept 2009 Shallow 1 3 15 25 NA 24.1 4878 0.05 1.3 3.0 3.8 0.9 3.8 2.4 Sept 2009 Shallow 2 1 0 5 106.1 26.2 1645 0.24 5.2 7.3 7.0 6.0 13.0 5.0 Sept 2009 Shallow 2 1 5 15 77.7 26.3 2862 0.20 4.7 6.8 6.5 5.7 11.9 4.1 Sept 2009 Shallow 2 1 15 25 25.3 25.3 5532 0.34 1.3 3.0 3.8 0.9 3.8 2.4 Sept 2009 Shallow 2 2 0 5 60.7 27.2 1645 0.39 5.2 7.3 7.0 6.0 13.0 5.0 Sept 2009 Shallow 2 2 5 15 82.8 26.9 2862 0.15 4.7 6.8 6.5 5.7 11.9 4.1 Sept 2009 Shallow 2 2 15 25 34.9 26.1 5532 0.23 1.3 3.0 3.8 0.9 3.8 2.4 Sept 2009 Shallow 2 3 0 5 NA 28.0 1645 0.20 5.2 7.3 7.0 6.0 13.0 5.0 Sept 2009 Shallow 2 3 5 15 NA 26.3 2862 0.13 4.7 6.8 6.5 5.7 11.9 4.1 Sept 2009 Shallow 2 3 15 25 NA 26.6 5532 0.08 1.3 3.0 3.8 0.9 3.8 2.4 Sept 2009 Shallow 3 1 0 5 51.1 28.0 1142 0.25 5.2 7.3 7.0 6.0 13.0 5.0 Sept 2009 Shallow 3 1 5 15 74.1 26.2 2336 0.12 4.7 6.8 6.5 5.7 11.9 4.1 Sept 2009 Shallow 3 1 15 25 89.1 26.1 3847 0.11 1.3 3.0 3.8 0.9 3.8 2.4 Sept 2009 Shallow 3 2 0 5 92.0 25.9 1142 0.14 5.2 7.3 7.0 6.0 13.0 5.0 Sept 2009 Shallow 3 2 5 15 48.3 25.1 2336 0.12 4.7 6.8 6.5 5.7 11.9 4.1 Sept 2009 Shallow 3 2 15 25 10.3 25.1 3847 0.16 1.3 3.0 3.8 0.9 3.8 2.4 Sept 2009 Shallow 3 3 0 5 NA 26.6 1142 0.31 5.2 7.3 7.0 6.0 13.0 5.0 Sept 2009 Shallow 3 3 5 15 NA 26.2 2336 0.29 4.7 6.8 6.5 5.7 11.9 4.1

PAGE 160

160 160 Table A 1. Continued Date Sampled Core Type Location Rep Depth cm 14 C 13 C Carbon Flux Soil Temperature (C) g C m 2 g C hr 1 Jun 08 Jul 08 Aug 08 May 09 Jul 09 Sep 09 Sept 2009 Shallow 3 3 15 25 NA 26.0 3847 0.21 1.3 3.0 3.8 0.9 3.8 2.4 May 2009 Deep 1 25 35 56.5 26.4 5317 0.07 0.5 1.5 2.3 0.1 1.7 1.6 May 2009 Deep 1 35 45 128.6 26.1 9997 0.10 0.1 0.8 1.6 0.4 1.0 1.2 May 2009 Deep 1 45 55 188.5 24.8 9848 0.06 0.1 0.3 1.1 0.6 0.7 0.9 May 2009 Deep 1 55 65 324.0 25.6 9398 0.02 0.3 0.1 0.8 0.7 0.5 0.7 May 2009 Deep 1 65 75 462.2 24.9 6674 0.05 0.4 0.1 0.6 0.8 0.4 0.6 May 2009 Deep 1 75 85 410.1 24.8 7364 0.06 0.5 0.2 0.4 0.8 0.3 0.6 May 2009 Deep 2 27 35 34.1 27.1 5317 0.36 0.4 1.4 2.2 0.1 1.6 1.5 May 2009 Deep 2 35 45 26.7 26.4 9997 0.21 0.1 0.8 1.6 0.4 1.0 1.2 May 2009 Deep 2 45 55 5.3 26.4 7878 0.28 0.1 0.3 1.1 0.6 0.7 0.9 May 2009 Deep 2 55 65 87.3 26.6 9398 0.14 0.3 0.1 0.8 0.7 0.5 0.7 May 2009 Deep 2 65 78 151.9 26.9 8676 0.06 0.4 0.1 0.5 0.8 0.3 0.6 May 2009 Deep 2 78 89 175.2 25.6 8252 0.09 0.5 0.3 0.3 0.8 0.3 0.5 May 2009 Deep 3 25 35 71.4 27.3 5317 0.27 0.5 1.5 2.3 0.1 1.7 1.6 May 2009 Deep 3 35 45 35.5 27.0 9997 0.29 0.1 0.8 1.6 0.4 1.0 1.2 May 2009 Deep 3 45 54 4.4 26.1 8863 0.32 0.1 0.4 1.1 0.6 0.7 0.9 May 2009 Deep 3 54 65 233.0 22.9 10338 0.03 0.1 0.4 1.1 0.6 0.7 0.9 May 2009 Deep 3 65 75 340.8 21.0 6674 0.03 0.4 0.1 0.6 0.8 0.4 0.6 May 2009 Deep 3 75 89.7 427.2 19.8 8422 0.03 0.5 0.3 0.3 0.8 0.3 0.5 May 2009 Deep 4 25 35 84.1 26.0 8656 0.08 0.5 1.5 2.3 0.1 1.7 1.6 May 2009 Deep 4 35 45 133.4 24.0 10504 0.02 0.1 0.8 1.6 0.4 1.0 1.2 May 2009 Deep 4 45 58 160.1 26.4 8833 0.09 0.2 0.3 1.0 0.6 0.6 0.9 May 2009 Deep 5 25 35 38.1 27.3 8656 0.12 0.5 1.5 2.3 0.1 1.7 1.6 May 2009 Deep 5 35 44 33.6 26.4 9453 0.09 0.1 0.8 1.6 0.4 1.0 1.2 May 2009 Deep 5 44 52 55.8 26.1 5770 0.17 0.1 0.4 1.2 0.6 0.7 0.9 May 2009 Deep 6 25 35 89.0 26.4 8656 0.09 0.5 1.5 2.3 0.1 1.7 1.6 May 2009 Deep 6 35 41 88.4 26.4 6302 0.04 0.2 0.9 1.7 0.4 1.1 1.2

PAGE 161

161 161 Table A 1. Continued Date Sampled Core Type Location Rep Depth cm 14 C 13 C Carbon Flux Soil Temperature (C) g C m 2 g C hr 1 Jun 08 Jul 08 Aug 08 May 09 Jul 09 Sep 09 May 2009 Deep 6 41 56 96.7 26.5 11560 0.06 0.1 0.4 1.2 0.6 0.7 0.9 May 2009 Deep 7 25 35 26.0 26.2 8809 0.37 0.5 1.5 2.3 0.1 1.7 1.6 May 2009 Deep 7 35 45 7.2 27.0 13709 0.24 0.1 0.8 1.6 0.4 1.0 1.2 May 2009 Deep 7 45 51 21.4 25.2 6016 0.09 0.1 0.4 1.2 0.6 0.7 0.9 May 2009 Deep 8 25 35 39.9 27.1 8809 0.19 0.5 1.5 2.3 0.1 1.7 1.6 May 2009 Deep 8 35 45 13.0 27.4 13709 0.17 0.1 0.8 1.6 0.4 1.0 1.2 May 2009 Deep 8 45 56.5 135.5 27.8 11438 0.06 0.2 0.3 1.1 0.6 0.6 0.9 May 2009 Deep 8 56.5 65 172.5 27.1 7721 0.07 0.3 0.1 0.8 0.7 0.5 0.7 May 2009 Deep 8 65 79 322.8 25.6 8185 0.06 0.4 0.1 0.5 0.8 0.3 0.6 May 2009 Deep 9 25 35 105.5 26.6 8809 0.08 0.5 1.5 2.3 0.1 1.7 1.6 May 2009 Deep 9 35 45 133.6 26.3 13709 0.05 0.1 0.8 1.6 0.4 1.0 1.2 May 2009 Deep 9 45 55 324.3 26.5 10027 0.08 0.1 0.3 1.1 0.6 0.7 0.9 May 2009 Deep 9 55 62.5 567.8 26.5 6813 0.09 0.3 0.1 0.8 0.7 0.5 0.8

PAGE 162

162 162 T able A 13 14 C of each R eco collar for all active layer (AL) categories and months. These are the mixture values used in the SIAR model. The R 2 are from the keeling plot regressions used t 13 C. Year Month AL Category 13 C 14 C R 2 Year Month AL Category 13 C 14 C R 2 2009 May Shallow 22.31 54.43 0.98 2008 June Shallow 22.89 61.47 0.98 2009 May Shallow 21.88 66.83 0.99 2008 June Shallow 22.45 62.40 0.97 200 9 May Shallow 25.65 68.81 0.97 2008 June Shallow 23.39 65.71 0.97 2009 May Shallow 24.32 70.23 0.99 2008 June Shallow 27.17 71.43 0.97 200 9 May Intermediate 23.60 61.72 1.00 2008 June Shallow 26.66 88.74 0.95 200 9 May Intermediate 21.61 62.91 0.95 2008 June Intermediate 22.98 56.70 1.00 200 9 May Deep 23.64 54.24 0.99 2008 June Intermediate 24.17 56.82 0.97 200 9 May Deep 24.15 55.14 1.00 2008 June Intermediate 24.13 58.65 0.99 2009 July Shallow 22.60 51.96 1.00 2008 June Deep 23.10 53.52 0.96 200 9 July Shallow 22.52 58.23 1.00 2008 June Deep 24.65 55.18 0.99 2009 July Shallow 23.12 59.71 0.99 2008 June Deep 24.09 56.54 0.98 200 9 July Shallow 22.94 66.85 0.99 2008 July Shallow 23.32 51.24 1.00 2009 July Shallow 22.24 72.00 1.00 2008 July Shallow 22.49 55.53 0.99 2009 July Intermediate 22.52 41.34 1.00 2008 July Shallow 22.23 61.15 1.00 200 9 July Intermediate 22.10 50.84 1.00 2008 July Shallow 22.39 66.02 1.00 2009 July Intermediate 22.24 57.64 0.98 2008 July Shallow 23.97 66.95 0.97 200 9 July Deep 22.35 46.37 1.00 2008 July Intermediate 21.87 43.61 0.99 2009 July Deep 24.16 47.61 1.00 2008 July Intermediate 23.17 48.26 1.00 200 9 July Deep 22.11 51.10 1.00 2008 July Intermediate 22.76 63.50 0.99 2009 September Shallow 23.83 39.30 0.98 2008 July Deep 22.36 46.98 1.00 200 9 September Shallow 23.06 55.34 0.99 2008 July Deep 23.37 48.97 1.00 200 9 September Shallow 23.31 58.19 0.96 2008 August Shallow 23.05 40.16 1.00 2009 September Shallow 24.28 64.32 0.98 2008 August Shallow 22.57 52.66 1.00 200 9 September Shallow 23.42 70.35 0.98 2008 August Shallow 23.40 52.70 0.99 2009 September Intermediate 23.78 47.78 1.00 2008 August Shallow 24.01 53.91 0.95 200 9 September Intermediate 27.88 55.14 0.96 2008 August Intermediate 23.15 45.30 1.00 200 9 September Intermediate 22.82 58.08 0.99 2008 August Intermediate 24.20 50.82 1.00 2009 September Deep 24.48 41.30 0.97 2008 August Intermediate 22.63 56.73 1.00

PAGE 163

163 163 Tab le A 2. Continued Year Month AL Category 13 C 14 C R 2 Year Month AL Category 13 C 14 C R 2 2009 September Deep 23.13 41.40 0.99 2008 August Deep 22.92 40.67 1.00 2009 September Deep 24.49 52.33 0.98 2008 August Deep 23.55 46.04 1.00 200 8 August Deep 22.84 49.96 1.00

PAGE 164

164 Table A 3 Mean (SE) R eco flux during the growing season for all AL categories in 2008 and 2009. GS stands for growing season. Data are averaged over different chamber bases but are from the same data set as Trucco et al. ( 2012). Respiration Flux (g C m 2 ) 2008 May June July Aug ust September GS Shallow AL 53.6 5 73.3 5 80.7 6 67.1 5 49.8 4 325 24 Intermediate AL 58.0 6 77.6 6 84.6 6 71.6 6 53.8 6 346 30 Deep AL 78.0 6 91.7 4 97.2 4 88.9 5 73.0 6 429 25 2009 Shallow AL 60.0 5 74.6 6 105.4 8 69.1 6 52.4 4 362 29 Intermediate AL 56.1 3 69.6 4 98.2 4 64.6 3 49.0 3 337 17 Deep AL 63.3 7 80.6 7 119 8 73.8 7 54.8 7 391 33 Table A 4 Sensitivity analysis using July 2009 data to quantify how changing the mean values of source isotopes affects model results. For this sensitivity analysis we changed 13 C shows t 13 C values that were not adjusted for differences between incubation and in situ temperatures (see Methods). OS SD and 14 14 C values that were min 14 13 C and Soil 13 13 C values that Change in percent mean contribution to R eco Aboveground Belowground Young Soil Old Soil Shallow AL Unadj. soil 13 C 10.0 3.0 5.1 8.0 OS 14 C 1.6 0.6 1.6 3.8 14 C 2.1 1.1 2.0 5.2 13 C 6.5 0.1 5.4 1.0 Soil 13 C 4.6 2.4 5.9 1.1 Intermediate AL Unadj. soil 13 C 7.8 3.2 2.7 8.3 OS 14 C 1.5 1.0 1.8 4.2 14 C 1.1 0.8 3.1 5.0 13 C 5.2 1.3 3.6 2.9 Soil 13 C 3.8 1.4 2.6 2.7 Deep AL Unadj. soil 13 C 6.4 1.6 4.2 3.8 OS 14 C 1.8 0.6 1.6 4.0 14 C 1.3 0.4 3.1 4.8 13 C 3.6 0.1 2.5 0.9 Soil 13 C 3.0 0.8 2.6 1.1

PAGE 165

165 APPENDIX B SUPPLEMENTAL TABLES FOR CHAPTER 4 Table B 13 14 C, flux rates, and estimated carbon pools and soil temperatures of individual soil sections used to calculate the depth 13 14 C of young soil and old soil. For shallow soil cores (0 25 cm) the block average % carbon and bulk density for each depth section were used to calculate the g carbo n per m 2 because they were not measured on the same cores. For the deep soil cores, % carbon and bulk density were measured on a subsample of the depth sections prior to incubation. Date Sampled Block Fence Treatment Depth Flux Carbon 13 C 14 C Soil Temperature (C) cm g C hr 1 g C m 2 Jul 2009 Sep 2009 Aug 2010 Aug 2011 June 2009 A control 0 5 0.25 1312 26.80 81.77 12.7 5.0 12.2 10.0 June 2009 A control 5 15 0.30 3988 26.08 122.19 6.3 3.0 7.2 5.8 June 2009 A control 15 25 0.27 10447 26.47 56.99 3.1 2.0 4.7 3.6 June 2009 A winter 0 5 0.21 1312 27.06 85.62 12.7 5.0 12.2 10.0 June 2009 A winter 5 15 0.35 3988 26.28 84.19 6.3 3.0 7.2 5.8 June 2009 A winter 15 25 0.35 10447 26.06 19.37 3.1 2.0 4.7 3.6 June 2009 B control 0 5 0.21 1184 26.76 118.33 12.7 5.0 12.2 10.0 June 2009 B control 5 15 0.23 4137 26.47 62.99 6.3 3.0 7.2 5.8 June 2009 B control 15 23 0.27 6268 25.98 11.34 3.1 2.0 4.7 3.6 June 2009 B winter 0 5 0.27 1184 26.95 107.92 12.7 5.0 12.2 10.0 June 2009 B winter 5 15 0.37 4137 26.14 72.29 6.3 3.0 7.2 5.8 June 2009 B winter 15 25 0.27 7835 26.17 0.81 3.1 2.0 4.7 3.6 June 2009 C control 0 5 0.23 1236 27.70 55.40 12.7 5.0 12.2 10.0 June 2009 C control 5 15 0.35 2431 25.63 100.69 6.3 3.0 7.2 5.8 June 2009 C control 15 25 0.22 4699 25.69 46.84 3.1 2.0 4.7 3.6 June 2009 C winter 0 5 0.09 1236 27.13 81.79 12.7 5.0 12.2 10.0 June 2009 C winter 5 15 0.61 2431 24.53 92.01 6.3 3.0 7.2 5.8 June 2009 C winter 15 25 0.41 4946 NA 62.98 3.1 2.0 4.7 3.6 Aug 2010 A annual 0 5 0.24 1312 25.87 94.24 12.7 5.0 12.2 10.0 Aug 2010 A annual 5 15 0.21 3988 25.63 57.71 6.3 3.0 7.2 5.8 Aug 2010 A annual 15 25 0.25 10447 25.35 18.87 3.1 2.0 4.7 3.6

