Download High latitude terrestrial ecosystems are considered key components in the... carbon (C) cycle and hold large reservoirs of soil organic... Abstract

Abstract
High latitude terrestrial ecosystems are considered key components in the global
carbon (C) cycle and hold large reservoirs of soil organic carbon (SOC). To a large
degree, this SOC is stored in permafrost soils and peatlands and is vulnerable to
remobilization under future global warming and permafrost thawing. Recent studies
estimate that soils in permafrost regions store SOC equivalent to ~ 1.5 times the
global atmospheric C pool. Ecosystems and soils interact with the atmospheric C
pool; photosynthesis sequesters CO2 into SOC whereas microbial decomposition
releases C based trace gases (mainly CO2 and CH4). Because of the radiative
greenhouse properties of these gases, soil processes also feedback on the global
climate system. Recent studies report increases in permafrost temperatures and under
future climate change scenarios permafrost environments stand to undergo further
changes. As permafrost thaws and surface hydrology changes, there is concern that
periglacial tundra and peatland ecosystems will switch from being sinks for
atmospheric C into sources, creating a potential for positive feedbacks on global
warming. The magnitude of change in C fluxes resulting from climate warming and
permafrost thawing depends on the remobilization processes affecting SOC stores, the
size of SOC stores that become available for remobilization and the lability of the
SOM compounds in these stores. While the large size and potential vulnerability of
arctic SOC reservoirs is recognized, detailed knowledge on the landscape partitioning
and quality of this SOC is poor.
Paper I of this thesis assesses landscape allocation and environmental gradients in
SOC storage in the Usa River Basin lowlands of northeastern European Russia. The
Russian study area ranges from taiga region with isolated permafrost patches to tundra
region with nearly continuous permafrost. Paper II of this thesis investigates total
storage, landscape partitioning and quality of soil organic carbon (SOC) in the tundra
and continuous permafrost terrain of the Tulemalu Lake area in the Central Canadian
Arctic. Databases on soil properties, permafrost, vegetation and modeled climate are
compiled and analyzed. Mean SOC storage in the two study regions is 38.3 kg C m-2
for the Usa River Basin and 33.8 kg C m-2 for Tulemalu Lake (for 1m depth in mineral
soils and total depth of peat deposits). Both estimates are higher than previous
estimates for the same study areas. Multivariate gradient analyses from the Usa Basin
show that local vegetation and permafrost are strong predictors of soil chemical
properties, overshadowing the effect of climate variables. The results highlight the
importance of peatlands, particularly bogs, in bulk SOC storage in all types of
permafrost terrain. In the Tulemalu Lake area significant amounts of SOC is stored in
cryoturbated soil horizons with C/N ratios indicating a relatively low degree of
decomposition. As this pool of cryoturbated SOC is mainly stored in the active layer,
no dramatic increases in remobilization are expected following a deepening of the
active layer. However, recent studies have demonstrated the importance of SOC
storage in deep (>1m) cryoturbated horizons. Perennially frozen peat deposits in
permafrost bogs constitute the main vulnerable SOC pool in the investigated regions.
Remobilization of this frozen C can occur through gradual but widespread deepening
of the active layer with subsequent talik formation, or through more rapid but
localized thermokarst erosion.
1
This thesis consists of a summary and two papers. The papers are referred to by
roman numerals.
Paper I:
Hugelius, G. and Kuhry, P. Landscape partitioning and environmental gradient
analyses of soil organic carbon in a permafrost environment. Global Biogeochemical
Cycles (in press).
Paper II:
Hugelius, G., Kuhry, P., Tarnocai C. and Virtanen T. Total storage and landscape
distribution of soil organic carbon in the Central Canadian Arctic. Submitted to
Permafrost and Periglacial Processes.
Paper I is a meta-analysis of data collected by many researchers during previous
research projects. I compiled the data and did all statistical and geomatic analyses. I
did most of the writing and constructed figures and tables. Peter Kuhry participated in
interpretation of results and writing.
For Paper II, Peter Kuhry, Charles Tarnocai and Tarmo Virtanen did the field
sampling and data collecting. I did most of the chemical analyses and all data
processing, statistical analyses and geomatic analyses. I did most of the writing and
constructed figures and tables. Peter Kuhry, Charles Tarnocai and Tarmo Virtanen
participated in interpretation of results and writing.