PAGE 166

166 Aug 2010 A control 0 5 0.25 1312 27.34 81.87 12.7 5.0 12.2 10.0 Aug 2010 A control 5 15 0.21 3988 24.75 65.63 6.3 3.0 7.2 5.8 Aug 2010 A control 15 25 0.21 10447 25.83 18.70 3.1 2.0 4.7 3.6 Aug 2010 A summer 0 5 0.14 1312 26.70 120.90 12.7 5.0 12.2 10.0 Aug 2010 A summer 5 15 0.16 3988 25.44 52.57 6.3 3.0 7.2 5.8 Aug 2010 A summer 15 25 0.15 10447 25.79 13.45 3.1 2.0 4.7 3.6 Aug 2010 A winter 0 5 0.20 1312 27.09 96.51 12.7 5.0 12.2 10.0 Aug 2010 A winter 5 15 0.19 3988 26.63 91.47 6.3 3.0 7.2 5.8 Aug 2010 A winter 15 25 0.11 10447 26.21 23.04 3.1 2.0 4.7 3.6 Aug 2010 B annual 0 5 0.17 1184 25.59 131.74 12.7 5.0 12.2 10.0 Aug 2010 B annual 5 15 0.12 4137 23.95 56.69 6.3 3.0 7.2 5.8 Aug 2010 B annual 15 25 0.20 7835 25.62 12.03 3.1 2.0 4.7 3.6 Aug 2010 B control 0 5 0.14 1184 25.37 114.40 12.7 5.0 12.2 10.0 Aug 2010 B control 5 15 0.13 4137 25.70 74.18 6.3 3.0 7.2 5.8 Aug 2010 B control 15 25 0.15 7835 24.38 20.64 3.1 2.0 4.7 3.6 Aug 2010 B summer 0 5 0.20 1184 25.16 79.90 12.7 5.0 12.2 10.0 Aug 2010 B summer 5 15 0.07 4137 24.96 80.70 6.3 3.0 7.2 5.8 Aug 2010 B summer 15 25 0.12 7835 24.84 2.10 3.1 2.0 4.7 3.6 Aug 2010 B winter 0 5 0.16 1184 25.35 78.70 12.7 5.0 12.2 10.0 Aug 2010 B winter 5 15 0.13 4137 24.85 104.20 6.3 3.0 7.2 5.8 Aug 2010 B winter 15 25 0.25 7835 23.13 12.70 3.1 2.0 4.7 3.6 Aug 2010 C annual 0 5 0.14 1236 25.59 131.74 12.7 5.0 12.2 10.0 Aug 2010 C annual 5 15 0.12 2431 23.95 56.69 6.3 3.0 7.2 5.8 Aug 2010 C annual 15 25 0.07 4946 25.62 12.03 3.1 2.0 4.7 3.6 Aug 2010 C control 0 5 0.22 1236 25.37 114.40 12.7 5.0 12.2 10.0 Aug 2010 C control 5 15 0.25 2431 25.70 74.18 6.3 3.0 7.2 5.8 Aug 2010 C control 15 25 0.29 4946 24.38 20.64 3.1 2.0 4.7 3.6 Aug 2010 C summer 0 5 0.24 1236 25.16 79.90 12.7 5.0 12.2 10.0 Aug 2010 C summer 5 15 0.09 2431 24.96 80.70 6.3 3.0 7.2 5.8 Aug 2010 C summer 15 25 0.06 4946 24.84 2.10 3.1 2.0 4.7 3.6 Aug 2010 C winter 0 5 0.18 1236 25.35 78.70 12.7 5.0 12.2 10.0

PAGE 167

167 Table B 1. Continued Date Sampled Block Fence Treatment Depth Flux Carbon 13 C 14 C Soil Temperature (C) cm g C hr 1 g C m 2 Jul 2009 Sep 2009 Aug 2010 Aug 2011 Aug 2010 C winter 5 15 0.16 2431 24.85 104.20 6.3 3.0 7.2 5.8 Aug 2010 C winter 15 25 0.12 4946 23.13 12.70 3.1 2.0 4.7 3.6 May 2010 A 1 control 25 35 0.02 3222 24.31 244.56 1.0 1.3 3.1 2.2 May 2010 A 1 control 35 45 0.03 4794 23.29 143.36 0.2 0.9 2.1 1.4 May 2010 A 1 control 45 55 0.05 8203 24.92 309.05 1.2 0.6 1.4 0.7 May 2010 A 1 control 55 65 0.06 8003 23.59 314.93 2.0 0.3 0.7 0.2 May 2010 A 1 control 65 72 0.02 4671 23.79 326.20 2.7 0.1 0.2 0.2 May 2010 A 1 winter 25 32 0.10 7245 26.02 34.35 1.5 1.4 3.5 2.6 May 2010 A 1 winter 32 42 0.03 2735 26.60 198.99 0.3 1.1 2.5 1.8 May 2010 A 1 winter 42 52 0.05 5488 24.70 186.39 0.8 0.7 1.7 1.0 May 2010 A 1 winter 52 62 0.05 3187 17.88 345.56 1.7 0.4 1.0 0.4 May 2010 A 1 winter 62 72 0.04 1414 21.18 466.83 2.4 0.2 0.4 0.1 May 2010 A 2 control 25 35 0.05 13102 25.21 97.34 1.3 1.4 3.3 2.4 May 2010 A 2 control 35 45 0.06 18812 26.81 193.92 0.1 0.9 2.3 1.5 May 2010 A 2 control 45 55 0.11 15565 25.81 138.58 1.1 0.6 1.5 0.8 May 2010 A 2 control 55 65 0.06 12311 23.82 253.26 1.9 0.3 0.8 0.3 May 2010 A 2 control 65 72 0.04 10251 25.88 440.14 2.5 0.1 0.3 0.1 May 2010 A 2 winter 25 35 0.06 6919 25.61 96.46 1.3 1.4 3.3 2.4 May 2010 A 2 winter 35 45 0.03 4026 23.77 215.44 0.1 0.9 2.3 1.5 May 2010 A 2 winter 45 55 0.07 8480 23.82 242.42 1.1 0.6 1.5 0.8 May 2010 A 2 winter 55 65 0.05 7506 14.98 286.23 1.9 0.3 0.8 0.3 May 2010 A 2 winter 65 75 0.13 7277 21.59 335.36 2.6 0.1 0.3 0.2 May 2010 B 3 control 25 30.5 0.05 7259 26.36 62.45 1.6 1.5 3.6 2.6 May 2010 B 3 control 30.5 38 0.03 3825 24.23 38.47 0.7 1.2 2.8 2.0 May 2010 B 3 control 38 45 0.04 8497 26.68 152.88 0.2 0.9 2.1 1.4 May 2010 B 3 control 45 54.5 0.05 9382 23.07 190.72 1.0 0.6 1.5 0.9 May 2010 B 3 winter 25 35 0.07 6971 25.24 96.46 1.3 1.4 3.3 2.4

PAGE 168

168 Table B 1. Continued Date Sampled Block Fence Treatment Depth Flux Carbon 13 C 14 C Soil Temperature (C) cm g C hr 1 g C m 2 Jul 2009 Sep 2009 Aug 2010 Aug 2011 May 2010 B 3 winter 35 45 0.04 9227 25.73 194.47 0.1 0.9 2.3 1.5 May 2010 B 3 winter 45 55 0.05 11243 26.70 340.90 0.1 0.9 2.3 1.5 May 2010 B 3 winter 55 65 0.01 6465 23.47 339.74 1.9 0.3 0.8 0.3 May 2010 B 3 winter 65 75 0.01 8866 22.38 395.29 2.6 0.1 0.3 0.2 May 2010 B 4 control 25 36.5 0.09 11907 26.26 58.21 1.2 1.3 3.2 2.3 May 2010 B 4 control 36.5 45 0.02 7642 24.98 356.71 0.1 0.9 2.2 1.5 May 2010 B 4 control 45 55 0.04 11279 27.50 296.87 1.1 0.6 1.5 0.8 May 2010 B 4 control 55 65 0.02 9404 23.76 326.90 1.9 0.3 0.8 0.3 May 2010 B 4 control 65 75 0.01 5846 22.59 449.18 2.6 0.1 0.3 0.2 May 2010 B 4 winter 25 37 0.11 13987 27.31 99.15 1.1 1.3 3.2 2.3 May 2010 B 4 winter 37 45 0.04 7292 25.42 190.41 0.2 0.9 2.2 1.4 May 2010 B 4 winter 45 55 0.03 8038 23.60 313.53 1.1 0.6 1.5 0.8 May 2010 B 4 winter 55 65 0.02 7688 24.69 432.74 1.9 0.3 0.8 0.3 May 2010 B 4 winter 65 75 0.02 10953 24.30 352.09 2.6 0.1 0.3 0.2 May 2010 C 5 control 25 33.5 0.07 5850 25.91 25.08 1.4 1.4 3.4 2.5 May 2010 C 5 control 33.5 43.5 0.05 9712 23.64 167.81 0.1 1.0 2.4 1.6 May 2010 C 5 control 43.5 53.5 0.05 3834 21.70 237.30 0.9 0.6 1.6 0.9 May 2010 C 5 control 53.5 65.5 0.04 4447 19.08 433.10 1.8 0.4 0.9 0.4 May 2010 C 5 winter 25 35 0.14 10660 25.54 9.11 1.3 1.4 3.3 2.4 May 2010 C 5 winter 35 45 0.11 11766 23.59 85.73 0.1 0.9 2.3 1.5 May 2010 C 5 winter 45 55 0.04 3458 22.62 272.68 1.1 0.6 1.5 0.8 May 2010 C 5 winter 55 64.5 0.12 3366 NA 320.25 1.9 0.3 0.8 0.3 May 2010 C 6 control 25 33.5 0.08 9404 26.66 183.19 1.4 1.4 3.4 2.5 May 2010 C 6 control 33.5 43.5 0.02 8535 25.42 148.27 0.1 1.0 2.4 1.6 May 2010 C 6 control 43.5 49 0.04 5501 26.23 33.04 0.7 0.7 1.7 1.1 May 2010 C 6 winter 25 32.5 0.12 8109 25.74 11.54 1.5 1.4 3.4 2.5 May 2010 C 6 winter 32.5 42.5 0.02 12418 22.22 120.86 0.2 1.0 2.5 1.7

PAGE 169

169 Table B 1. Continued Date Sampled Block Fence Treatment Depth Flux Carbon 13 C 14 C Soil Temperature (C) cm g C hr 1 g C m 2 Jul 2009 Sep 2009 Aug 2010 Aug 2011 May 2010 C 6 winter 42.5 52.5 0.04 6529 24.79 247.28 0.8 0.7 1.6 1.0 May 2010 C 6 winter 52.5 62.5 0.03 6603 25.25 343.21 1.7 0.4 1.0 0.4 May 2010 C 6 winter 62.5 72.5 0.03 3476 18.36 461.84 2.5 0.2 0.4 0.1

PAGE 170

170 T able B 2 Soil environmental variables by plot during each sampling period: dept to thaw, depth to the water table (WT), volumetric water content in the top 20 cm (VWC), and temperature at soil depths of 5 cm, 20 cm, and 40 cm. Date Plot Treatment Thaw WT VWC 1 5 cm T 1 20 cm T 1 VWC 2 5 cm T 2 20 cm T 2 c m cm % C C % C C Jul 2009 1 2 control 24.5 16.0 32.6 9.33 1.93 32.1 8.42 1.84 Jul 2009 4 2 control 26.9 30.5 33.5 7.77 2.22 33.3 6.52 1.80 Jul 2009 5 2 control 25.3 29.0 45.1 9.97 2.60 45.6 6.43 1.49 Jul 2009 1 1 summer 29.7 22.5 40.1 8.62 1.23 41.4 7.53 1.33 Jul 2009 4 1 summer 23.2 23.0 51.4 5.55 0.70 53.5 4.57 0.51 Jul 2009 5 1 summer 29.2 16.5 47.5 8.65 2.58 48.7 6.27 1.99 Jul 2009 2 6 winter 40.7 22.5 56.8 8.92 1.48 57.4 8.44 1.34 Jul 2009 4 6 winter 36.4 27.0 35.6 9.16 0.98 35.3 7.79 0.89 Jul 2009 6 6 winter 38.2 19.0 44.1 8.98 3.51 46.0 6.82 3.01 Jul 2009 1 5 annual 29.0 34.3 46.1 9.16 2.24 46.7 8.24 2.26 Jul 2009 4 5 annual 33.0 28.5 39.5 9.22 1.07 39.9 8.07 0.87 Jul 2009 6 5 annual 36.9 20.5 46.5 7.94 2.26 46.9 5.80 0.58 Sep 2009 1 2 control 46.2 18.0 43.6 6.04 2.66 44.3 6.13 2.72 Sep 2009 4 2 control 56.0 29.0 44.7 4.81 3.03 44.3 5.69 2.90 Sep 2009 5 2 control 48.5 22.0 45.2 5.77 2.48 45.1 6.80 2.74 Sep 2009 1 1 summer 48.6 23.0 41.4 6.08 2.80 42.8 6.74 3.05 Sep 2009 4 1 summer 51.3 22.0 52.2 5.65 3.68 51.8 6.12 3.50 Sep 2009 5 1 summer 59.0 13.5 48.7 5.78 2.93 48.6 6.48 3.19 Sep 2009 2 6 winter 63.1 18.0 76.7 10.62 2.96 74.6 7.69 2.68 Sep 2009 4 6 winter 63.7 30.0 41.3 5.05 2.34 41.5 6.20 2.23 Sep 2009 6 6 winter 64.8 14.0 45.6 5.41 3.54 45.6 6.31 3.91 Sep 2009 2 5 annual 54.3 12.5 51.3 9.94 2.59 50.7 8.14 2.26 Sep 2009 4 5 annual 52.7 26.0 40.8 6.29 2.79 39.9 6.71 2.67 Sep 2009 6 5 annual 55.0 13.0 47.3 5.28 2.77 47.4 5.26 3.02 Aug 2010 1 2 control 47.1 17.0 50.0 9.30 3.82 49.8 9.44 4.25 Aug 2010 2 2 control 59.6 25.0 55.8 9.57 3.61 55.6 10.00 3.97 Aug 2010 3 2 control 52.8 31.0 56.6 7.75 3.24 56.4 7.65 3.48

PAGE 171

171 Table B 2 Continued Date Plot Treatment Thaw WT VWC 1 5 cm T 1 20 cm T 1 VWC 2 5 cm T 2 20 cm T 2 c m cm % C C % C C Aug 2010 4 2 control 57.7 32.0 49.9 9.36 4.69 50.4 9.05 4.87 Aug 2010 5 2 control 53.2 22.0 50.5 11.00 5.70 51.5 9.89 4.51 Aug 2010 6 2 control 56.4 13.0 62.7 7.21 5.17 59.0 11.59 6.41 Aug 2010 1 1 summer 52.2 20.0 47.6 10.56 5.01 48.0 11.26 5.02 Aug 2010 2 1 summer 54.1 33.0 53.5 11.33 3.88 53.1 12.36 4.11 Aug 2010 3 1 summer 58.3 25.0 60.4 8.67 5.45 61.0 8.58 5.74 Aug 2010 4 1 summer 52.6 29.0 58.6 10.09 5.29 58.4 10.29 5.89 Aug 2010 5 1 summer 64.0 16.0 53.9 10.32 6.56 56.8 8.86 5.37 Aug 2010 6 1 summer 53.6 14.0 51.2 7.41 4.14 50.1 9.27 4.96 Aug 2010 1 6 winter 70.9 11.0 61.8 8.50 4.49 62.0 8.85 4.54 Aug 2010 2 6 winter 68.4 19.0 60.8 10.29 6.46 62.7 10.23 6.37 Aug 2010 3 6 winter 60.0 20.0 56.8 9.93 5.92 56.7 9.82 6.17 Aug 2010 4 6 winter 65.8 31.0 NA 10.57 3.51 NA 9.97 3.67 Aug 2010 5 6 winter 60.6 12.0 53.9 11.76 6.98 55.7 10.27 5.79 Aug 2010 6 6 winter 66.8 10.0 57.3 6.98 5.55 55.9 10.55 6.60 Aug 2010 1 5 annual 62.9 35.0 58.7 8.34 3.80 58.6 9.02 4.26 Aug 2010 2 5 annual 59.8 15.0 58.2 8.87 4.62 58.8 9.61 4.94 Aug 2010 3 5 annual 57.7 16.0 64.4 11.05 6.84 64.3 10.46 6.92 Aug 2010 4 5 annual 54.7 29.0 70.9 9.36 6.19 70.7 8.98 6.34 Aug 2010 5 5 annual 59.6 29.0 56.2 9.79 5.36 57.8 8.91 4.56 Aug 2010 6 5 annual 59.8 7.0 65.2 6.20 4.34 63.1 8.32 5.07 Aug 2011 1 2 control 45.1 38.0 45.9 6.68 2.25 45.6 7.09 2.66 Aug 2011 2 2 control 49.6 41.0 51.0 7.24 2.20 50.6 7.85 2.54 Aug 2011 3 2 control 40.0 39.0 55.9 6.27 2.08 56.0 5.92 2.05 Aug 2011 4 2 control 50.1 38.0 48.1 7.26 3.31 48.2 6.68 3.11 Aug 2011 5 2 control 40.7 34.0 51.2 7.46 2.89 51.3 7.10 2.78 Aug 2011 6 2 control 47.6 33.0 52.3 9.01 4.01 53.2 7.47 3.38 Aug 2011 2 1 summer 40.0 35.0 48.3 7.49 1.75 48.0 7.49 1.93

PAGE 172

172 Table B 2 Continued Date Plot Treatment Thaw WT VWC 1 5 cm T 1 20 cm T 1 VWC 2 5 cm T 2 20 cm T 2 cm cm % C C % C C Aug 2011 3 1 summer 45.7 31.0 54.9 5.93 3.19 54.8 6.33 3.33 Aug 2011 4 1 summer 45.1 39.0 46.6 6.98 4.62 46.9 6.93 3.14 Aug 2011 5 1 summer 52.0 33.0 51.6 7.20 2.90 52.2 5.80 2.71 Aug 2011 6 1 summer 50.9 29.0 47.8 9.31 3.59 48.6 7.04 3.00 Aug 2011 1 6 winter 66.1 31.0 58.7 6.18 2.33 58.5 6.23 2.54 Aug 2011 2 6 winter 65.3 36.0 57.0 7.62 3.38 56.8 7.89 3.69 Aug 2011 4 6 winter 63.4 43.0 52.2 7.19 3.29 52.1 7.31 3.20 Aug 2011 5 6 winter 61.9 25.0 47.1 10.13 4.32 48.2 7.87 3.90 Aug 2011 6 6 winter 61.5 27.0 55.8 9.03 4.07 56.3 7.42 3.89 Aug 2011 1 5 annual 63.0 37.0 55.0 5.73 2.77 54.5 6.62 2.93 Aug 2011 2 5 annual 59.8 35.0 52.6 6.50 3.06 52.3 7.15 3.20 Aug 2011 3 5 annual 56.3 31.0 56.4 6.67 4.25 56.2 7.16 4.43 Aug 2011 4 5 annual 51.4 44.0 65.6 7.86 3.27 65.5 7.13 4.70 Aug 2011 5 5 annual 52.2 40.0 52.6 7.86 2.60 53.1 6.33 2.53 Aug 2011 6 5 annual 57.7 25.0 55.9 8.96 3.30 56.4 7.07 3.05 1 These soil variables were averaged over the day R eco 14 C was sampled at each plot. 2 These soil variables were averaged for the whole time period over which R eco 14 C and 13 C were sampled.