2
SOIL ORGANIC CARBON IN PERMAFROST TERRAIN
Total storage, landscape distribution and environmental controls
A wider perspective on soil organic carbon
The origins of soil organic matter
Soil organic matter (SOM) is defined as the mixture of recognizable plant and animal
detritus and material that has been altered so that it no longer contains its original
structure (Oades 1989, in this thesis also including SOM in peat deposits and lake
sediments). SOM is rich in carbon (C) and accumulates over time as organic detritus
from biota is incorporated into mineral soils and peat deposits. The main source of
organic detritus is from photosynthetic plants sequestering atmospheric carbon
dioxide (CO2). However, decomposition of organic material by soil microbes (soil
respiration) returns C to the atmosphere, mainly as CO2 or methane (CH4). These dual
fluxes of C maintain a turnover, or cycling, of soil organic carbon (SOC) pools.
The formation of SOM attracted the attention of 19th century naturalist Charles
Darwin whose last work (Darwin 1881) described the role of earth worms in SOM
formation. Meanwhile, contemporary Russian geographer Vasily Dokuchaev was
developing the ideas that laid the foundations of modern soil science (Dokuchaev
1898, in Jenny 1980). During extensive travels throughout the Eurasian continent,
Dokuchaev made observations on the linkages between biota, climate, topography and
soil. Hans Jenny (1930, 1941) expanded on Dokuchaev’s work and did pioneering
efforts in describing pedogenic processes and the formation of SOM along
environmental gradients. He formulated semi-quantitative equations of soil forming
processes and confirmed links between soil carbon inventories and pedogenic factors,
i.e. parent material, climate, biota, topography, hydrology, human influence and time
(Jenny 1980).
Soils in the global C cycle
Today, the integral role of soils in the global C cycle is well recognized (figure 1).
The very first attempt at quantification of the global SOC pool was made by
Waksman in 1938 (estimated 400 Pg global pool, Pg = g*1015). Schlesinger (1977)
made the first comprehensive quantification based on global ecosystem distributions,
thereby producing a benchmark study. He presented separate estimates for a range of
ecosystem types (1456 Pg total pool, 173 Pg in “Tundra and Alpine”, 137 Pg in global
“Swamp and Marsh”) and established the importance of soils in the global C cycle.
Post et al. (1982) made estimates based on world life zones, providing a global
estimate (1395 Pg) and estimates for the different life zones (192 Pg in “Tundra”, 202
Pg in “Global Wetlands”). Bohn (1982) estimated global storage from the FAO Soil
Map of the World to be 2200 Pg (with 400 Pg in peatlands). Most recent global
estimates converge around 1500-1600 Pg for the top meter of global soils, with an
additional 500-900 Pg of SOC stored in the second meter (Eswaran et al. 1993, Batjes
1996, Jobbagy and Jacksson 2000, Kirschbaum 2000, Amundson 2001).
3
Soil Organic Carbon in northern environments
Gorham’s (1991) review on northern peatlands highlighted the importance of northern
peat deposits in C cycling and storage (an estimated 455 Pg in peatlands of boreal
forests and tundra regions,). Based on the Northern Circumpolar Soil Carbon
Database (NCSCD, Tarnocai et al. 2007), Tarnocai and Broll (2008) estimated the
SOC storage in northern permafrost regions to be 496 Pg, of which 238 Pg is stored in
peatlands (calculated to 1 m depth in mineral soils and to full depth in peat deposits).
Schuur et al. (2008) included deeper mineral deposits (to 3 m depth) and estimated
SOC storage in circum-arctic permafrost regions to be 1024 Pg, with peatlands
contributing 277 Pg.
Results from these and other studies focusing on SOC storage in permafrost terrain
(see also Zimov et al., 2006; Ping et al., 2008; Tarnocai et al., in press), suggest that
frequently cited estimates of global SOC stores (Schlesinger 1977 and 1997, Post et
al. 1982, Jobbagy and Jackson 2000) may underestimate storage in subarctic and
arctic ecosystems by a factor 2 or more to 1 m depth and a factor ~3.5 to 3 m depth.
The lower global estimates may be caused by an underestimation of storage in
cryoturbated horizons of permafrost soils and peatlands (see e.g. Kuhry et al. 2002,
Bockheim 2007, Tarnocai and Broll 2008, Ping et al. 2008, Papers I and II of this
thesis). However, comparison between different studies is difficult as terminology and
definitions differ, as do the definitions and sizes of the upscaling regions. The
northern permafrost regions (18.7 * 106 km2 of land outlined by Brown et al., 1997)
used in the studies of Schuur et al. (2008) and Tarnocai and Broll (2008) include
alpine and tundra ecosystems (~8.8 * 106 km2 in global estimates), as well as regions
of boreal forest in North America and Siberia (boreal forest cover ~12 * 106 km2 in
global estimates).