PAGE 173

173 Table B 3 13 14 C, average R eco flux, and mean estimates of aboveground autotrophic (AG), belowground autotrophic (BG), young soil (YS), and old soil (OS) proportional contributions to R eco Note the R eco flux for Jul 2009 was actually calculated from June 15 to July 15, 2009 since samp ling occurred from June 26 through July 1. Date Treatment Plot 13 C 14 C R eco AG BG YS OS g C m 2 hour 1 Jul 2009 control 1 2 24.26 44.08 0.13 0.22 0.38 0.27 0.13 Jul 2009 control 4 2 23.88 39.57 0.18 0.27 0.27 0.28 0.18 Jul 2009 control 5 2 20.49 41.30 0.32 0.28 0.03 0.37 0.32 Jul 2009 summer 1 1 26.23 18.63 0.23 0.05 0.68 0.04 0.23 Jul 2009 summer 4 1 23.07 39.19 0.21 0.28 0.18 0.33 0.21 Jul 2009 summer 5 1 21.83 24.96 0.34 0.25 0.07 0.34 0.34 Jul 2009 winter 2 6 24.58 37.80 0.16 0.19 0.44 0.21 0.16 Jul 2009 winter 4 6 25.33 39.50 0.11 0.31 0.46 0.12 0.11 Jul 2009 winter 6 6 24.40 33.45 0.18 0.24 0.35 0.23 0.18 Jul 2009 annual 1 5 20.83 44.88 0.06 0.71 0.09 0.14 0.06 Jul 2009 annual 4 5 22.15 48.97 0.17 0.31 0.07 0.45 0.17 Jul 2009 annual 6 5 23.20 44.83 0.18 0.25 0.20 0.37 0.18 Sep 2009 control 1 2 23.28 62.16 0.04 0.24 0.26 0.46 0.04 Sep 2009 control 4 2 24.30 37.76 0.14 0.28 0.38 0.20 0.14 Sep 2009 control 5 2 21.50 55.08 0.13 0.30 0.05 0.52 0.13 Sep 2009 summer 1 1 22.85 67.20 0.04 0.23 0.17 0.56 0.04 Sep 2009 summer 4 1 25.36 35.96 0.11 0.31 0.50 0.08 0.11 Sep 2009 summer 5 1 22.45 60.66 0.09 0.30 0.12 0.49 0.09 Sep 2009 winter 2 6 25.12 56.66 0.03 0.13 0.59 0.24 0.03 Sep 2009 winter 4 6 25.16 58.32 0.03 0.23 0.46 0.28 0.03 Sep 2009 winter 6 6 22.76 59.80 0.09 0.29 0.16 0.46 0.09 Sep 2009 annual 2 5 24.31 36.07 0.15 0.22 0.45 0.18 0.15 Sep 2009 annual 4 5 23.97 32.70 0.18 0.26 0.34 0.22 0.18 Sep 2009 annual 6 5 22.82 48.51 0.14 0.26 0.19 0.41 0.14 Aug 2010 control 1 2 23.05 15.75 0.48 0.18 0.20 0.14 0.48 Aug 2010 control 2 2 22.93 52.28 0.66 0.11 0.13 0.09 0.66 Aug 2010 control 3 2 22.74 5.35 0.38 0.25 0.16 0.21 0.38 Aug 2010 control 4 2 24.66 28.82 0.50 0.07 0.37 0.07 0.50 Aug 2010 control 5 2 26.20 68.52 0.73 0.10 0.14 0.03 0.73 Aug 2010 control 6 2 25.36 326.10 0.71 0.28 0.01 0.01 0.71 Aug 2010 summer 1 1 24.08 31.50 0.21 0.20 0.34 0.24 0.21 Aug 2010 summer 2 1 23.07 13.45 0.34 0.24 0.19 0.23 0.34 Aug 2010 summer 3 1 23.28 38.90 0.19 0.24 0.21 0.35 0.19 Aug 2010 summer 4 1 22.93 41.75 0.61 0.13 0.16 0.10 0.61 Aug 2010 summer 5 1 26.88 9.89 0.25 0.11 0.61 0.03 0.25 Aug 2010 summer 6 1 24.16 22.86 0.38 0.25 0.28 0.09 0.38 Aug 2010 winter 1 6 23.72 41.03 0.59 0.09 0.22 0.09 0.59

PAGE 174

174 Table B 3 Continued Date Treatment Plot 13 C 14 C R eco AG BG YS OS g C m 2 hour 1 Aug 2010 winter 2 6 23.21 30.64 0.25 0.27 0.20 0.29 0.25 Aug 2010 winter 3 6 25.93 22.61 0.33 0.25 0.39 0.03 0.33 Aug 2010 winter 4 6 26.30 18.82 0.40 0.03 0.54 0.03 0.40 Aug 2010 winter 5 6 25.07 5.07 0.37 0.14 0.42 0.07 0.37 Aug 2010 winter 6 6 23.74 7.72 0.35 0.24 0.26 0.15 0.35 Aug 2010 annual 1 5 24.12 28.43 0.23 0.19 0.35 0.23 0.23 Aug 2010 annual 2 5 23.47 30.30 0.24 0.26 0.23 0.28 0.24 Aug 2010 annual 3 5 25.27 15.93 0.21 0.27 0.44 0.08 0.21 Aug 2010 annual 4 5 24.22 29.36 0.22 0.19 0.36 0.23 0.22 Aug 2010 annual 5 5 23.56 20.24 0.29 0.24 0.24 0.23 0.29 Aug 2010 annual 6 5 24.29 23.88 0.22 0.23 0.32 0.22 0.22 Aug 2011 control 1 2 23.75 47.84 0.09 0.26 0.30 0.35 0.09 Aug 2011 control 2 2 21.35 47.34 0.16 0.32 0.03 0.49 0.16 Aug 2011 control 3 2 21.80 45.80 0.20 0.28 0.05 0.47 0.20 Aug 2011 control 4 2 24.30 45.69 0.09 0.22 0.37 0.32 0.09 Aug 2011 control 5 2 21.48 46.59 0.17 0.31 0.03 0.49 0.17 Aug 2011 control 6 2 24.65 39.62 0.12 0.26 0.37 0.26 0.12 Aug 2011 summer 2 1 24.11 30.37 0.19 0.21 0.35 0.25 0.19 Aug 2011 summer 3 1 23.96 50.60 0.08 0.27 0.26 0.39 0.08 Aug 2011 summer 4 1 21.51 28.94 0.31 0.26 0.05 0.38 0.31 Aug 2011 summer 5 1 25.01 57.59 0.04 0.13 0.44 0.40 0.04 Aug 2011 summer 6 1 24.86 49.63 0.06 0.28 0.35 0.32 0.06 Aug 2011 winter 1 6 23.98 48.06 0.08 0.24 0.32 0.35 0.08 Aug 2011 winter 2 6 23.62 48.14 0.10 0.28 0.22 0.40 0.10 Aug 2011 winter 4 6 24.50 56.28 0.06 0.26 0.32 0.36 0.06 Aug 2011 winter 5 6 23.03 31.07 0.21 0.29 0.22 0.28 0.21 Aug 2011 winter 6 6 25.47 52.64 0.04 0.25 0.41 0.30 0.04 Aug 2011 annual 1 5 24.27 46.63 0.09 0.22 0.36 0.33 0.09 Aug 2011 annual 2 5 23.64 48.55 0.10 0.28 0.22 0.40 0.10 Aug 2011 annual 3 5 23.68 46.73 0.13 0.26 0.24 0.37 0.13 Aug 2011 annual 4 5 23.85 45.21 0.10 0.26 0.31 0.33 0.10 Aug 2011 annual 5 5 22.17 38.98 0.22 0.30 0.07 0.41 0.22 Aug 2011 annual 6 5 25.55 47.83 0.05 0.28 0.47 0.20 0.05

PAGE 175

175 APPENDIX C SUPPLEMENTAL T ABLES F OR C HAPTER 5 Table C 1. Mean aboveground net primary production (ANPP; standard error) for the thaw gradient in 2009 and CiPEHR in 2011 by species and tissue type. These are the site or treatment means of the data used to calculate plant community weighted decomposition consta nts (see Methods in Chapter 5 ). Year Site/Treatment Species Leaf ANPP (g biomass m 2 y 1 ) Stem ANPP (g biomass m 2 y 1 ) 2009 Extensive AND 1 3.40 0.56 1.70 0.28 2009 Extensive BN 23.41 8.77 6.14 2.30 2009 Extensive CX 12.62 1.73 4.47 0.61 2009 Extensive EN 10.58 4.72 3.17 1.42 2009 Extensive EV 60.65 38.27 44.21 27.89 2009 Extensive LED 28.88 5.79 14.44 2.89 2009 Extensive OXY 6.06 3.61 1.40 0.83 2009 Extensive RC 31.20 4.66 2.90 0.43 2009 Extensive VUG 43.80 14.85 42.46 14.40 2009 Extensive VVI 18.25 5.52 4.21 1.28 2009 Minimal AND 3.58 1.02 1.79 0.51 2009 Minimal BN 19.41 6.49 5.09 1.70 2009 Minimal CX 16.74 3.11 5.92 1.10 2009 Minimal EN 5.18 2.47 1.55 0.74 2009 Minimal EV 154.63 33.45 112.72 24.38 2009 Minimal LED 16.92 2.19 8.46 1.10 2009 Minimal OXY 2.61 0.16 0.61 0.04 2009 Minimal RC 13.58 2.13 1.26 0.20 2009 Minimal VUG 21.35 1.28 20.70 1.25 2009 Minimal VVI 8.79 2.26 2.03 0.52 2009 Moderate AND 2.56 0.00 1.28 0.00 2009 Moderate BN 8.09 1.17 2.12 0.31 2009 Moderate CX 11.59 1.43 4.10 0.51 2009 Moderate EN 8.88 3.40 2.66 1.02 2009 Moderate EV 165.41 22.62 120.57 16.49 2009 Moderate LED 23.83 3.80 11.92 1.90 2009 Moderate OXY 3.01 0.26 0.70 0.06 2009 Moderate RC 19.11 2.38 1.77 0.22 2009 Moderate VUG 19.86 4.26 19.25 4.14 2009 Moderate VVI 12.48 3.42 2.88 0.79 2011 Annual AND 3.32 0.12 1.10 0.04 2011 Annual BN 9.42 3.48 2.47 0.91 2011 Annual CX 21.62 3.87 7.65 1.37 2011 Annual EN 7.30 2.19 2.19 0.66 2011 Annual EV 133.42 37.73 97.25 27.50 2011 Annual LED 16.85 1.85 8.42 0.92 2011 Annual OXY 2.78 0.18 0.64 0.04 2011 Annual RC 14.72 1.33 1.37 0.12 2011 Annual VUG 20.65 1.57 20.01 1.53 2011 Annual VVI 13.58 2.05 3.14 0.47 2011 Control AND 3.56 0.28 1.17 0.09 2011 Control BN 8.50 2.77 2.23 0.73 2011 Control CX 11.49 1.69 4.06 0.60

PAGE 176

176 Table C 1. Continued Year Site/Treatment Species Leaf ANPP (g biomass m 2 y 1 ) Stem ANPP (g biomass m 2 y 1 ) 2011 Control EN 5.83 1.87 1.75 0.56 2011 Control EV 112.80 18.08 82.22 13.18 2011 Control LED 15.50 1.71 7.75 0.85 2011 Control OXY 4.25 1.39 0.98 0.32 2011 Control RC 12.18 1.22 1.13 0.11 2011 Control VUG 15.67 1.49 15.20 1.45 2011 Control VVI 9.49 1.55 2.19 0.36 2011 Summer AND 3.63 0.30 1.20 0.10 2011 Summer BN 10.87 2.64 2.85 0.69 2011 Summer CX 14.34 2.69 5.07 0.95 2011 Summer EN 3.59 0.65 1.07 0.19 2011 Summer EV 121.26 33.11 88.39 24.13 2011 Summer LED 13.84 1.70 6.92 0.85 2011 Summer OXY 3.51 0.65 0.81 0.15 2011 Summer RC 12.87 0.96 1.19 0.09 2011 Summer VUG 18.29 2.35 17.73 2.28 2011 Summer VVI 12.03 1.97 2.78 0.46 2011 Winter AND 3.74 0.25 1.23 0.08 2011 Winter BN 6.85 2.39 1.80 0.63 2011 Winter CX 15.39 2.15 5.44 0.76 2011 Winter EN 8.71 2.19 2.61 0.66 2011 Winter EV 183.60 34.69 133.83 25.29 2011 Winter LED 12.99 1.40 6.49 0.70 2011 Winter OXY 2.53 0.08 0.59 0.02 2011 Winter RC 12.76 1.17 1.18 0.11 2011 Winter VUG 20.81 2.07 20.18 2.01 2011 Winter VVI 11.45 1.73 2.64 0.40 1 AND ( Andromeda polifolia ), BN ( Betula nana ), CX ( Carex bigelowii ), EN ( Empetrum nigr um ), EV ( Eriophorum vaginatum ), LED ( Rhododendron subarcticum ), OXY ( Oxycoccus microcarpus ), RC ( Rubus chamaemorus ). VUG ( Vaccinium uliginosum ), VVI ( Vaccinium vitis idaea )

PAGE 177

177 Table C 2. Annual mass loss of the common substrate at two depths in the permafrost thaw gradient. These data were used in the mixed model multiple regression (see Methods in Chapter 5 ).Environmental data included in the regression were the growing season (GS) average soil temperature at 10 cm per site, the average water table (WT) per s ite, and the active layer depth (ALD; by site for WGW locations and by location for the rest), as well as the total growing season precipitation (P ) The growing season was defined as May through September. Year Site Location Bag ID Depth (cm) Mass Loss (%) GS Soil T (C) P (mm) WT (cm) ALD (cm) 2004 2005 Minimal 13 13 0 10 3.39 7.28 145 20.2 60.5 2004 2005 Minimal 14 14 0 10 10.70 7.28 145 20.2 58.8 2004 2005 Minimal 15 15 0 10 4.70 7.28 145 20.2 59.5 2004 2005 Minimal 16 16 0 10 20.10 7.28 145 20.2 66.3 2004 2005 Minimal 17 17 0 10 14.23 7.28 145 20.2 65.8 2004 2005 Minimal 18 18 0 10 6.79 7.28 145 20.2 79.8 2004 2005 Minimal WGW11 141 0 10 7.96 7.28 145 20.2 68.6 2004 2005 Minimal WGW12 146 0 10 11.10 7.28 145 20.2 68.6 2004 2005 Minimal WGW14 154 0 10 3.79 7.28 145 20.2 68.6 2004 2005 Minimal WGW15 161 0 10 7.44 7.28 145 20.2 68.6 2004 2005 Moderate 7 7 0 10 12.66 8.39 145 18.4 72.1 2004 2005 Moderate 8 8 0 10 6.27 8.39 145 18.4 57.4 2004 2005 Moderate 9 9 0 10 3.79 8.39 145 18.4 61.9 2004 2005 Moderate 10 10 0 10 3.79 8.39 145 18.4 66.3 2004 2005 Moderate 11 11 0 10 20.50 8.39 145 18.4 66.6 2004 2005 Moderate 12 12 0 10 3.66 8.39 145 18.4 82.1 2004 2005 Moderate WGW10 134 0 10 12.01 8.39 145 18.4 77.9 2004 2005 Moderate WGW6 190 0 10 8.88 8.39 145 18.4 77.9 2004 2005 Moderate WGW7 194 0 10 10.18 8.39 145 18.4 77.9 2004 2005 Moderate WGW9 204 0 10 4.05 8.39 145 18.4 77.9 2004 2005 Extensive 1 1 0 10 12.01 6.67 145 15.6 59.5 2004 2005 Extensive 2 2 0 10 14.23 6.67 145 15.6 60.8 2004 2005 Extensive 3 3 0 10 10.57 6.67 145 15.6 57.5 2004 2005 Extensive 4 4 0 10 38.25 6.67 145 15.6 87.5 2004 2005 Extensive 5 5 0 10 18.67 6.67 145 15.6 61.8 2004 2005 Extensive 6 6 0 10 4.83 6.67 145 15.6 110.8 2004 2005 Extensive WGW1 127 0 10 9.27 6.67 145 15.6 77.8 2004 2005 Extensive WGW2 169 0 10 25.46 6.67 145 15.6 77.8 2004 2005 Extensive WGW3 176 0 10 3.66 6.67 145 15.6 77.8 2004 2005 Extensive WGW5 185 0 10 13.32 6.67 145 15.6 77.8 2005 2006 Minimal 13 31 0 10 4.96 6.84 227 22.4 66.3 2005 2006 Minimal 14 32 0 10 25.98 6.84 227 22.4 54.3 2005 2006 Minimal 15 33 0 10 28.20 6.84 227 22.4 53.4 2005 2006 Minimal 16 34 0 10 46.48 6.84 227 22.4 60.3 2005 2006 Minimal 17 35 0 10 25.98 6.84 227 22.4 55.6 2005 2006 Minimal 18 36 0 10 38.38 6.84 227 22.4 56.8 2005 2006 Minimal WGW11 144 0 10 56.92 6.84 227 22.4 62.7 2005 2006 Minimal WGW12 149 0 10 49.35 6.84 227 22.4 62.7 2005 2006 Minimal WGW13 151 0 10 27.68 6.84 227 22.4 62.7 2005 2006 Minimal WGW14 157 0 10 28.46 6.84 227 22.4 62.7 2005 2006 Moderate 7 25 0 10 33.94 5.55 227 19.8 67.5 2005 2006 Moderate 8 26 0 10 49.35 5.55 227 19.8 50.9