Northern soils in a changing environment
From the 19th century, following the end of the Little Ice Age in the Arctic,
temperatures in arctic regions have increased significantly (Overpeck et al. 1997).
Since the start of the industrial revolution the global average temperatures have risen
by ∼0,6oC (ACIA 2004). In the last few decades average temperature increases in the
Arctic have been near twice as high as mean global increases (ACIA 2004). Under
many future climate change scenarios, permafrost environments stand to undergo
massive changes (Stendel and Christensen 2002). The IPCC predicts increases above
global averages in arctic mean temperature and winter precipitation (Christensen et al.
2007), both key factors regulating permafrost distribution. Recent warming of
permafrost has been reported (Lemke et al. 2007). In Alaska and Siberia, permafrost
thaw is changing the size and distribution patterns of thermokarst lakes (Yoshikawa
and Hinzman 2003, Smith et al. 2005).
Upland soils and peatlands in permafrost regions hold a large SOC pool. Northern
peatlands have been expanding throughout the Holocene, sequestering C and acting as
a sink for atmospheric C (Smith et al. 2004). However, as permafrost thaws and
surface hydrology changes, there is concern that periglacial ecosystems will switch
from being sinks for atmospheric C into sources, creating a potential for positive
feedbacks on global warming (Oechel et al. 1993, Heikkinen et al. 2004). The surface
hydrology of peatlands is a key factor regulating the exchange of C greenhouse gases,
4
as it affects soil respiration. Respiration by aerobic microbes in aerated soils and peat
deposits primarily leads to CO2 emissions, while CH4 is the main by-product of
anaerobic decay in waterlogged peat deposits or thermokarst lakes (Christensen et al.
2008). While most northern peatlands sequester CO2, they are an important source of
CH4, contributing 3 to 5 % of global CH4 emissions (Frolking et al. 2006). On a 100
year timeline, the greenhouse warming potential of CH4 is 25 times that of CO2
(Forster et al. 2007). As a consequence, on shorter timescales, a peatland can maintain
net sequestration of CO2-C through photosynthesis (leading to sustained peat
accumulation), while sustained CH4 emissions (from anaerobic decay of SOM) lead to
a positive net radiative forcing. However, because of the relatively short mean
residence time (MRT) of CH4 in the atmosphere, the long term effect over millennial
timescales is a negative net radiative forcing (Frolking et al. 2006).
Gruber et al. (2004) identified permafrost soils and peatlands to be key C pools
(together with oceans and forests), vulnerable to remobilization through permafrost
thaw and changed surface hydrological conditions (figure 1).
Figure 1. Photograph on the left shows thermokarst developing in thawing palsa, Tavvavuoma, NorthWestern Sweden. Photograph on the right shows close-up of segregated ice inside the eroding palsa.
Cryosuction has freeze-dried the fen peat surrounding the ice (pen added for scale).
An estimated half of the global SOC pool consists of refractory compounds. They
account for a disproportionate amount of long-term terrestrial carbon storage due to
their long MRT in soils (Amundson 2001). But general knowledge concerning the
absolute and relative sizes of labile and recalcitrant SOC pools in the Arctic is poor.
The intrinsic chemical lability of SOM compounds varies with botanical origin
(Kuhry et al. 1996). Other environmental factors that affect the MRT of SOC include
protection within soil aggregates or through electro-chemical bonds, drought, flooding
and subzero temperatures (Davidson and Janssens 2006). While the latter is an
obvious rationale for focusing on permafrost soils, the other factors are also
susceptible to change if permafrost thaw alters the physical environment. A rapid
response to environmental changes can be expected from labile SOC compounds, and
improved knowledge of these SOC pools is crucial for predicting transient responses
of high latitude soils in this century.
There are many potential feedback mechanisms on global warming that may affect C
fluxes between high latitude soils and the atmosphere. The magnitude of change in C
5
fluxes resulting from global warming and permafrost thawing depends on: (i) the
nature, rate and extent of remobilization processes affecting SOC stores, (ii) the size
of SOC stores that become available for remobilization and (iii) the lability of the
SOM compounds in these stores. Improved process understanding is a research
challenge that is currently being met by a diverse scientific community.