PAGE 178

178 Table C 2. Continued Year Site Location Bag ID Depth (cm) Mass Loss (%) GS Soil T (C) P (mm) WT (cm) ALD (cm) 2005 2006 Moderate 9 27 0 10 45.95 5.55 227 19.8 60.2 2005 2006 Moderate 10 28 0 10 24.28 5.55 227 19.8 54.5 2005 2006 Moderate 11 29 0 10 39.43 5.55 227 19.8 64.8 2005 2006 Moderate 12 30 0 10 3.92 5.55 227 19.8 68.3 2005 2006 Moderate WGW10 137 0 10 5.22 5.55 227 19.8 65.8 2005 2006 Moderate WGW7 197 0 10 45.56 5.55 227 19.8 65.8 2005 2006 Moderate WGW8 201 0 10 37.08 5.55 227 19.8 65.8 2005 2006 Moderate WGW9 207 0 10 47.78 5.55 227 19.8 65.8 2005 2006 Extensive 1 19 0 10 42.04 6.19 227 19.3 56.1 2005 2006 Extensive 2 20 0 10 35.38 6.19 227 19.3 59.7 2005 2006 Extensive 3 21 0 10 34.99 6.19 227 19.3 44.4 2005 2006 Extensive 4 22 0 10 59.92 6.19 227 19.3 84.7 2005 2006 Extensive 5 23 0 10 47.26 6.19 227 19.3 50.8 2005 2006 Extensive 6 24 0 10 38.38 6.19 227 19.3 106.1 2005 2006 Extensive WGW1 130 0 10 34.20 6.19 227 19.3 75.2 2005 2006 Extensive WGW2 172 0 10 53.66 6.19 227 19.3 75.2 2005 2006 Extensive WGW3 179 0 10 29.24 6.19 227 19.3 75.2 2005 2006 Extensive WGW5 188 0 10 41.78 6.19 227 19.3 75.2 2006 2007 Minimal 13 49 0 10 11.39 7.14 331 18.8 73.1 2006 2007 Minimal 14 50 0 10 28.41 7.14 331 18.8 67.5 2006 2007 Minimal 15 51 0 10 41.01 7.14 331 18.8 60.0 2006 2007 Minimal 16 52 0 10 46.86 7.14 331 18.8 67.4 2006 2007 Minimal 17 53 0 10 45.24 7.14 331 18.8 60.4 2006 2007 Minimal 18 54 0 10 27.02 7.14 331 18.8 62.0 2006 2007 Minimal WGW13 152 0 10 55.04 7.14 331 18.8 70.8 2006 2007 Minimal WGW14 159 0 10 49.80 7.14 331 18.8 70.8 2006 2007 Minimal WGW16 165 0 10 42.07 7.14 331 18.8 70.8 2006 2007 Minimal WGW17 167 0 10 12.53 7.14 331 18.8 70.8 2006 2007 Moderate 7 43 0 10 30.25 7.16 331 16.1 71.5 2006 2007 Moderate 8 44 0 10 51.13 7.16 331 16.1 58.3 2006 2007 Moderate 9 45 0 10 59.79 7.16 331 16.1 69.1 2006 2007 Moderate 10 46 0 10 27.28 7.16 331 16.1 61.0 2006 2007 Moderate 11 47 0 10 29.34 7.16 331 16.1 67.9 2006 2007 Moderate 12 48 0 10 25.47 7.16 331 16.1 79.8 2006 2007 Moderate WGW10 139 0 10 31.54 7.16 331 16.1 71.4 2006 2007 Moderate WGW7 199 0 10 16.49 7.16 331 16.1 71.4 2006 2007 Moderate WGW8 202 0 10 45.56 7.16 331 16.1 71.4 2006 2007 Moderate WGW9 209 0 10 74.73 7.16 331 16.1 71.4 2006 2007 Extensive 1 37 0 10 70.61 7.68 331 14.3 61.5 2006 2007 Extensive 2 38 0 10 40.47 7.68 331 14.3 65.3 2006 2007 Extensive 3 39 0 10 37.89 7.68 331 14.3 53.8 2006 2007 Extensive 4 40 0 10 86.12 7.68 331 14.3 83.4 2006 2007 Extensive 5 41 0 10 39.91 7.68 331 14.3 55.8 2006 2007 Extensive 6 42 0 10 38.97 7.68 331 14.3 106.6 2006 2007 Extensive WGW1 132 0 10 43.47 7.68 331 14.3 79.1 2006 2007 Extensive WGW2 174 0 10 35.42 7.68 331 14.3 79.1 2006 2007 Extensive WGW3 181 0 10 36.85 7.68 331 14.3 79.1 2006 2007 Extensive WGW4 183 0 10 41.54 7.68 331 14.3 79.1 2007 2008 Minimal 13 67 0 10 13.71 6.00 346 15.2 76.4

PAGE 179

179 Table C 2. Continued Year Site Location Bag ID Depth (cm) Mass Loss (%) GS Soil T (C) P (mm) WT (cm) ALD (cm) 2007 2008 Minimal 14 68 0 10 18.02 6.00 346 15.2 68.0 2007 2008 Minimal 15 69 0 10 18.02 6.00 346 15.2 57.8 2007 2008 Minimal 16 70 0 10 53.39 6.00 346 15.2 69.3 2007 2008 Minimal 17 71 0 10 37.19 6.00 346 15.2 59.0 2007 2008 Minimal 18 72 0 10 22.70 6.00 346 15.2 57.8 2007 2008 Minimal WGW13 153 0 10 15.91 6.00 346 15.2 81.2 2007 2008 Minimal WGW14 160 0 10 32.21 6.00 346 15.2 81.2 2007 2008 Minimal WGW16 166 0 10 44.57 6.00 346 15.2 81.2 2007 2008 Minimal WGW18 168 0 10 26.33 6.00 346 15.2 81.2 2007 2008 Moderate 7 61 0 10 18.02 4.46 346 10.9 69.6 2007 2008 Moderate 8 62 0 10 35.64 4.46 346 10.9 53.6 2007 2008 Moderate 9 63 0 10 24.15 4.46 346 10.9 70.0 2007 2008 Moderate 10 64 0 10 17.10 4.46 346 10.9 58.0 2007 2008 Moderate 11 65 0 10 23.37 4.46 346 10.9 65.3 2007 2008 Moderate 12 66 0 10 17.36 4.46 346 10.9 77.2 2007 2008 Moderate WGW10 140 0 10 23.63 4.46 346 10.9 80.3 2007 2008 Moderate WGW7 200 0 10 8.22 4.46 346 10.9 80.3 2007 2008 Moderate WGW8 203 0 10 47.26 4.46 346 10.9 80.3 2007 2008 Moderate WGW9 210 0 10 56.14 4.46 346 10.9 80.3 2007 2008 Extensive 1 55 0 10 40.08 3.08 346 10.8 57.6 2007 2008 Extensive 2 56 0 10 39.43 3.08 346 10.8 60.9 2007 2008 Extensive 3 57 0 10 21.28 3.08 346 10.8 52.5 2007 2008 Extensive 4 58 0 10 31.46 3.08 346 10.8 89.6 2007 2008 Extensive 5 59 0 10 38.77 3.08 346 10.8 55.3 2007 2008 Extensive 6 60 0 10 15.80 3.08 346 10.8 122.3 2007 2008 Extensive WGW1 133 0 10 31.33 3.08 346 10.8 84.5 2007 2008 Extensive WGW2 175 0 10 28.33 3.08 346 10.8 84.5 2007 2008 Extensive WGW3 182 0 10 31.07 3.08 346 10.8 84.5 2007 2008 Extensive WGW4 184 0 10 14.62 3.08 346 10.8 84.5 2008 2009 Minimal 12 84 0 10 31.51 6.90 178 18.4 78.0 2008 2009 Minimal 13 86 0 10 28.37 6.90 178 18.4 57.8 2008 2009 Minimal 14 87 0 10 20.17 6.90 178 18.4 53.3 2008 2009 Minimal 15 88 0 10 7.35 6.90 178 18.4 67.8 2008 2009 Minimal 17 89 0 10 16.58 6.90 178 18.4 57.0 2008 2009 Minimal 18 90 0 10 9.07 6.90 178 18.4 57.0 2008 2009 Minimal WGW11 143 0 10 3.90 6.90 178 18.4 66.0 2008 2009 Minimal WGW12 148 0 10 17.78 6.90 178 18.4 66.0 2008 2009 Minimal WGW14 156 0 10 30.92 6.90 178 18.4 66.0 2008 2009 Minimal WGW15 163 0 10 24.21 6.90 178 18.4 66.0 2008 2009 Moderate 7 79 0 10 30.41 4.80 178 17.0 70.8 2008 2009 Moderate 8 80 0 10 3.99 4.80 178 17.0 54.7 2008 2009 Moderate 9 81 0 10 12.42 4.80 178 17.0 64.3 2008 2009 Moderate 10 82 0 10 17.61 4.80 178 17.0 56.3 2008 2009 Moderate 11 83 0 10 6.88 4.80 178 17.0 68.7 2008 2009 Moderate 12 85 0 10 15.71 4.80 178 17.0 68.3 2008 2009 Moderate WGW10 136 0 10 0.00 4.80 178 17.0 68.1 2008 2009 Moderate WGW6 192 0 10 21.28 4.80 178 17.0 68.1 2008 2009 Moderate WGW7 196 0 10 11.81 4.80 178 17.0 68.1 2008 2009 Moderate WGW9 206 0 10 31.30 4.80 178 17.0 68.1

PAGE 180

180 Table C 2. Continued Year Site Location Bag ID Depth (cm) Mass Loss (%) GS Soil T (C) P (mm) WT (cm) ALD (cm) 2008 2009 Extensive 1 73 0 10 30.78 7.15 178 18.7 60.3 2008 2009 Extensive 2 74 0 10 14.67 7.15 178 18.7 58.2 2008 2009 Extensive 3 75 0 10 6.85 7.15 178 18.7 49.5 2008 2009 Extensive 4 76 0 10 36.91 7.15 178 18.7 102.0 2008 2009 Extensive 5 77 0 10 18.26 7.15 178 18.7 52.8 2008 2009 Extensive 6 78 0 10 13.63 7.15 178 18.7 112.7 2008 2009 Extensive WGW1 129 0 10 32.94 7.15 178 18.7 73.8 2008 2009 Extensive WGW2 171 0 10 37.40 7.15 178 18.7 73.8 2008 2009 Extensive WGW3 178 0 10 0.72 7.15 178 18.7 73.8 2008 2009 Extensive WGW5 187 0 10 16.74 7.15 178 18.7 73.8 2009 2010 Minimal 13 103 0 10 6.08 7.80 249 15.2 75.2 2009 2010 Minimal 14 104 0 10 29.99 7.80 249 15.2 69.6 2009 2010 Minimal 15 105 0 10 22.35 7.80 249 15.2 62.5 2009 2010 Minimal 16 106 0 10 6.03 7.80 249 15.2 72.4 2009 2010 Minimal 17 107 0 10 34.35 7.80 249 15.2 61.4 2009 2010 Minimal 18 108 0 10 13.97 7.80 249 15.2 62.1 2009 2010 Minimal WGW11 145 0 10 22.76 7.80 249 15.2 68.9 2009 2010 Minimal WGW12 150 0 10 24.44 7.80 249 15.2 68.9 2009 2010 Minimal WGW14 158 0 10 26.37 7.80 249 15.2 68.9 2009 2010 Minimal WGW15 164 0 10 37.38 7.80 249 15.2 68.9 2009 2010 Moderate 7 97 0 10 33.29 7.41 249 11.8 73.1 2009 2010 Moderate 8 98 0 10 24.88 7.41 249 11.8 57.8 2009 2010 Moderate 9 99 0 10 47.75 7.41 249 11.8 66.0 2009 2010 Moderate 10 100 0 10 0.99 7.41 249 11.8 58.9 2009 2010 Moderate 11 101 0 10 18.12 7.41 249 11.8 67.6 2009 2010 Moderate 12 102 0 10 31.42 7.41 249 11.8 78.1 2009 2010 Moderate WGW10 138 0 10 16.23 7.41 249 11.8 69.3 2009 2010 Moderate WGW6 193 0 10 34.38 7.41 249 11.8 69.3 2009 2010 Moderate WGW7 198 0 10 24.60 7.41 249 11.8 69.3 2009 2010 Moderate WGW9 208 0 10 43.08 7.41 249 11.8 69.3 2009 2010 Extensive 1 91 0 10 31.01 6.44 249 13.9 64.6 2009 2010 Extensive 2 92 0 10 33.46 6.44 249 13.9 62.5 2009 2010 Extensive 3 93 0 10 42.00 6.44 249 13.9 56.6 2009 2010 Extensive 4 94 0 10 27.22 6.44 249 13.9 87.1 2009 2010 Extensive 5 95 0 10 43.64 6.44 249 13.9 59.7 2009 2010 Extensive 6 96 0 10 22.96 6.44 249 13.9 106.3 2009 2010 Extensive WGW1 131 0 10 31.37 6.44 249 13.9 75.6 2009 2010 Extensive WGW2 173 0 10 38.82 6.44 249 13.9 75.6 2009 2010 Extensive WGW3 180 0 10 17.24 6.44 249 13.9 75.6 2009 2010 Extensive WGW5 189 0 10 38.73 6.44 249 13.9 75.6 2010 2011 Minimal 13 121 0 10 10.42 4.68 164 13.6 64.3 2010 2011 Minimal 14 122 0 10 20.88 4.68 164 13.6 62.0 2010 2011 Minimal 15 123 0 10 9.45 4.68 164 13.6 52.8 2010 2011 Minimal 16 124 0 10 9.69 4.68 164 13.6 60.8 2010 2011 Minimal 17 125 0 10 18.22 4.68 164 13.6 58.0 2010 2011 Minimal 18 126 0 10 17.04 4.68 164 13.6 56.3 2010 2011 Minimal WGW11 142 0 10 25.83 4.68 164 13.6 59.1 2010 2011 Minimal WGW12 147 0 10 16.93 4.68 164 13.6 59.1 2010 2011 Minimal WGW14 155 0 10 28.04 4.68 164 13.6 59.1

PAGE 181

181 Table C 2. Continued Year Site Location Bag ID Depth (cm) Mass Loss (%) GS Soil T (C) P (mm) WT (cm) ALD (cm) 2010 2011 Minimal WGW15 162 0 10 18.27 4.68 164 13.6 59.1 2010 2011 Moderate 7 115 0 10 28.94 5.84 164 13.0 61.7 2010 2011 Moderate 8 116 0 10 18.03 5.84 164 13.0 48.0 2010 2011 Moderate 9 117 0 10 22.25 5.84 164 13.0 50.7 2010 2011 Moderate 10 118 0 10 1.59 5.84 164 13.0 49.0 2010 2011 Moderate 11 119 0 10 8.42 5.84 164 13.0 61.0 2010 2011 Moderate 12 120 0 10 6.42 5.84 164 13.0 65.5 2010 2011 Moderate WGW10 135 0 10 NA 5.84 164 13.0 56.0 2010 2011 Moderate WGW6 191 0 10 21.12 5.84 164 13.0 56.0 2010 2011 Moderate WGW7 195 0 10 12.71 5.84 164 13.0 56.0 2010 2011 Moderate WGW9 205 0 10 21.93 5.84 164 13.0 56.0 2010 2011 Extensive 1 109 0 10 19.21 5.58 164 17.3 55.2 2010 2011 Extensive 2 110 0 10 26.64 5.58 164 17.3 55.5 2010 2011 Extensive 3 111 0 10 15.20 5.58 164 17.3 48.3 2010 2011 Extensive 4 112 0 10 35.06 5.58 164 17.3 76.0 2010 2011 Extensive 5 113 0 10 48.57 5.58 164 17.3 52.3 2010 2011 Extensive 6 114 0 10 28.43 5.58 164 17.3 91.0 2010 2011 Extensive WGW1 128 0 10 19.54 5.58 164 17.3 63.1 2010 2011 Extensive WGW2 170 0 10 21.30 5.58 164 17.3 63.1 2010 2011 Extensive WGW3 177 0 10 17.01 5.58 164 17.3 63.1 2010 2011 Extensive WGW5 186 0 10 27.87 5.58 164 17.3 63.1 2004 2005 Minimal 13 13 10 20 0.00 7.28 145 20.2 60.5 2004 2005 Minimal 14 14 10 20 0.00 7.28 145 20.2 58.8 2004 2005 Minimal 15 15 10 20 0.00 7.28 145 20.2 59.5 2004 2005 Minimal 16 16 10 20 0.00 7.28 145 20.2 66.3 2004 2005 Minimal 17 17 10 20 0.00 7.28 145 20.2 65.8 2004 2005 Minimal 18 18 10 20 1.04 7.28 145 20.2 79.8 2004 2005 Minimal WGW11 141 10 20 1.17 7.28 145 20.2 68.6 2004 2005 Minimal WGW12 146 10 20 0.00 7.28 145 20.2 68.6 2004 2005 Minimal WGW14 154 10 20 0.00 7.28 145 20.2 68.6 2004 2005 Minimal WGW15 161 10 20 2.48 7.28 145 20.2 68.6 2004 2005 Moderate 7 7 10 20 0.00 8.39 145 18.4 72.1 2004 2005 Moderate 8 8 10 20 0.00 8.39 145 18.4 57.4 2004 2005 Moderate 9 9 10 20 0.00 8.39 145 18.4 61.9 2004 2005 Moderate 10 10 10 20 0.00 8.39 145 18.4 66.3 2004 2005 Moderate 11 11 10 20 0.00 8.39 145 18.4 66.6 2004 2005 Moderate 12 12 10 20 0.00 8.39 145 18.4 82.1 2004 2005 Moderate WGW10 134 10 20 1.96 8.39 145 18.4 77.9 2004 2005 Moderate WGW6 190 10 20 0.00 8.39 145 18.4 77.9 2004 2005 Moderate WGW7 194 10 20 0.00 8.39 145 18.4 77.9 2004 2005 Moderate WGW9 204 10 20 0.78 8.39 145 18.4 77.9 2004 2005 Extensive 1 1 10 20 1.31 6.67 145 15.6 59.5 2004 2005 Extensive 2 2 10 20 0.00 6.67 145 15.6 60.8 2004 2005 Extensive 3 3 10 20 0.65 6.67 145 15.6 57.5 2004 2005 Extensive 4 4 10 20 0.00 6.67 145 15.6 87.5 2004 2005 Extensive 5 5 10 20 1.83 6.67 145 15.6 61.8 2004 2005 Extensive 6 6 10 20 5.87 6.67 145 15.6 110.8 2004 2005 Extensive WGW1 127 10 20 0.00 6.67 145 15.6 77.8 2004 2005 Extensive WGW2 169 10 20 2.74 6.67 145 15.6 77.8

PAGE 182

182 Table C 2. Continued Year Site Location Bag ID Depth (cm) Mass Loss (%) GS Soil T (C) P (mm) WT (cm) ALD (cm) 2004 2005 Extensive WGW3 176 10 20 0.00 6.67 145 15.6 77.8 2004 2005 Extensive WGW5 185 10 20 0.00 6.67 145 15.6 77.8 2005 2006 Minimal 13 31 10 20 0.00 6.84 227 22.4 66.3 2005 2006 Minimal 14 32 10 20 11.36 6.84 227 22.4 54.3 2005 2006 Minimal 15 33 10 20 19.97 6.84 227 22.4 53.4 2005 2006 Minimal 16 34 10 20 12.27 6.84 227 22.4 60.3 2005 2006 Minimal 17 35 10 20 14.49 6.84 227 22.4 55.6 2005 2006 Minimal 18 36 10 20 9.66 6.84 227 22.4 56.8 2005 2006 Minimal WGW11 144 10 20 5.48 6.84 227 22.4 62.7 2005 2006 Minimal WGW12 149 10 20 17.23 6.84 227 22.4 62.7 2005 2006 Minimal WGW13 151 10 20 21.54 6.84 227 22.4 62.7 2005 2006 Minimal WGW14 157 10 20 15.01 6.84 227 22.4 62.7 2005 2006 Moderate 7 25 10 20 0.00 5.55 227 19.8 67.5 2005 2006 Moderate 8 26 10 20 13.71 5.55 227 19.8 50.9 2005 2006 Moderate 9 27 10 20 0.00 5.55 227 19.8 60.2 2005 2006 Moderate 10 28 10 20 1.44 5.55 227 19.8 54.5 2005 2006 Moderate 11 29 10 20 0.00 5.55 227 19.8 64.8 2005 2006 Moderate 12 30 10 20 4.44 5.55 227 19.8 68.3 2005 2006 Moderate WGW10 137 10 20 8.49 5.55 227 19.8 65.8 2005 2006 Moderate WGW7 197 10 20 11.62 5.55 227 19.8 65.8 2005 2006 Moderate WGW8 201 10 20 14.10 5.55 227 19.8 65.8 2005 2006 Moderate WGW9 207 10 20 15.01 5.55 227 19.8 65.8 2005 2006 Extensive 1 19 10 20 24.80 6.19 227 19.3 56.1 2005 2006 Extensive 2 20 10 20 10.18 6.19 227 19.3 59.7 2005 2006 Extensive 3 21 10 20 14.23 6.19 227 19.3 44.4 2005 2006 Extensive 4 22 10 20 15.14 6.19 227 19.3 84.7 2005 2006 Extensive 5 23 10 20 22.19 6.19 227 19.3 50.8 2005 2006 Extensive 6 24 10 20 38.51 6.19 227 19.3 106.1 2005 2006 Extensive WGW1 130 10 20 41.78 6.19 227 19.3 75.2 2005 2006 Extensive WGW2 172 10 20 50.26 6.19 227 19.3 75.2 2005 2006 Extensive WGW3 179 10 20 9.53 6.19 227 19.3 75.2 2005 2006 Extensive WGW5 188 10 20 13.71 6.19 227 19.3 75.2 2006 2007 Minimal 13 49 10 20 4.24 7.14 331 18.8 73.1 2006 2007 Minimal 14 50 10 20 9.19 7.14 331 18.8 67.5 2006 2007 Minimal 15 51 10 20 54.41 7.14 331 18.8 60.0 2006 2007 Minimal 16 52 10 20 10.86 7.14 331 18.8 67.4 2006 2007 Minimal 17 53 10 20 5.84 7.14 331 18.8 60.4 2006 2007 Minimal 18 54 10 20 8.61 7.14 331 18.8 62.0 2006 2007 Minimal WGW13 152 10 20 10.09 7.14 331 18.8 70.8 2006 2007 Minimal WGW14 159 10 20 17.50 7.14 331 18.8 70.8 2006 2007 Minimal WGW16 165 10 20 16.21 7.14 331 18.8 70.8 2006 2007 Minimal WGW17 167 10 20 22.13 7.14 331 18.8 70.8 2006 2007 Moderate 7 43 10 20 7.83 7.16 331 16.1 71.5 2006 2007 Moderate 8 44 10 20 0.00 7.16 331 16.1 58.3 2006 2007 Moderate 9 45 10 20 1.49 7.16 331 16.1 69.1 2006 2007 Moderate 10 46 10 20 16.95 7.16 331 16.1 61.0 2006 2007 Moderate 11 47 10 20 0.27 7.16 331 16.1 67.9 2006 2007 Moderate 12 48 10 20 8.29 7.16 331 16.1 79.8 2006 2007 Moderate WGW10 139 10 20 21.82 7.16 331 16.1 71.4

PAGE 183

183 Table C 2. Continued Year Site Location Bag ID Depth (cm) Mass Loss (%) GS Soil T (C) P (mm) WT (cm) ALD (cm) 2006 2007 Moderate WGW7 199 10 20 12.76 7.16 331 16.1 71.4 2006 2007 Moderate WGW8 202 10 20 11.27 7.16 331 16.1 71.4 2006 2007 Moderate WGW9 209 10 20 35.90 7.16 331 16.1 71.4 2006 2007 Extensive 1 37 10 20 19.80 7.68 331 14.3 61.5 2006 2007 Extensive 2 38 10 20 14.33 7.68 331 14.3 65.3 2006 2007 Extensive 3 39 10 20 22.60 7.68 331 14.3 53.8 2006 2007 Extensive 4 40 10 20 28.07 7.68 331 14.3 83.4 2006 2007 Extensive 5 41 10 20 20.74 7.68 331 14.3 55.8 2006 2007 Extensive 6 42 10 20 27.36 7.68 331 14.3 106.6 2006 2007 Extensive WGW1 132 10 20 21.24 7.68 331 14.3 79.1 2006 2007 Extensive WGW2 174 10 20 95.65 7.68 331 14.3 79.1 2006 2007 Extensive WGW3 181 10 20 75.49 7.68 331 14.3 79.1 2006 2007 Extensive WGW4 183 10 20 17.13 7.68 331 14.3 79.1 2007 2008 Minimal 13 67 10 20 11.23 6.00 346 15.2 76.4 2007 2008 Minimal 14 68 10 20 4.44 6.00 346 15.2 68.0 2007 2008 Minimal 15 69 10 20 17.65 6.00 346 15.2 57.8 2007 2008 Minimal 16 70 10 20 11.75 6.00 346 15.2 69.3 2007 2008 Minimal 17 71 10 20 15.52 6.00 346 15.2 59.0 2007 2008 Minimal 18 72 10 20 5.29 6.00 346 15.2 57.8 2007 2008 Minimal WGW13 153 10 20 12.28 6.00 346 15.2 81.2 2007 2008 Minimal WGW14 160 10 20 9.75 6.00 346 15.2 81.2 2007 2008 Minimal WGW16 166 10 20 16.92 6.00 346 15.2 81.2 2007 2008 Minimal WGW18 168 10 20 37.39 6.00 346 15.2 81.2 2007 2008 Moderate 7 61 10 20 0.52 4.46 346 10.9 69.6 2007 2008 Moderate 8 62 10 20 21.80 4.46 346 10.9 53.6 2007 2008 Moderate 9 63 10 20 0.00 4.46 346 10.9 70.0 2007 2008 Moderate 10 64 10 20 18.67 4.46 346 10.9 58.0 2007 2008 Moderate 11 65 10 20 3.00 4.46 346 10.9 65.3 2007 2008 Moderate 12 66 10 20 5.22 4.46 346 10.9 77.2 2007 2008 Moderate WGW10 140 10 20 24.15 4.46 346 10.9 80.3 2007 2008 Moderate WGW7 200 10 20 3.52 4.46 346 10.9 80.3 2007 2008 Moderate WGW8 203 10 20 22.45 4.46 346 10.9 80.3 2007 2008 Moderate WGW9 210 10 20 16.84 4.46 346 10.9 80.3 2007 2008 Extensive 1 55 10 20 7.18 3.08 346 10.8 57.6 2007 2008 Extensive 2 56 10 20 1.31 3.08 346 10.8 60.9 2007 2008 Extensive 3 57 10 20 19.97 3.08 346 10.8 52.5 2007 2008 Extensive 4 58 10 20 34.46 3.08 346 10.8 89.6 2007 2008 Extensive 5 59 10 20 24.15 3.08 346 10.8 55.3 2007 2008 Extensive 6 60 10 20 20.10 3.08 346 10.8 122.3 2007 2008 Extensive WGW1 133 10 20 28.20 3.08 346 10.8 84.5 2007 2008 Extensive WGW2 175 10 20 70.01 3.08 346 10.8 84.5 2007 2008 Extensive WGW3 182 10 20 50.65 3.08 346 10.8 84.5 2007 2008 Extensive WGW4 184 10 20 0.00 3.08 346 10.8 84.5 2008 2009 Minimal 12 84 10 20 6.06 6.90 178 18.4 78.0 2008 2009 Minimal 13 86 10 20 0.28 6.90 178 18.4 57.8 2008 2009 Minimal 14 87 10 20 13.01 6.90 178 18.4 53.3 2008 2009 Minimal 15 88 10 20 13.20 6.90 178 18.4 67.8 2008 2009 Minimal 17 89 10 20 0.92 6.90 178 18.4 57.0 2008 2009 Minimal 18 90 10 20 0.00 6.90 178 18.4 57.0

PAGE 184

184 Table C 2. Continued Year Site Location Bag ID Depth (cm) Mass Loss (%) GS Soil T (C) P (mm) WT (cm) ALD (cm) 2008 2009 Minimal WGW11 143 10 20 12.55 6.90 178 18.4 66.0 2008 2009 Minimal WGW12 148 10 20 0.79 6.90 178 18.4 66.0 2008 2009 Minimal WGW14 156 10 20 0.60 6.90 178 18.4 66.0 2008 2009 Minimal WGW15 163 10 20 27.46 6.90 178 18.4 66.0 2008 2009 Moderate 7 79 10 20 0.00 4.80 178 17.0 70.8 2008 2009 Moderate 8 80 10 20 0.00 4.80 178 17.0 54.7 2008 2009 Moderate 9 81 10 20 0.00 4.80 178 17.0 64.3 2008 2009 Moderate 10 82 10 20 8.32 4.80 178 17.0 56.3 2008 2009 Moderate 11 83 10 20 0.00 4.80 178 17.0 68.7 2008 2009 Moderate 12 85 10 20 1.33 4.80 178 17.0 68.3 2008 2009 Moderate WGW10 136 10 20 0.00 4.80 178 17.0 68.1 2008 2009 Moderate WGW6 192 10 20 7.82 4.80 178 17.0 68.1 2008 2009 Moderate WGW7 196 10 20 0.63 4.80 178 17.0 68.1 2008 2009 Moderate WGW9 206 10 20 6.31 4.80 178 17.0 68.1 2008 2009 Extensive 1 73 10 20 3.85 7.15 178 18.7 60.3 2008 2009 Extensive 2 74 10 20 6.28 7.15 178 18.7 58.2 2008 2009 Extensive 3 75 10 20 27.07 7.15 178 18.7 49.5 2008 2009 Extensive 4 76 10 20 2.23 7.15 178 18.7 102.0 2008 2009 Extensive 5 77 10 20 7.19 7.15 178 18.7 52.8 2008 2009 Extensive 6 78 10 20 0.39 7.15 178 18.7 112.7 2008 2009 Extensive WGW1 129 10 20 41.98 7.15 178 18.7 73.8 2008 2009 Extensive WGW2 171 10 20 48.45 7.15 178 18.7 73.8 2008 2009 Extensive WGW3 178 10 20 5.20 7.15 178 18.7 73.8 2008 2009 Extensive WGW5 187 10 20 23.64 7.15 178 18.7 73.8 2009 2010 Minimal 13 103 10 20 3.45 7.80 249 15.2 75.2 2009 2010 Minimal 14 104 10 20 15.45 7.80 249 15.2 69.6 2009 2010 Minimal 15 105 10 20 21.82 7.80 249 15.2 62.5 2009 2010 Minimal 16 106 10 20 2.01 7.80 249 15.2 72.4 2009 2010 Minimal 17 107 10 20 15.98 7.80 249 15.2 61.4 2009 2010 Minimal 18 108 10 20 4.48 7.80 249 15.2 62.1 2009 2010 Minimal WGW11 145 10 20 16.34 7.80 249 15.2 68.9 2009 2010 Minimal WGW12 150 10 20 9.97 7.80 249 15.2 68.9 2009 2010 Minimal WGW14 158 10 20 7.54 7.80 249 15.2 68.9 2009 2010 Minimal WGW15 164 10 20 28.85 7.80 249 15.2 68.9 2009 2010 Moderate 7 97 10 20 6.32 7.41 249 11.8 73.1 2009 2010 Moderate 8 98 10 20 4.14 7.41 249 11.8 57.8 2009 2010 Moderate 9 99 10 20 10.34 7.41 249 11.8 66.0 2009 2010 Moderate 10 100 10 20 4.37 7.41 249 11.8 58.9 2009 2010 Moderate 11 101 10 20 5.33 7.41 249 11.8 67.6 2009 2010 Moderate 12 102 10 20 1.98 7.41 249 11.8 78.1 2009 2010 Moderate WGW10 138 10 20 2.21 7.41 249 11.8 69.3 2009 2010 Moderate WGW6 193 10 20 24.34 7.41 249 11.8 69.3 2009 2010 Moderate WGW7 198 10 20 0.00 7.41 249 11.8 69.3 2009 2010 Moderate WGW9 208 10 20 2.86 7.41 249 11.8 69.3 2009 2010 Extensive 1 91 10 20 28.39 6.44 249 13.9 64.6 2009 2010 Extensive 2 92 10 20 6.65 6.44 249 13.9 62.5 2009 2010 Extensive 3 93 10 20 29.26 6.44 249 13.9 56.6 2009 2010 Extensive 4 94 10 20 1.18 6.44 249 13.9 87.1 2009 2010 Extensive 5 95 10 20 24.79 6.44 249 13.9 59.7