Objective of PhD-project
Although there is a scientific consensus that arctic and sub-arctic ecosystems harbour
large reservoirs of SOC, detailed knowledge concerning the exact amounts,
distribution and quality of SOC is lacking. The main objective of my PhD-studies is
to: (1) assess the quantity, distribution and quality of SOC in permafrost terrain. Other
objectives are to: (2) assess the quality and susceptibility to decay of SOM in
permafrost and evaluate the possibility of using low-cost methods for this purpose
(C/N ratios and peat colorimetry); (3) improve the understanding of environmental
controls on SOC storage and quality through multivariate analyses; (4) further
develop current geomatic methodology for soil sampling and upscaling results at
local-regional scales; (5) a final goal of the PhD-project is to provide a synthesis
comparing results from different types of permafrost terrain (from sporadic to
continuous) in different climatic and geographical settings (continental to maritime,
lowland to alpine).
The main study area is the Usa River Basin, North Eastern European Russia. A
regional analysis of this area is presented in Paper I. In addition, detailed landscape
level inventories of SOC were conducted in the Usa Basin and in more southerly taiga
areas in 2007-2008. Results from these studies will form the basis for future work.
Additional study areas are Lake Tulemalu in Central Canada (Paper II), Tavvavuoma
in northern Sweden (fieldwork 2007-2009) and Zackenberg in North Eastern
Greenland (fieldwork 2009). Figure 2 shows the location of all study areas on a
circum-arctic map.
Figure 2. Circum-arctic map showing the locations of present and future study areas.
6
Upscaling methodology and statistical analyses
Global studies of SOC storage
Attempts to estimate global SOC stores through the link to vegetation have typically
used an approach where the arithmetic mean of SOC storage from a number of sites
within a biome or life zone are multiplied by the respective region's areal coverage.
Most global estimates of SOC suffer from a poor sample representation in large and
remote high-latitude areas. The combined "Tundra and Alpine Ecosystems of the
World" (~ 8-8.8 * 106 km2) are represented by 21, 48 and 51 sites in the studies by
Schlesinger (1977), Post et al. (1982) and Jobbagy and Jackson (2000), respectively.
The main challenges to overcome are availability of reliable global SOC data and
finding a reliable upscaling proxy with global coverage. The choice of upscaling
proxies is essentially limited to satellite land cover classifications (LCC:s) or soil
maps. A difficulty with global LCC:s is development of a classification system that
works across the diverse regions of the Earth. Maintaining thematic accuracy and
translating different thematic classes across regions is a challenge, but there are
promising products and initiatives that could be useful in SOC upscaling (GLC 2000
and 2002, NASA 2007). There are quite extensive global soil databases available for
use in upscaling, ordered according to international standardized soil classification
schemes (FAO 1995, Batjes 2002). However, they suffer some drawbacks that may
limit their usability for more detailed studies of SOC storage. Many pedons lack
values for soil bulk density and C % content, variables that are needed to calculate
SOC storage.
Local and regional SOC studies in the Arctic
Local or regional scale studies of SOC storage in remote arctic areas have typically
used an upscaling approach similar to global studies, but with greater spatial and
thematic resolution (e.g. Michaelson et al. 1996, Tarnocai and Lacelle 1996, Kuhry et
al. 2002, Mazhitova et al. 2003). Estimates of total storage in the investigated
landscape are based on an upscaling methodology where an empirical connection
between nominal vegetation or soil classes and SOC storage is established through in
situ sampling of soils. Areal coverage is subsequently obtained through a satellite
LCC or regional soil map. The estimated mean SOC storage in each land cover or soil
class is multiplied by the areal coverage of that class.
As this upscaling method is based on arithmetic means of sites the standard error of
the mean (SE= StD/√n), is a logical statistical measure of accuracy. The best way to
directly affect SE is to increase site density (number of sites per upscaling unit) of the
estimate. But to get an accurate result, the site selection must provide a correct
representation of the landscape that is being upscaled. A narrow site selection (e.g.
insufficient representation of deep peatlands or cryoturbated soils) may not fully
represent the variety present in the landscape. This leads to a low StD in the upscaling
dataset, causing an illusion of upscaling accuracy as the SE remains low.
There are a number of inherent assumptions that you accept when you upscale by
multiplying the mean SOC storage of a class with spatial coverage. You assume that
the selected upscaling classes (land cover or soil classes) accurately depict the diverse
7
natural environment of the landscape you are describing and that all upscaling classes
are separate statistical populations that can be accurately described through empirical
studies. Any gradients or ecotones within the upscaling classes are ignored (or more
accurately they are described by a large SE). There are some ways to approach the
fulfillment of above mentioned criteria by simple statistical tests. The data within each
class should be (log)normally distributed and unimodal. Furthermore, data from all
upscaling classes can be compared in student t-test to see whether they do form
separate populations.
One of the challenges with estimating arctic soil C pools is the construction of a
sampling program that correctly describes the environment. The arctic landscape
displays variation on many different spatial scales, and different methods for
calculating landscape C pools demand varying approaches to sampling. In field
studies we have used either a transect based sampling method, or a weighted stratified
random sampling approach. The transect approach involves a subjective selection of
'representative' transect locations following field reconnaissance. Once transects are
established, soil samples are collected equidistantly without further subjective bias.
This sampling scheme combines selective representation of what is considered
representative in the landscape with a measure of randomization introduced by smallscale vegetation and micro-topography patterns. In stratified random sampling
program, the study area is subdivided a priori, based on a preliminary LCC, and a
weighted number of sampling sites are randomly distributed within each class. The
weighting of sites is based on the expected variability in SOC storage within the class
(data from previous studies).While the latter is a statistically more sound approach
(e.g. Simbahan and Dobermann 2006), it is labour intensive and consumes more field
time (mainly because of longer distances between sites).
Multivariate statistical analyses
Multivariate gradient analysis techniques, such as Principal Component Analysis
(PCA), allow integrated analysis of multiple variables on different scales of
measurement (Jongman et al. 1995, Kent 2006). In Papers I and II, we use PCA to
investigate connections between site specific soil properties and a range of
environmental variables (vegetation, climate and permafrost). In Paper I, constrained
gradient analyses (Redundancy Analysis, RDA) is combined with Monte Carlo
permutations to model connections between soil properties, permafrost, vegetation
and climate.
PCA is an unconstrained ordination technique commonly used for reducing
dimensionality and extracting patterns from multivariate datasets with linear
responses to gradients of change. PCA calculates theoretical variables (termed
Principal Components, PCs) that minimize the residual sums of squares between all
observations in a multidimensional matrix and the PC axes. All PCs are orthogonal in
multidimensional space (i.e. completely uncorrelated), maximizing the amount of
information that is retained in each PC.
In earth-science and ecology there are several direct gradient analysis methods in use,
these are generally ordination based techniques that are expanded to include the
influence of external independent variables, or environmental gradients (Wagner
2004, Kent 2006). RDA is an extension of PCA where the information contained in an
explanatory dataset of environmental variables is used to constrain the response
8
variable dataset. For each site, it only extracts the information in the soil response
variables that can be explained through variations in the environmental variables. In
RDA the ordination axes are derived from PCA and the technique assumes a linear
response model (as opposed to a Gaussian response model used in Canonical
Correspondence Analysis). The intended use of RDA analysis is to relate the
abundance of a set of species (dependent variables, Dataset 1) to data describing a
suite of environmental variables for the sampling sites (independent variables, Dataset
2).
In an extension of RDA, Monte Carlo permutations are used to evaluate the null
hypothesis that soil response variables are unaffected by environmental variables (Ter
Braak and Smilauer 2002). In what is called a “Forward Selection of variables”, the
environmental variable with the highest explanatory power is tested first (and
included if the p-value is accepted) after which the explanatory power of the
remaining variables are recalculated, factoring in the variance already accounted for
by the included variable. The process is used to avoid over-fitting the RDA model
with redundant environmental variables.
A spatial matrix based on geographical locations of sampling sites can be used to
partial out spatial autocorrelation and ecoclimatic location from the dataset, allowing
more direct analyses of the impact of other environmental variables (implemented in
Paper I). A matrix is constructed by adding all terms for a cubic trend surface
regression (Borcard et al. 1992). To avoid over fitting of the model, stepwise Forward
Selection of variables is used to identify the variables with highest explanatory power.
Multivariate ordination techniques, coupled with Monte Carlo permutations and covariable matrixes have proved to be good tools for investigating complex relationships
between soil properties and the environment. Future work will include modeling of
spatial structures and autocorrelation in SOM distribution through variance
partitioning (Peres-Neto et al. 2006) and semivariogram modeling (Wagner 2003,
Wagner and Fortin 2005).
9
Summary of the papers
Paper I
Hugelius, G. and Kuhry, P. Landscape partitioning and environmental gradient
analyses of soil organic carbon in a permafrost environment. Global Biogeochemical
Cycles (in press).