PAGE 185

185 Table C 2. Continued Year Site Location Bag ID Depth (cm) Mass Loss (%) GS Soil T (C) P (mm) WT (cm) ALD (cm) 2009 2010 Extensive 6 96 10 20 1.71 6.44 249 13.9 106.3 2009 2010 Extensive WGW1 131 10 20 46.07 6.44 249 13.9 75.6 2009 2010 Extensive WGW2 173 10 20 69.08 6.44 249 13.9 75.6 2009 2010 Extensive WGW3 180 10 20 4.66 6.44 249 13.9 75.6 2009 2010 Extensive WGW5 189 10 20 23.41 6.44 249 13.9 75.6 2010 2011 Minimal 13 121 10 20 0.00 4.68 164 13.6 64.3 2010 2011 Minimal 14 122 10 20 11.92 4.68 164 13.6 62.0 2010 2011 Minimal 15 123 10 20 12.01 4.68 164 13.6 52.8 2010 2011 Minimal 16 124 10 20 1.88 4.68 164 13.6 60.8 2010 2011 Minimal 17 125 10 20 2.55 4.68 164 13.6 58.0 2010 2011 Minimal 18 126 10 20 2.51 4.68 164 13.6 56.3 2010 2011 Minimal WGW11 142 10 20 4.10 4.68 164 13.6 59.1 2010 2011 Minimal WGW12 147 10 20 4.71 4.68 164 13.6 59.1 2010 2011 Minimal WGW14 155 10 20 6.85 4.68 164 13.6 59.1 2010 2011 Minimal WGW15 162 10 20 21.84 4.68 164 13.6 59.1 2010 2011 Moderate 7 115 10 20 0.75 5.84 164 13.0 61.7 2010 2011 Moderate 8 116 10 20 0.80 5.84 164 13.0 48.0 2010 2011 Moderate 9 117 10 20 11.64 5.84 164 13.0 50.7 2010 2011 Moderate 10 118 10 20 3.14 5.84 164 13.0 49.0 2010 2011 Moderate 11 119 10 20 1.99 5.84 164 13.0 61.0 2010 2011 Moderate 12 120 10 20 0.00 5.84 164 13.0 65.5 2010 2011 Moderate WGW10 135 10 20 NA 5.84 164 13.0 56.0 2010 2011 Moderate WGW6 191 10 20 23.67 5.84 164 13.0 56.0 2010 2011 Moderate WGW7 195 10 20 0.17 5.84 164 13.0 56.0 2010 2011 Moderate WGW9 205 10 20 10.46 5.84 164 13.0 56.0 2010 2011 Extensive 1 109 10 20 10.94 5.58 164 17.3 55.2 2010 2011 Extensive 2 110 10 20 8.15 5.58 164 17.3 55.5 2010 2011 Extensive 3 111 10 20 39.83 5.58 164 17.3 48.3 2010 2011 Extensive 4 112 10 20 9.37 5.58 164 17.3 76.0 2010 2011 Extensive 5 113 10 20 40.27 5.58 164 17.3 52.3 2010 2011 Extensive 6 114 10 20 14.39 5.58 164 17.3 91.0 2010 2011 Extensive WGW1 128 10 20 36.08 5.58 164 17.3 63.1 2010 2011 Extensive WGW2 170 10 20 35.24 5.58 164 17.3 63.1 2010 2011 Extensive WGW3 177 10 20 1.16 5.58 164 17.3 63.1 2010 2011 Extensive WGW5 186 10 20 10.69 5.58 164 17.3 63.1

PAGE 186

186 Table C 3. Annual and growing season (GS) mass loss of the common substrate in CiPEHR. These data were used in mixed model multiple regressions for annual and GS mass loss (see Methods). Environmental data included in the regression were the GS average soil temperature at 10 cm, the winter (W) average soil temperature at 10 cm, the average water table (WT), and the active layer depth (ALD) for each plot, as well as the total growing season precipitation (precip) The growing season was defined as May through September. Year Treatment Plot Bag ID Depth (cm) Mass Loss (%) GS Mass Loss (%) GS S oil T (C) W Soil T (C) Precip (mm) ALD (cm) WTD (cm) GS R eco (g C m 2 y 1 ) 2008 2009 Control 1 1 1 0 10 1.03 4.12 5.08 178 52.4 27.6 211 2008 2009 Control 1 2 2 0 10 4.64 6.58 4.71 178 51.0 26.5 159 2008 2009 Winter 1 7 3 0 10 9.30 4.37 4.28 178 61.8 36.9 274 2008 2009 Winter 1 8 4 0 10 16.57 5.47 3.27 178 62.0 13.5 288 2008 2009 Control 2 1 5 0 10 37.68 3.53 4.44 178 48.1 32.2 192 2008 2009 Winter 2 8 6 0 10 28.58 3.91 2.92 178 58.4 24.7 208 2008 2009 Control 3 1 7 0 10 3.81 4.41 4.29 178 58.0 28.4 206 2008 2009 Control 3 2 8 0 10 16.59 2.29 4.05 178 49.3 31.2 164 2008 2009 Winter 3 7 9 0 10 36.16 3.52 2.75 178 50.9 21.6 250 2008 2009 Winter 3 8 10 0 10 38.08 4.75 2.21 178 60.6 27.0 243 2008 2009 Control 4 3 11 0 10 48.60 3.84 4.52 178 55.8 29.2 197 2008 2009 Control 4 4 12 0 10 5.81 3.77 3.72 178 59.4 33.8 195 2008 2009 Winter 4 7 13 0 10 10.58 3.64 2.94 178 56.5 34.4 235 2008 2009 Winter 4 8 14 0 10 38.37 3.99 2.89 178 56.8 33.5 242 2008 2009 Control 5 1 15 0 10 22.16 4.48 4.15 178 54.4 19.5 266 2008 2009 Control 5 2 16 0 10 7.64 5.14 4.67 178 48.4 28.4 185 2008 2009 Winter 5 7 17 0 10 2.94 4.82 3.14 178 51.1 31.5 275 2008 2009 Winter 5 8 18 0 10 32.05 4.99 2.42 178 58.6 20.2 209 2008 2009 Control 6 1 19 0 10 15.49 4.15 4.8 178 42.3 26.7 224 2008 2009 Control 6 2 20 0 10 9.29 4.96 5.01 178 54.2 21.3 153 2008 2009 Winter 6 7 21 0 10 12.17 5.09 4.28 178 55.1 19.3 241 2008 2009 Winter 6 7 22 0 10 11.17 3.97 3.02 178 55.1 20.1 241 2009 2010 Summer 1 1 23 0 10 71.68 38.71 5.41 5.08 249 57.7 19.1 331 2009 2010 Control 1 2 24 0 10 48.65 28.39 5.25 4.71 249 56.5 16.6 201 2009 2010 Annual 1 5 25 0 10 19.78 48.30 4.37 2.82 249 59.8 30.3 328 2009 2010 Winter 1 6 26 0 10 41.28 5.75 3.74 2.2 249 68.2 8.7 306 2009 2010 Summer 2 1 27 0 10 23.33 21.60 5.93 4.44 249 56.7 26.3 333 2009 2010 Control 2 2 28 0 10 27.88 34.71 4.83 4.42 249 62.0 21.5 296 2009 2010 Annual 2 5 29 0 10 26.17 11.15 4.79 4.11 249 65.0 12.2 308 2009 2010 Winter 2 6 30 0 10 54.60 13.30 5.62 3.41 249 69.7 18.3 260

PAGE 187

187 Table C 3. Continued Year Treatment Plot Bag ID Depth (cm) Mass Loss (%) GS Mass Loss (%) GS Soil T (C) W Soil T (C) Precip (mm) ALD (cm) WTD (cm) GS R eco (g C m 2 y 1 ) 2009 2010 Summer 3 1 31 0 10 35.99 29.21 5.1 0 4.29 249 60.8 19.8 262 2009 2010 Control 3 2 32 0 10 10.46 14.09 3.05 4.05 249 52.8 25.4 275 2009 2010 Annual 3 5 33 0 10 29.95 27.62 5.63 3.63 249 61.3 13.2 305 2009 2010 Winter 3 6 34 0 10 62.84 21.38 5.39 3.74 249 61.0 18.0 248 2009 2010 Summer 4 1 35 0 10 6.60 14.27 5.27 4.52 249 60.3 21.1 292 2009 2010 Control 4 2 36 0 10 28.75 13.19 4.14 3.72 249 62.0 27.3 273 2009 2010 Annual 4 5 37 0 10 34.73 13.46 4.55 2.92 249 58.0 24.8 239 2009 2010 Winter 4 6 38 0 10 18.80 29.28 2.68 3.65 249 56.8 26.1 339 2009 2010 Summer 5 1 39 0 10 25.28 8.46 5.03 4.15 249 59.5 12.8 352 2009 2010 Control 5 2 40 0 10 34.05 17.58 5.33 4.67 249 54.3 19.9 277 2009 2010 Annual 5 5 41 0 10 25.44 9.07 4.88 2.59 249 61.7 24.8 307 2009 2010 Winter 5 6 42 0 10 51.39 28.86 6.29 3.19 249 62.8 10.8 346 2009 2010 Summer 6 1 43 0 10 58.72 15.11 5.05 4.8 249 53.0 18.2 324 2009 2010 Control 6 2 44 0 10 41.84 23.22 7.03 5.01 249 61.2 15.7 238 2009 2010 Annual 6 5 45 0 10 81.07 25.77 4.25 2.31 249 65.8 9.8 246 2009 2010 Winter 6 6 46 0 10 70.65 96.71 6.04 2.79 249 64.5 12.8 325 2010 2011 Summer 1 1 47 0 10 71.63 33.81 4.07 4.09 164 57.7 20.7 305 2010 2011 Control 1 2 48 0 10 34.52 14.57 4.99 4.56 164 57.5 21.9 240 2010 2011 Annual 1 5 49 0 10 15.50 0.00 4.04 1.5 164 63.2 30.5 357 2010 2011 Winter 1 6 50 0 10 14.87 7.16 3.4 0 0.95 164 67.3 11.2 307 2010 2011 Summer 2 1 51 0 10 21.85 8.01 3.73 3.73 164 52.5 28.0 241 2010 2011 Control 2 2 52 0 10 29.93 7.77 4.64 4.16 164 59.3 26.9 274 2010 2011 Annual 2 5 53 0 10 21.32 8.81 5.29 1.81 164 64.8 14.9 291 2010 2011 Winter 2 6 54 0 10 15.06 0.69 5.16 1.7 164 70.5 19.6 278 2010 2011 Summer 3 1 55 0 10 24.25 15.05 4.7 0 4.23 164 61.0 21.9 273 2010 2011 Control 3 2 56 0 10 16.61 9.99 2.75 3.83 164 56.5 27.2 254 2010 2011 Annual 3 5 57 0 10 18.62 10.67 5.04 1.97 164 58.8 16.3 276 2010 2011 Winter 3 6 58 0 10 56.15 14.66 5.04 2.15 164 59.0 21.7 278 2010 2011 Summer 4 1 59 0 10 9.59 0.28 4.46 3.75 164 58.8 25.4 266 2010 2011 Control 4 2 60 0 10 14.03 8.19 3.61 3.08 164 61.7 30.8 242 2010 2011 Annual 4 5 61 0 10 16.42 12.15 4.56 1.5 164 59.2 27.6 268 2010 2011 Winter 4 6 62 0 10 41.22 15.19 4.42 0.98 164 53.8 27.8 356 2010 2011 Summer 5 1 63 0 10 19.70 4.86 3.76 4.12 164 57.8 19.0 349 2010 2011 Control 5 2 64 0 10 17.90 3.77 3.95 4.35 164 52.5 21.4 238 2010 2011 Annual 5 5 65 0 10 4.98 2.66 3.73 1.35 164 61.0 28.8 343

PAGE 188

188 Table C 3. Continued Year Treatment Plot Bag ID Depth (cm) Mass Loss (%) GS Mass Loss (%) GS Soil T (C) W Soil T (C) Precip (mm) ALD (cm) WTD (cm) GS R eco (g C m 2 y 1 ) 2010 2011 Winter 5 6 66 0 10 41.89 5.89 5.58 1.69 164 60.0 12.9 367 2010 2011 Summer 6 1 67 0 10 32.99 20.24 4.14 4.57 164 48.7 19.3 287 2010 2011 Control 6 2 68 0 10 13.33 4.40 5.3 0 4.36 164 60.8 18.5 183 2010 2011 Annual 6 5 69 0 10 88.60 34.59 4.28 1.47 164 63.0 11.6 253 2010 2011 Winter 6 6 70 0 10 88.81 68.26 5.68 1.99 164 63.2 16.3 410 2008 2009 Control 1 1 1 10 20 19.80 4.12 5.08 178 52.4 27.6 211 2008 2009 Control 1 2 2 10 20 1.15 6.58 4.71 178 51.0 26.5 159 2008 2009 Winter 1 7 3 10 20 5.66 4.37 4.28 178 61.8 36.9 274 2008 2009 Winter 1 8 4 10 20 62.24 5.47 3.27 178 62.0 13.5 288 2008 2009 Control 2 1 5 10 20 41.45 3.53 4.44 178 48.1 32.2 192 2008 2009 Winter 2 8 6 10 20 0.97 3.91 2.92 178 58.4 24.7 208 2008 2009 Control 3 1 7 10 20 4.12 4.41 4.29 178 58.0 28.4 206 2008 2009 Control 3 2 8 10 20 2.03 2.29 4.05 178 49.3 31.2 164 2008 2009 Winter 3 7 9 10 20 17.87 3.52 2.75 178 50.9 21.6 250 2008 2009 Winter 3 8 10 10 20 4.89 4.75 2.21 178 60.6 27.0 243 2008 2009 Control 4 3 11 10 20 22.43 3.84 4.52 178 55.8 29.2 197 2008 2009 Control 4 4 12 10 20 3.42 3.77 3.72 178 59.4 33.8 195 2008 2009 Winter 4 7 13 10 20 21.94 3.64 2.94 178 56.5 34.4 235 2008 2009 Winter 4 8 14 10 20 19.24 3.99 2.89 178 56.8 33.5 242 2008 2009 Control 5 1 15 10 20 2.29 4.48 4.15 178 54.4 19.5 266 2008 2009 Control 5 2 16 10 20 8.69 5.14 4.67 178 48.4 28.4 185 2008 2009 Winter 5 7 17 10 20 8.90 4.82 3.14 178 51.1 31.5 275 2008 2009 Winter 5 8 18 10 20 16.35 4.99 2.42 178 58.6 20.2 209 2008 2009 Control 6 1 19 10 20 4.71 4.15 4.8 178 42.3 26.7 224 2008 2009 Control 6 2 20 10 20 11.55 4.96 5.01 178 54.2 21.3 153 2008 2009 Winter 6 7 21 10 20 4.18 5.09 4.28 178 55.1 19.3 241 2008 2009 Winter 6 7 22 10 20 4.18 3.97 3.02 178 55.1 20.1 241 2009 2010 Summer 1 1 23 10 20 47.43 5.41 5.08 249 57.7 19.1 331 2009 2010 Control 1 2 24 10 20 27.58 5.25 4.71 249 56.5 16.6 201 2009 2010 Annual 1 5 25 10 20 8.38 4.37 2.82 249 59.8 30.3 328 2009 2010 Winter 1 6 26 10 20 15.13 3.74 2.2 249 68.2 8.7 306 2009 2010 Summer 2 1 27 10 20 22.17 5.93 4.44 249 56.7 26.3 333 2009 2010 Control 2 2 28 10 20 1.33 4.83 4.42 249 62.0 21.5 296 2009 2010 Annual 2 5 29 10 20 5.01 4.79 4.11 249 65.0 12.2 308 2009 2010 Winter 2 6 30 10 20 0.00 5.62 3.41 249 69.7 18.3 260

PAGE 189

189 Table C 3. Continued Year Treatment Plot Bag ID Depth (cm) Mass Loss (%) GS Mass Loss (%) GS Soil T (C) W Soil T (C) Precip (mm) ALD (cm) WTD (cm) GS R eco (g C m 2 y 1 ) 2009 2010 Summer 3 1 31 10 20 21.17 5.1 0 4.29 249 60.8 19.8 262 2009 2010 Control 3 2 32 10 20 7.31 3.05 4.05 249 52.8 25.4 275 2009 2010 Annual 3 5 33 10 20 3.15 5.63 3.63 249 61.3 13.2 305 2009 2010 Winter 3 6 34 10 20 16.10 5.39 3.74 249 61.0 18.0 248 2009 2010 Summer 4 1 35 10 20 29.11 5.27 4.52 249 60.3 21.1 292 2009 2010 Control 4 2 36 10 20 8.22 4.14 3.72 249 62.0 27.3 273 2009 2010 Annual 4 5 37 10 20 15.54 4.55 2.92 249 58.0 24.8 239 2009 2010 Winter 4 6 38 10 20 26.03 2.68 3.65 249 56.8 26.1 339 2009 2010 Summer 5 1 39 10 20 2.18 5.03 4.15 249 59.5 12.8 352 2009 2010 Control 5 2 40 10 20 5.32 5.33 4.67 249 54.3 19.9 277 2009 2010 Annual 5 5 41 10 20 1.25 4.88 2.59 249 61.7 24.8 307 2009 2010 Winter 5 6 42 10 20 53.14 6.29 3.19 249 62.8 10.8 346 2009 2010 Summer 6 1 43 10 20 9.44 5.05 4.8 249 53.0 18.2 324 2009 2010 Control 6 2 44 10 20 11.06 7.03 5.01 249 61.2 15.7 238 2009 2010 Annual 6 5 45 10 20 34.81 4.25 2.31 249 65.8 9.8 246 2009 2010 Winter 6 6 46 10 20 54.53 6.04 2.79 249 64.5 12.8 325 2010 2011 Summer 1 1 47 10 20 23.67 4.07 4.09 164 57.7 20.7 305 2010 2011 Control 1 2 48 10 20 13.97 4.99 4.56 164 57.5 21.9 240 2010 2011 Annual 1 5 49 10 20 15.46 4.04 1.5 164 63.2 30.5 357 2010 2011 Winter 1 6 50 10 20 13.21 3.4 0 0.95 164 67.3 11.2 307 2010 2011 Summer 2 1 51 10 20 9.94 3.73 3.73 164 52.5 28.0 241 2010 2011 Control 2 2 52 10 20 3.66 4.64 4.16 164 59.3 26.9 274 2010 2011 Annual 2 5 53 10 20 6.23 5.29 1.81 164 64.8 14.9 291 2010 2011 Winter 2 6 54 10 20 0.00 5.16 1.7 164 70.5 19.6 278 2010 2011 Summer 3 1 55 10 20 11.47 4.7 0 4.23 164 61.0 21.9 273 2010 2011 Control 3 2 56 10 20 3.70 2.75 3.83 164 56.5 27.2 254 2010 2011 Annual 3 5 57 10 20 27.78 5.04 1.97 164 58.8 16.3 276 2010 2011 Winter 3 6 58 10 20 18.49 5.04 2.15 164 59.0 21.7 278 2010 2011 Summer 4 1 59 10 20 19.14 4.46 3.75 164 58.8 25.4 266 2010 2011 Control 4 2 60 10 20 10.04 3.61 3.08 164 61.7 30.8 242 2010 2011 Annual 4 5 61 10 20 25.79 4.56 1.5 164 59.2 27.6 268 2010 2011 Winter 4 6 62 10 20 38.44 4.42 0.98 164 53.8 27.8 356 2010 2011 Summer 5 1 63 10 20 4.03 3.76 4.12 164 57.8 19.0 349 2010 2011 Control 5 2 64 10 20 6.80 3.95 4.35 164 52.5 21.4 238 2010 2011 Annual 5 5 65 10 20 27.92 3.73 1.35 164 61.0 28.8 343