In this study we describe the landscape distribution, quantity and quality of SOM
(including SOC storage in upland soils, lake sediments and peat deposits) across the
lowland taiga-tundra continuum of the Usa River Basin, northeastern European Russia
(93 500 km2, figure 3). The study area ranges from taiga with isolated patches of
permafrost in the South, through the taiga-tundra transition zone with discontinuous
permafrost to tundra and continuous permafrost in the North.
Figure 3. The large map shows main vegetation patterns and boundaries of the regions defined for
upscaling: 1. taiga region 2. (forest)-tundra region, and 3. mountain region. No vegetation data is
shown for the mountain region. The inset map shows the location of the Usa River Basin and the
Pechora and Ob rivers (west and east of the basin, respectively). Sites used for ordination and upscaling
analyses are shown (an additional 23 taiga sites and 87 (forest)-tundra sites are lacking exact
coordinates and are not displayed). Projection UTM 40 N (WGS 84).
We compile and analyse databases describing soil chemical and physical properties,
permafrost conditions and vegetation for 325 separate sites across the taiga-tundra
continuum. Previous estimates of landscape SOC storage in the Usa Basin (see Kuhry
et al. 2002) are revised and updated with new calculations and addition of new sites.
We describe the partitioning of SOC in 24 different vegetation types, also including
estimates of SOC storage in permafrost soils (above and below the active layer
boundary). A subset of the database is analyzed in multivariate gradient analyses to
determine how an array of environmental variables are interconnected and linked to
site specific SOM quantity and quality. Constrained gradient analyses with Monte
10
Carlo permutations are used to model connections between SOM, permafrost,
vegetation and climate patterns.
Mean SOC storage in taiga and (forest)-tundra regions are estimated at 37.3 and 38.7
kg C m-2, respectively. Peatlands account for 72 % of SOC storage with only 30 % of
the surface area (figure 4). Permafrost within the upper meter (signifying Cryosols or
cryic Histosols) is found in 13 % of investigated sites. These permafrost soils store 42
% of the total SOC pool, of which nearly three quarters is perennially frozen below
the active layer. Peatlands hold 95 % of all SOC in permafrost terrain and 98 % of all
perennially frozen SOC, mainly in bog peatlands of the (forest)-tundra region.
Multivariate gradient analyses further emphasize the role of bog peatlands in SOC
storage. SOM stored in permafrost has higher C/N ratios than unfrozen material,
indicating that this material is more labile and susceptible to decay if thawed.
Figure 4. Graph shows the estimated percentage of total land cover and total soil C storage (bars) for
different land cover types in the taiga and (forest)-tundra regions. The bars are subdivided to show
storage in non permafrost soils as well as storage above and below active layers in permafrost soils.
Some of the land cover types used for upscaling have been amalgamated into groups.
Gradient analyses of climatic patterns show that at this regional scale, site permafrost
is equally affected by temperature and precipitation variables. Site vegetation and
permafrost are strong predictors of soil chemical properties, also when regional
ecoclimatic patterns are accounted for (through introduction of a spatial co-variable
matrix). While site permafrost loses its predictive power when the vegetation covariable is used, vegetation remains a strong predictor despite introduction of covariables and explains a significant part of the variance in soil properties. Local
vegetation and permafrost conditions overshadow the effect of climate variables on
soil properties. Introducing co-variables to mask out the local conditions shows that a
11
combination of mean annual air temperature and annual precipitation is a strong
predictor of site specific soil properties.
The results from this study highlight the importance of permafrost bogs as stores of
large amounts of labile C. It is important that the representation of these SOC hotspots
in global C estimates and models is as accurate as possible. Permafrost bogs are
common throughout the sub-arctic and arctic, often found in areas no longer
climatically optimal for permafrost formation (Christensen et al. 2004). Permafrost
thaw resulting in remobilization of this frozen SOC pool may occur through
deepening of the active layer or through thermokarst formation. Active layer
deepening is a gradual process with large areal extent that may eventually lead to talik
formation (a layer or body of perennially unfrozen ground occurring in permafrost
terrain). Thermokarst formation is more localized but may occur rapidly, generally
leading to altered surface hydrological conditions. While there is evidence for the
active occurrence of both active layer deepening and thermokarst (Lemke et al. 2007),
the relative and total magnitude of changes in C fluxes resulting from these processes
remain uncertain.
Paper II
Hugelius, G., Kuhry, P., Tarnocai C. and Virtanen T. Total storage and landscape
distribution of soil organic carbon in the Central Canadian Arctic. Submitted to
Permafrost and Periglacial Processes.
A substantial part of high latitude stores of SOC are found in the Canadian Arctic.
Tarnocai and Lacelle (1996) concluded that peatlands and permafrost soils contained
the largest total SOC masses in Canada, respectively storing 106 Pg and 103 Pg SOC.
Tarnocai (2006) subsequently updated the C storage estimate for Canadian peatlands
(1.1 million km2) to 147 Pg SOC and predicts that 60 % of Canada’s peatlands are
expected to be severely or extremely severely affected by projected climate change.
Permafrost temperatures in Canada have been rising since the mid-1990s (Brown et
al. 2000, Smith et al. 2003, Smith et al. 2005).
Only generalized soil C inventories are available for large parts of the Canadian
Arctic.
In this study we assess the total storage and spatial distribution of soil C in continuous
permafrost terrain of Central Canada. The study area is located on the shore of
Tulemalu Lake (62° 55´ N, 99° 10´ W, figure 5) in the Low Arctic zone, 100 km north
of the treeline. Mean annual temperatures in the area are between - 9.4 to -14.3 °C and
total annual precipitation is < 300 mm. The region is located in the Continuous
Permafrost Zone, with low to medium ice content. Deglaciation of the Laurentide ice
occurred between 9000-8000 cal yrs BP.
12
Figure 5. Overview map showing the location of the Tulemalu study area in relation to
hydrography, tree line and permafrost zonation in the Western Hudson Bay lowlands. Map also
shows the location of nearest climate stations and permafrost temperatures. Data on permafrost
zonation from Canada EcoATLAS (Government of Canada, 1998). Treeline and permafrost
temperatures are from Brown et al. (1997).
The landscape allocation of soil C is assessed using a transect-based soil sampling
program in a 21 km2 study area. We compare upscaling using transect vegetation
inventories to full landscape upscaling using land cover data derived from Landsat 7
ETM+ satellite imagery. We analyse the partitioning of soil C in surface organic
deposits and peat deposits, deeper cryoturbated organic pockets and the mineral
subsoil, also separating active layer from permafrost storage. The quality of soil
organic compounds is assessed using C/N weight ratios. Basal peat and cryoturbated
pockets of organic matter have been AMS 14C dated to describe the Holocene
landscape development of the study area.
Mean SOC storage in the study area for 0-100 cm depth is estimated to be 33.8 kg C
m-2, which is much higher than previous estimates for this part of the Central
Canadian Arctic (14-24 kg C m-2). The oldest basal and fossil wood dates indicate that
peat deposits started to accumulate around ~6000 cal yrs BP, in a time when spruce
was present in the study area. Bog peatlands hold 39 % of the total SOC pool, 58 % of
organic deposits and 59 % of the perennially frozen SOC (figure 6). Fen peatlands
cover larger areas than bogs, but store relatively little SOC (17 % of total). There are
large differences in SOC storage between different upland tundra classes, following a
landscape moisture gradient.
Radiocarbon dating of cryoturbated soil horizons shows that cryoturbation processes
have been occurring in the study area since the Middle Holocene (~6500 cal yrs BP).
13
Figure 6. Graph shows percentage of total C storage, divided by land cover types in the study area.
Subdivision shows partitioning between top organic/peat layers, cryoturbated pockets and mineral
subsoil with active layer storage separated from permafrost storage. Crosses show percentage of land
cover in the study area (lakes excluded from calculation).
Cryosols developed on till parent material store significant amounts of SOC in buried
cryoturbated soil horizons, contributing to 17 % of the total storage for the 0-100 cm
depth interval (mostly in the active layer). Using default values from the literature to
include deeper (1–3 m) cryoturbated deposits and 30 cm of mineral subsoil underlying
peat deposits significantly increases the mean landscape storage to 52.9 kg C m-2.
Analyses of C/N ratios show that the organic matter in cryoturbated pockets is less
decomposed than the organic matter in the surrounding mineral materials.
14
Summary of conclusions
Mean SOC storage in the two study regions is 38.3 kg C m-2 for the Usa River Basin
and 33.8 kg C m-2 for Lake Tulemalu. Both estimates are higher than previous
estimates for the same study areas. The results highlight the importance of peatlands,
particularly bogs, in bulk SOC storage in all types of permafrost terrain. Peatlands
hold 72 % of SOC in the Usa Basin and 56 % of SOC in the Tulemalu Lake study
area. In all, 42 % of SOC in the Usa Basin is stored in permafrost soils, and 30 % of
all SOC is perennially frozen below the active layer. In the continuous permafrost
terrain of Lake Tulemalu all SOC is stored in permafrost soils, but the amount of
perennially frozen SOC is 35 %, similar to the Usa Basin. In the (forest)-tundra region
of the Usa Basin there is a high prevalence of very deep frozen peat deposits and 98 %
of the perennially frozen SOC is stored in bog peatlands. Peat deposits at Lake
Tulemalu are generally younger and shallower, leading to lower storage of SOC in
perennially frozen peat.