PAGE 190

190 Table C 3. Continued Year Treatment Plot Bag ID Depth (cm) Mass Loss (%) GS Mass Loss (%) GS Soil T (C) W Soil T (C) Precip (mm) ALD (cm) WTD (cm) GS R eco (g C m 2 y 1 ) 2010 2011 Winter 5 6 66 10 20 26.09 5.58 1.69 164 60.0 12.9 367 2010 2011 Summer 6 1 67 10 20 7.70 4.14 4.57 164 48.7 19.3 287 2010 2011 Control 6 2 68 10 20 18.49 5.3 0 4.36 164 60.8 18.5 183 2010 2011 Annual 6 5 69 10 20 13.99 4.28 1.47 164 63.0 11.6 253 2010 2011 Winter 6 6 70 10 20 41.04 5.68 1.99 164 63.2 16.3 410

PAGE 191

191 LIST OF REFERENCES ACIA (2005) Arctic Impact Climate Assessment, Cambridge University Press, New York. Adair EC, Hobbie SE, Hobbie RK (2010) Single pool exponential decomposition models: potential pitfalls in their use in ecological studies. Ecology, 91 1225 1236. Aerts R (2006) The freezer defrosting: global warming and litter decomposition rates in cold biomes. Journal of Ecology, 94 713 724. Amthor JS (2000) The McCree de Wit Penning de Vries Thornley respiration paradigms: 30 years later. Annals of Botany, 86 1 20. Andrews JA, Matamala R, Westover KM, Schlesinger WH (2000) Temperature effects on the d iversity of soil heterotrophs and the delta C 13 of soil respired CO2. Soil Biology & Biochemistry, 32 699 706. Avis CA, Weaver AJ, Meissner KJ (2011) Reduction in areal extent of high latitude wetlands in response to permafrost thaw. Nature Geoscience, 4 444 448. Badeck FW, Tcherkez G, Nogues S, Piel C, Ghashghaie J (2005) Post photo synthetic fractionation of stable carbon isotopes between plant organs a widespread phenomenon. Rapid Communications in Mass Spectrometry, 19 1381 1391. Baptist F, Yoccoz NG, Choler P (2010) Direct and indirect control by snow cover over decomposition in alpine tundra along a snowmelt gradient. Plant and Soil, 328 397 410. Bauer JE, Williams PM, Druffel ERM (1992) Recovery of submilligram quantities of carbon dioxide from gas streams by molecular sieve for subsequent determination of isotopic (C 13 and C 14) natural abundances. Analytical Chemistry, 64 824 827. Biasi C, Rusalimova O, Meyer H et al. (2005) Temperature dependent shift from labile to recalcitrant carbon sour ces of arctic heterotrophs. John Wiley & Sons Ltd. Billett MF, Garnett MH, Harvey F (2007) UK peatland streams release old carbon dioxide to the atmosphere and young dissolved organic carbon to rivers. Geophysical Research Letters, 34 Blagodatskaya E, Yuy ukina T, Blagodatsky S, Kuzyakov Y (2011) Three source partitioning of microbial biomass and of CO2 efflux from soil to evaluate mechanisms of priming effects. Soil Biology & Biochemistry, 43 778 786. Borken W, Savage K, Davidson EA, Trumbore SE (2006) Ef fects of experimental drought on soil respiration and radiocarbon efflux from a temperate forest soil. Global Change Biology, 12 177 193.

PAGE 192

192 Bostrom B, Comstedt D, Ekblad A (2007) Isotope fractionation and C 13 enrichment in soil profiles during the decompos ition of soil organic matter. Oecologia, 153 89 98. Bowling DR, Mcdowell NG, Bond BJ, Law BE, Ehleringer JR (2002) C 13 content of ecosystem respiration is linked to precipitation and vapor pressure deficit. Oecologia, 131 113 124. Bowling DR, Pataki DE, Randerson JT (2008) Carbon isotopes in terrestrial ecosystem pools and CO2 fluxes. New Phytologist, 178 24 40. Bret Harte MS, Shaver GR, Chapin FS (2002) Primary and secondary stem growth in arctic shrubs: Implications for community response to environme ntal change. Journal of Ecology, 90 251 267. Callaghan TV, Bjorn LO, Chernov Y et al. (2004) Effects on the function of arctic ecosystems in the short and long term perspectives. Ambio, 33 448 458. Carbone MS, Still CJ, Ambrose AR et al. (2011) Seasonal and episodic moisture controls on plant and microbial contributions to soil respiration. Oecologia, 167 265 278. Chapin FS, Matson PA, Mooney HA (2002) Principles of Ecosystem Ecology, Springer, New York. Chapin FS, Woodwell GM, Randerson JT et al. (2006 ) Reconciling carbon cycle concepts, terminology, and methods. Ecosystems, 9 1041 1050. Chiti T, Neubert R, Janssens I, Yuste J, Sirignano C, Certini G (2011) Radiocarbon based assessment of soil organic matter contribution to soil respiration in a pine s tand of the Campine region, Belgium. Plant and Soil, 344 273 282. Cisneros Dozal LM, Trumbore S, Hanson PJ (2006) Partitioning sources of soil respired CO2 and their seasonal variation using a unique radiocarbon tracer. Global Change Biology, 12 194 204. Clymo RS (1965) Experiments on breakdown of sphagnum in 2 bogs. Journal of Ecology, 53 747 758. Conant RT, Ryan MG, Agren GI et al. (2011) Temperature and soil organic matter decomposition rates synthesis of current knowledge and a way forward. Global Change Biology, 17 3392 3404. Cornelissen JHC, Van Bodegom PM, Aerts R et al. (2007) Global negative vegetation feedback to climate warming responses of leaf litter decomposition rates in cold biomes. Ecology Letters, 10 619 627.

PAGE 193

193 Czimczik CI, Trumbore SE (2007) Short term controls on the age of microbial carbon sources in boreal forest soils. Journal of Geophysical Research Biogeosciences, 112 Czimczik CI, Trumbore SE, Carbone MS, Winston GC (2006) Changing sources of soil respiration with time since fire in a boreal forest. Global Change Biology, 12 957 971. Dagg J, Lafleur P (2011) Vegetation c ommunity, f oliar n itrogen, and t emperature e ffects on t undra CO2 e xcha nge across a s oil m oisture g radient. Arctic Antarctic and Alpine Research, 43 189 197. Davidson EA, Janssens IA (2006) Temperature sensitivity of soil carbon decomposition and feedbacks to climate change. Nature, 440 165 173. Dawson TE, Mambelli S, Plamb oeck AH, Templer PH, Tu KP (2002) Stable isotopes in plant ecology. Annual Review of Ecology and Systematics 33 507 559 De Graaff MA, Classen AT, Castro HF, Schadt CW (2010) Labile soil carbon inputs mediate the soil microbial community composition and plant residue decomposition rates. New Phytologist, 188 1055 1064. Demarco J (2011) Plant soil feedbacks with changing vegetation structure and composition in a warming Arctic. Unpublished PhD thesis, University of Florida, Gainesville, FL, 142 pp. Diouma eva I, Trumbore S, Schuur EAG, Goulden ML, Litvak M, Hirsch AI (2002) Decomposition of peat from upland boreal forest: Temperature dependence and sources of respired carbon. Journal of Geophysical Research Atmospheres, 108 12. Dorrepaal E, Cornelissen JHC Aerts R, Wallen B, Van Logtestijn RSP (2005) Are growth forms consistent predictors of leaf litter quality and decomposability across peatlands along a latitudinal gradient? Journal of Ecology, 93 817 828. Dorrepaal E, Toet S, Van Logtestijn RSP, Swart E, Van De Weg MJ, Callaghan TV, Aerts R (2009) Carbon respiration from subsurface peat accelerated by climate warming in the subarctic. Nature, 460 616 U679. Dutta K, Schuur EAG, Neff JC, Zimov SA (2006) Potential carbon release from permafrost soils of N ortheastern Siberia. Global Change Biology, 12 2336 2351. Ehleringer JR, Buchmann N, Flanagan LB (2000) Carbon isotope ratios in belowground carbon cycle processes. Ecological Applications, 10 412 422. Follett RF, Paul EA, Pruessner EG (2007) Soil carbon dynamics during a long term incubation study involving C 13 and C 14 measurements. Soil Science, 172 189 208.

PAGE 194

194 Fox JF, Vancleve K (1983) Relationships between cellulose decomposition, Jennys k, forest floor nitrogen, and soil tempera ture in Alaskan taiga forests. Canadian Journal of Forest Research Revue Canadienne De Recherche Forestiere, 13 789 794. Gaudinski JB, Dawson TE, Quideau S et al. (2005) Comparative analysis of cellulose preparation techniques for use with C 13, C 14, and O 18 isotopic measurements. Analytical Chemistry, 77 7212 7224. Gaudinski JB, Trumbore SE, Davidson EA, Zheng SH (2000) Soil carbon cycling in a temperate forest: radiocarbon based estimates of residence times, sequestration rates and partitioning of flu xes. Biogeochemistry, 51 33 69. Gebauer RLE, Tenhunen JD, Reynolds JF (1996) Soil aeration in relation to soil physical properties, nitrogen availability, and root characteristics within an arctic watershed. Plant and Soil, 178 37 48. Gomez Casanovas N, Matamala R, Cook DR, Gonzalez Meler MA (2012) Net ecosystem exchange modifies the relationship between the autotrophic and heterotrophic components of soil respiration with abiotic factors in prairie grasslands. Global Change Biology, 18 2532 2545. Goulde n ML, Mcmillan AMS, Winston GC, Rocha AV, Manies KL, Harden JW, Bond Lamberty BP (2011) Patterns of NPP, GPP, respiration, and NEP during boreal forest succession. Global Change Biology, 17 855 871. Goulden ML, Wofsy SC, Harden JW et al. (1998) Sensitivit y of boreal forest carbon balance to soil thaw. Science, 279 214 217. Gower ST, Krankina O, Olson RJ, Apps M, Linder S, Wang C (2001) Net primary production and carbon allocation patterns of boreal forest ecosystems. Ecological Applications, 11 1395 1411 Grant RF, Humphreys ER, Lafleur PM, Dimitrov DD (2011) Ecological controls on net ecosystem productivity of a mesic arctic tundra under current and future climates. Journal of Geophysical Research Biogeosciences, 116 17. Hardie SML, Garnett MH, Fallick AE, Ostle NJ, Rowland AP (2009) Bomb C 14 analysis of ecosystem respiration reveals that peatland vegetation facilitates release of old carbon. Geoderma, 153 393 401. Hardie SML, Garnett MH, Fallick AE, Rowland AP, Ostle NJ (2005) Carbon dioxide capture u sing a zeolite molecular sieve sampling system for isotopic studies (C 13 and C 14) of respiration. Radiocarbon, 47 441 451. Hicks Pries CE, Schuur EAG, Crummer KG (2012) Holocene c arbon s tocks and c arbon a ccumulation r ates a ltered in s oils u ndergoing p er mafrost t haw. Ecosystems, 12 162 173.

PAGE 195

195 Hobbie EA, Ouimette AP (2009) Controls of nitrogen isotope patterns in soil profiles. Biogeochemistry, 95 355 371. Hobbie SE (1996) Temperature and plant species control over litter decomposition in Alaskan tundra. E cological Monographs, 66 503 522. Hobbie SE (2005) Contrasting effects of substrate and fertilizer nitrogen on the early stages of litter decomposition. Ecosystems, 8 644 656. Hobbie SE, Chapin FS (1996) Winter regulation of tundra litter carbon and nitrogen dynamics. Biogeochemistry, 35 327 338. Hobbie SE, Chapin FS (1998) Response of tundra plant biomass, aboveground production, nitrogen, and CO2 flux to experimental warming. Ecology, 79 1526 1544. Hobbie SE, Gough L (2004) Litter decomposition in moist acidic and non acidic tundra with different glacial histories. Oecologia, 140 113 124. Hobbie SE, Schimel JP, Trumbore SE, Randerson JR (2000) Controls over carbon storage and turnover in high latitude soi ls. Global Change Biology, 6 196 210. Hgberg P, Ekblad A, Nordgren A, Plamboeck AH, Ohlsson A, Bhupinderpal Singh, Hgberg MH (2005) Factors determining the 13C abundance of soil respired CO2 in boreal forests. In: Stable Isotopes and Biosphere Atmospher e Interactions: Processes and Biological Controls. (eds Flanagan LB, Ehleringer JR, Pataki DE) pp Page. Elsevier Academic Press. Hogberg P, Nordgren A, Buchmann N et al. (2001) Large scale forest girdling shows that current photosynthesis drives soil respi ration. Nature, 411 789 792. Hua Q, Barbetti M (2004) Review of tropospheric bomb C 14 data for carbon cycle modeling and age calibration purposes. Radiocarbon, 46 1273 1298. Huemmrich KF, Kinoshita G, Gamon JA, Houston S, Kwon H, Oechel WC (2010) Tundra carbon balance under varying temperature and moisture regimes. Journal of Geophysical Research Biogeosciences, 115 Hugelius G, Kuhry P, Tarnocai C, Virtanen T (2010) Soil o rganic c arbon p ools in a p eriglacial l andscape: A c ase s tudy from the c entral Cana dian Arctic. Permafrost and Periglacial Processes, 21 16 29. Inger R, Jackson A, Parnell A, Bearhop S (2010) SIAR V4 (Stable Isotope Analysis in http://www.tcd.ie/Zoology/research/research/ theoretical /siar/SIAR_For_Ecologists.p df > IPCC (2007) Climate Change 2007: The Physical Science Basis, Cambridge University Press, New York.

PAGE 196

196 Jia GSJ, Epstein HE, Walker DA (2003) Greening of arctic Alaska, 1981 2001. Geophysical Research Letters, 30 Jorgenson MT, Osterkamp TE (2005) Response of boreal ecosystems to varying modes of permafrost degradation. Canadian Journal of Forest Research Revue Canadienne De Recherche Forestiere, 35 2100 2111. Kasischke ES, Turetsky MR (2006) Recent changes in the fire regime across the North American borea l region Spatial and temporal patterns of burning across Canada and Alaska. Geophysical Research Letters, 33 Keeling CD (1958) The concentration and isotopic abundances of atmospheric carbon dioxide in rural areas. Geochimica Et Cosmochimica Acta, 13 322 334. Klumpp K, Schaufele R, Lotscher M, Lattanzi FA, Feneis W, Schnyder H (2005) C isotope composition of CO2 respired by shoots and roots: fractionation during dark respiration? Plant Cell and Environment, 28 241 250. Koven CD, Ringeval B, Friedlings tein P et al. (2011) Permafrost carbon climate feedbacks accelerate global warming. Proceedings of the National Academy of Sciences of the United States of America, 108 14769 14774. Kuhry P, Vitt DH (1996) Fossil carbon/nitrogen ratios as a measure of pea t decomposition. Ecology, 77 271 275. Kuzyakov Y (2006) Sources of CO2 efflux from soil and review of partitioning methods. Soil Biology & Biochemistry, 38 425 448. Kuzyakov Y, Hill PW, Jones DL (2007) Root exudate components change litter decomposition in a simulated rhizosphere depending on temperature. Plant and Soil, 290 293 305. Lambers H, Chapin FS, Pons TL (2008) Plant Physiological Ecology, Springer. Lang SI, Cornelissen JHC, Klahn T, Van Logtestijn RSP, Broekman R, Schweikert W, Aerts R (2009) A n experimental comparison of chemical traits and litter decomposition rates in a diverse range of subarctic bryophyte, lichen and vascular plant species. Journal of Ecology, 97 886 900. Lavelle P, Blanchart E, Martin A et al. (1993) A hierarchal model for decomposition in terrestrial ecosystems application to soils of the humid tropics. Biotropica, 25 130 150. Lavoie M, Mack MC, Schuur EAG (2011) Effects of elevated nitrogen and temperature on carbon and nitrogen dynamics in Alaskan arctic and boreal soil s. Journal of Geophysical Research Biogeosciences, 116

PAGE 197

197 Lawrence DM, Slater AG, Romanovsky VE, Nicolsky DJ (2008) Sensitivity of a model projection of near surface permafrost degradation to soil column depth and representation of soil organic matter. Journ al of Geophysical Research Earth Surface, 113 14. Lee H, Schuur EAG, Inglett KS, Lavoie M, Chanton JP (2012) The rate of permafrost carbon release under aerobic and anaerobic conditions and its potential effects on climate. Global Change Biology, 18 515 527. Lee H, Schuur EAG, Vogel JG (2010) Soil CO(2) production in upland tundra where permafrost is thawing. Journal of Geophysical Research Biogeosciences, 115 Levin I, Hesshaimer V (2000) Radiocarbon A unique tracer of global carbon cycle dynamics. Rad iocarbon, 42 69 80. Liu WX, Zhang Z, Wan SQ (2009) Predominant role of water in regulating soil and microbial respiration and their responses to climate change in a semiarid grassland. Global Change Biology, 15 184 195. Luan JW, Liu SR, Wang JX, Zhu XL, Shi ZM (2011) Rhizospheric and heterotrophic respiration of a warm temperate oak chronosequence in China. Soil Biology & Biochemistry, 43 503 512. Lund M, Lafleur PM, Roulet NT et al. (2010) Variability in exchange of CO2 across 12 northern peatland and t undra sites. Global Change Biology, 16 2436 2448. Mackelprang R, Waldrop MP, Deangelis KM et al. (2011) Metagenomic analysis of a permafrost microbial community reveals a rapid response to thaw. Nature, 480 368 U120. Malmer N, Holm E (1984) Variation in the C/N quotient of peat in relation to decomposition rate and age determination with PB 210. Oikos, 43 171 182. Mayorga E, Aufdenkampe AK, Masiello CA et al. (2005) Young organic matter as a source of carbon dioxide outgass ing from Amazonian rivers. Nature, 436 538 541. Mc C laugherty CA, Pastor J, Aber JD, Melillo JM (1985) Forest litter decomposition in relation to soil nitrogen dynamics and litter quality. Ecology, 66 266 275. McClellan MH, Bormann BT, Cromack K (1990) Cellulose decomposition in southeast Alaskan forests Effects of pit and mound microrelief and burial depth. Canadian Journal of Forest Research Revue Canadienne De Recherche Forestiere, 20 1242 1246. McDowell NG, Bowling DR, Schauer A, Irvine J, Bond BJ, Law BE, Ehleringer JR (2004) Associations between carbon isotope ratios of ecosystem respiration, water availability and canopy conductance. Global Change Biology, 10 1767 1784.

PAGE 198

198 Melillo JM, Steudler PA, Aber JD et al. (2002) Soil warming and c arbon cycle feedbacks to the climate system. Science, 298 2173 2176. Michaelson GJ, Ping CL, Kimble JM (1996) Carbon storage and distribution in tundra soils of Arctic Alaska, USA. Arctic and Alpine Research, 28 414 424. Midwood AJ, Gebbing T, Wendler R, Sommerkorn M, Hunt JE, Millard P (2006) Collection and storage of CO2 for C 13 analysis: an application to separate soil CO2 efflux into root and soil derived components. Rapid Communications in Mass Spectrometry, 20 3379 3384. Millard P, Midwood AJ, Hu nt JE, Whitehead D, Boutton TW (2008) Partitioning soil surface CO2 efflux into autotrophic and heterotrophic components, using natural gradients in soil delta C 13 in an undisturbed savannah soil. Soil Biology & Biochemistry, 40 1575 1582. Mortazavi B, C hanton JP (2004) Use of Keeling plots to determine sources of dissolved organic carbon in nearshore and open ocean systems. Limnology and Oceanography, 49 102 108. Muhr J, Borken W, Matzner E (2009) Effects of soil frost on soil respiration and its radioc arbon signature in a Norway spruce forest soil. Global Change Biology, 15 782 793. Natali SM, Schuur EAG, Webb E, Pries CEH, Crummer KG (in review) Permafrost degradation stimulates carbon loss from experimentally warmed tundra. Natali SM, Schuur EAG, Rub in RL (2012) Increased plant productivity in Alaskan tundra as a result of experimental warming of soil and permafrost. Journal of Ecology, 100 Natali SM, Schuur EAG, Trucco C, Pries CEH, Crummer KG, Lopez AFB (2011) Effects of experimental warming of air soil and permafrost on carbon balance in Alaskan tundra. Global Change Biology, 17 1394 1407. Ngao J, Epron D, Brechet C, Granier A (2005) Estimating the contribution of leaf litter decomposition to soil CO2 efflux in a beech forest using C 13 depleted litter. Global Change Biology, 11 1768 1776. Nowinski NS, Taneva L, Trumbore SE, Welker JM (2010) Decomposition of old organic matter as a result of deeper active layers in a snow depth manipulation experiment. Oecologia, 163 785 792. Oberbauer SF, Starr G, Pop EW (1998) Effects of extended growing season and soil warming on carbon dioxide and methane exchange of tussock tundra in Alaska. Journal of Geophysical Research Atmospheres, 103 29075 29082.

PAGE 199

199 Oberbauer SF, Tweedie CE, Welker JM et al. (2007) Tundr a CO2 fluxes in response to experimental warming across latitudinal and moisture gradients. Ecological Monographs, 77 221 238. O'Donnell JA, Harden JW, Mcguire AD, Kanevskiy MZ, Jorgenson MT, Xu XM (2011) The effect of fire and permafrost interactions on soil carbon accumulation in an upland black spruce ecosystem of interior Alaska: implications for post thaw carbon loss. Global Change Biology, 17 1461 1474. Olsrud M, Christensen TR (2011) Carbon partitioning in a wet and a semiwet subarctic mire ecosyst em based on in situ C 14 pulse labelling. Soil Biology & Biochemistry, 43 231 239. Osterkamp TE (2007) Characteristics of the recent warming of permafrost in Alaska. Journal of Geophysical Research Earth Surface, 112 Osterkamp TE, Jorgenson MT, Schuur EA G, Shur YL, Kanevskiy MZ, Vogel JG, Tumskoy VE (2009) Physical and e cological c hanges a ssociated with w arming p ermafrost and t hermokarst in i nterior Alaska. Permafrost and Periglacial Processes, 20 235 256. Osterkamp TE, Romanovsky VE (1999) Evidence for warming and thawing of discontinuous permafrost in Alaska. Permafrost and Periglacial Processes, 10 17 37. Parnell AC, Inger R, Bearhop S, Jackson AL (2010) Source p artitioning u sing s table i sotopes: Coping with t oo m uch v ariation. Plos One, 5 Peterjohn WT, Melillo JM, Bowles FP, Steudler PA (1993) Soil warming and trace gas fluxes experimental design and preliminary flux results. Oecologia, 93 18 24. Phillips DL, Gregg JW (2003) Source partitioning using stable isotopes: coping with too many sources. Oecologia, 136 261 269. Ping CL, Michaelson GJ, Kane ES, Packee EC, Stiles CA, Swanson DK, Zaman ND (2010) Carbon Stores and Biogeochemical Properties of Soils under Black Spruce Forest, Alaska. Soil Science Society of America Jour nal, 74 969 978. Pinheiro J, Bates D, Debroy A, Sarcar D, R Development Core Team (2012) nlme: Linear and Nonlinear Mixed Effects Models R package V ersion 3.1 103. R Development Core Team (2012) R: A language and environment for statistical computing., R Foundation for Statistical Computing, Vienna, Austria. Raich JW, Russell AE, Kitayama K, Parton WJ, Vitousek PM (2006) Temperature influences carbon accumulation in moist tropical forests. Ecology, 87 76 87.

PAGE 200

200 Raich JW, Schlesinger WH (1992) The global c arbon dioxide flux in soil respiration and its relationship to vegetation and climate. Tellus Series B Chemical and Physical Meteorology, 44 81 99. Ruehr NK, Buchmann N (2010) Soil respiration fluxes in a temperate mixed forest: seasonality and temperatur e sensitivities differ among microbial and root rhizosphere respiration. Tree Physiology, 30 165 176. Rustad LE, Campbell JL, Marion GM et al. (2001) A meta analysis of the response of soil respiration, net nitrogen mineralization, and aboveground plant g rowth to experimental ecosystem warming. Oecologia, 126 543 562. Rustad LE, Fernandez IJ (1998) Experimental soil warming effects on CO2 and CH4 flux from a low elevation spruce fir forest soil in Maine, USA. Global Change Biology, 4 597 605. Ryan MG, Me lillo JM, Ricca A (1990) A comparison of methods for determining proximate carbon fractions of forest litter. Canadian Journal of Forest Research Revue Canadienne De Recherche Forestiere, 20 166 171. Savage KE, Davidson EA (2003) A comparison of manual and automated systems for soil co2 flux measurements: Trade offs between spatial and temporal resolution. Journal of Experimental Botany, 54 891 899. Schaefer K, Zhang T, Bruhwiler L, Barrett AP (2011) Amount and t iming of p ermafrost c arbon r elease in r esponse to c limate w arming. Tellus B, 63 165 180 Schielzeth H (2010) Simple means to improve the interpretability of regression coefficients. Methods in Ecology and Evolution, 1 103 113. Schimel JP, Gulledge JM, Clein Curley JS, Lind strom JE, Braddock JF (1999) Moisture effects on microbial activity and community structure in decomposing birch litter in the Alaskan taiga. Soil Biology & Biochemistry, 31 831 838. Schuur EAG, Abbott B (2011) Climate change: High risk of permafrost thaw Nature 480 32 33. Schuur EAG, Bockheim J, Canadell JG et al. (2008) Vulnerability of permafrost carbon to climate change: Implications for the global carbon cycle. Bioscience, 58 701 714. Schuur EAG, Crummer KG, Vogel JG, Mack MC (2007) Plant species composition and productivity following permafrost thaw and thermokarst in alaskan tundra. Ecosystems, 10 280 292. Schuur EAG, Trumbore SE (2006) Partitioning sources of soil respiration in boreal black spruce forest using radiocarbon. Global Change Biology, 12 165 176.

PAGE 201

201 Schuur EAG, Vogel JG, Crummer KG, Lee H, Sickman JO, Osterkamp TE (2009) The effect of permafrost thaw on old carbon release and net carbon exchange from tundra. Nature, 459 556 559. Shaver GR, Billings WD, Chapin FS, Giblin AE, Nade lhoffer KJ, Oechel WC, Rastetter EB (1992) Global change and the carbon balance of arctic ecosystems. Bioscience, 42 433 441. Shaver GR, Bret Harte SM, Jones MH, Johnstone J, Gough L, Laundre J, Chapin FS (2001) Species composition interacts with fertiliz er to control long term change in tundra productivity. Ecology, 82 3163 3181. Shaver GR, Chapin FS (1991) Production biomass relationships and element cycling in contrasting arctic vegetation types. Ecological Monographs, 61 1 31. Shaver GR, Johnson LC Cades DH et al. (1998) Biomass and CO2 flux in wet sedge tundras: Responses to nutrients, temperature, and light. Ecological Monographs, 68 75 97. Sterner RW, Elser JJ (2002) Ecological Stoichiometry, Princeton University Press, Princeton, New Jersey. S tuiver M, Reimer P (2010) Calib 14C Calibration Program. Sturm M, Racine C, Tape K (2001) Climate change Increasing shrub abundance in the Arctic. Nature, 411 546 547. Subke JA, Voke NR, Leronni V, Garnett MH, Ineson P (2011) Dynamics and pathways of au totrophic and heterotrophic soil CO(2) efflux revealed by forest girdling. Journal of Ecology, 99 186 193. Tarnocai C, Canadell JG, Schuur EAG, Kuhry P, Mazhitova G, Zimov S (2009) Soil organic carbon pools in the northern circumpolar permafrost region. G lobal Biogeochemical Cycles, 23 11. Trucco C, Schuur EAG, Natali SM, Belshe EF, Bracho R, Vogel J (2012) Seven year trends of CO2 exchange in a tundra ecosystem affected by long term permafrost thaw. Journal of Geophysical Research Biogeosciences, 117 Tr umbore S (2000) Age of soil organic matter and soil respiration: Radiocarbon constraints on belowground C dynamics. Ecological Applications, 10 399 411. Trumbore SE (1997) Potential responses of soil organic carbon to global environmental change. Proceedi ngs of the National Academy of Sciences of the United States of America, 94 8284 8291.

PAGE 202

202 Trumbore SE, Bubier JL, Harden JW, Crill PM (1999) Carbon cycling in boreal wetlands: A comparison of three approaches. Journal of Geophysical Research Atmospheres, 104 27673 27682. Trumbore SE, Czimczik CI (2008) Geology An uncertain future for soil carbon. Science, 321 1455 1456. Trumbore SE, Harden JW (1997) Accumulation and turnover of carbon in organic and mineral soils of the BOREAS northern study area. Journal of Geophysical Research Atmospheres, 102 28817 28830. Turetsky MR, Wieder RK, Vitt DH, Evans RJ, Scott KD (2007) The disappearance of relict permafrost in boreal north America: Effects on peatland carbon storage and fluxes. Global Change Biology, 13 192 2 1934. Vitousek PM, Turner DR, Parton WJ, Sanford RL (1994) Litter decomposition on the Mauna Loa environmental matrix, Hawaii patterns, mechanisms, and models. Ecology, 75 418 429. Vogel J, Schuur EAG, Trucco C, Lee H (2009) Response of CO2 exchange in a tussock tundra ecosystem to permafrost thaw and thermokarst development. Journal of Geophysical Research Biogeosciences, 114 14. Vogel JS, Nelson DE, Southon JR (1987) C 14 background levels in an accelerator mass spectrometry system. Radiocarbon, 29 323 333. Von Deimling TS, Meinshausen M, Levermann A, Huber V, Frieler K, Lawrence DM, Brovkin V (2012) Estimating the near surface permafrost carbon feedback on global warming. Biogeosciences, 9 649 665. Wagner D, Jones JB (2006) The impact of harvester ants on decomposition, n mineralization, litter quality, and the availability of N to plants in the Mojave desert. Soil Biology & Biochemistry, 38 2593 2601. Wickland KP, Striegl RG, Neff JC, Sachs T (2006) Effe cts of permafrost melting on CO2 and CH4 exchange of a poorly drained black spruce lowland. Journal of Geophysical Research Biogeosciences, 111 13. Wu ZT, Dijkstra P, Koch GW, Penuelas J, Hungate BA (2011) Responses of terrestrial ecosystems to temperatur e and precipitation change: a meta analysis of experimental manipulation. Global Change Biology, 17 927 942. Yi ZG, Fu SL, Yi WM et al. (2007) Partitioning soil respiration of subtropical forests with different successional stages in south China. Forest E cology and Management, 243 178 186.

PAGE 203

203 Zhou XH, Sherry RA, An Y, Wallace LL, Luo YQ (2006) Main and interactive effects of warming, clipping, and doubled precipitation on soil CO2 efflux in a grassland ecosystem. Global Biogeochemical Cycles, 20 12. Zimov S A, Davydov SP, Zimova GM, Davydova AI, Schuur EAG, Dutta K, Chapin FS (2006) Permafrost carbon: Stock and decomposability of a globally significant carbon pool. Geophysical Research Letters, 33 5. Zona D, Lipson DA, Paw KT, Oberbauer SF, Olivas P, Gioli B Oechel WC (2012) Increased CO2 loss from vegetated drained lake tundra ecosystems due to flooding. Global Biogeochemical Cycles, 26 16. Zuur AF, Ieno EN, Walker N, Saveliev AA, Smith GM (2009) Mixed Effects Models and Extensions in Ecology with R, Sprin ger.

PAGE 204

204 BIOGRAPHICAL SKETCH Caitlin Hicks Pries was born in Framingham, Massachusetts and grew up in Western Massachusetts. She spent her summer vacations on Cape Cod where she She attended Middlebury College in Vermont where she enjoyed views of the Green Mountains and Adirondacks every day. She pursued a joint major in biology and environmental studies, graduating in May 2004. During college, she spent a challenging but immen sely rewarding semester at the Ecosystem Center of the Woods Hole Marine Biology Laboratory where she came to appreciate the cycling of elements within ecosystems. Upon graduation, Caitlin worked as a research assistant for Dr. Steward Pickett at the Insti tute of Ecosystem Studies in Millbrook, New York. While at that job, she ran a laboratory, worked on GIS projects, and enjoyed arduous field work in Kruger National Park, South Africa. In August 2005, seeking a change from the northeast climate, biota, and culture she had grown up with, Caitlin moved to Gainesville, Florida to begin her master s work with Dr. Ramesh l and Water Science Department. For her master s work, Caitlin was given the freedom to pick her own project, so she chose to work in a coastal environment, like the ones she loved as a child, and on carbon, her favorite element. After completing her master s, Caitlin continued at the University of Florida working on a dissertation with Dr. Ted Schuur in the Biology Department She switched field sites from the flat, hot, buggy coast of Florida to the mountainous, cold, buggy interior of Alaska. She continued her study of the carbon cycle. During her dissertation, she spent three summers in Healy, Alaska waking up at 5 am to sample tundra respiration with Dale, her loyal hound mix, as her sole field companion. During the fourth summer, she

PAGE 205

205 went back to Healy but only after travelling abroad to Abisko, Sweden to collaborate on a related field project While working on her dissertation, Caitlin married Alexander Pries, a fellow master s graduate of the University of Florida and an all around hardworking, kindhearted Midwestern guy. After graduation in December 2012, Caitlin, Alex, Dale, and their cat, Posey, will embark on a new adventure together. Caitlin has a postdoctoral position at the Lawrence Berkeley Laboratory that will take them across the country to Northern California.