Cryoturbation is an important process for burial of C rich soil horizons in continuous
permafrost terrain in Canada (representing 17 % of total SOC storage, mainly in the
active layer), but seems of little importance in the Usa Basin lowlands. As a
consequence upland tundra soils in Lake Tulemalu have higher mean SOC storage.
Multivariate gradient analyses show that local vegetation and permafrost are strong
predictors of soil chemical properties, overshadowing the effect of climate variables.
In the Usa Basin SOM stored in permafrost has higher C/N ratios than unfrozen
material. Cryoturbated organic soil pockets in Lake Tulemalu have higher C/N ratios
than the surrounding mineral horizons.
Permafrost bogs constitute the main vulnerable SOC pool in the investigated regions.
Permafrost bogs are common throughout sub-arctic and arctic regions and it is
important to have an accurate representation of these SOC hotspots for global
estimates of C cycling. Remobilization of this frozen C can occur through gradual but
widespread deepening of the active layer with subsequent talik formation, or through
more rapid but localized thermokarst erosion. In the Lake Tulemalu area significant
amounts of SOC is also stored in cryoturbated soil horizons with C/N ratios indicating
a relatively low degree of decomposition. As this pool of cryoturbated SOC is mainly
stored in the active layer, we would not expect dramatic increases in remobilization
following a deepening of the active layer. However, recent studies have demonstrated
the importance of SOC storage in deep (>1m) cryoturbated, mostly perennially frozen,
horizons.
15
Ongoing work and future plans
The taiga-tundra continuum in northeastern European Russia
The study area of the CARBO-North research program (www.carbonorth.net) ranges
from typical (permafrost free) boreal taiga near Syktyvkar in the south to continuous
permafrost terrain in the north (figure 7). During two extended summer field
campaigns in 2007 and 2008 I have carried out soil sampling programs in all
designated study areas.
Figure 7. Overview of the CARBO-North study areas (1-4) in North-Eastern
European Russia. 1. The Seida intensive study area. 2. The Rogovaya river (three
study areas in transect across Forest-Tundra transition). 3. The Khosedayu study
area. 4. The taiga study areas, Laly and Sludka. (source: www.carbonorth.net).
Within the Usa River Basin we have five study areas. In the Seida intensive study area
we have sampled soils in an extensive program combining a transect approach to a
weighted stratified random sampling scheme. Along the Rogovaya River, three
separate study sites form a transect across the wide forest-tundra transition. The
Khosedayu study area, furthest to the west, is in a different geological setting from the
other areas. The results of the soil sampling programs and chemical analyses will be
upscaled using high resolution satellite LCC:s provided within the CARBO-North
project by colleagues from the University of Helsinki, Finland. The upscaling results
will be integrated with results from phytomass inventories (from University of
Helsinki) to increase understanding of C storage and cycling in the region.
A detailed inventory of thermokarst lakes in peat plateaus in Seida and at the
Rogovaya River was carried out (figure 8a). The results give information on lake
morphology and processes behind thermokarst formation. Radiocarbon dating of
selected soil and peat horizons and stratigraphic analyses helps in recreating the
Holocene development of the study area. From the Seida intensive study area
representative samples from all different vegetation types have been selected for more
detailed analyses. In cooperation with colleagues from the Department of Geology
and Geochemistry (Stockholm University) we have extracted the humic and fulvic
16
acid components from freeze-dried SOM. These extracts are analyzed for dissolved
organic C content, elemental and stable isotope content for C, N and hydrogen (H).
Using photo-spectrometry, absorption in different wavelengths is measured (400, 465,
600 and 665 nm). These combined analyses of humic and fulvic extracts will give us
in depth information on the lability of SOM compounds stored in different landscape
settings (Ikeya and Watanabe 2003). This will be combined with the landscape
estimates of SOC storage and partitioning to provide an overall view of potential for
SOC remobilization in the area.
Figure 8. Photograph to the left (a) shows an elevated peat plateau with eroded surface in Seida, there is
a steep erosion edge into adjacent thermokarst lake. Photograph to the right (b) shows stream draining
the pristine taiga catchment in the Laly study area.
We have also collected soil samples from boreal taiga catchments (figure 8b). The
taiga study areas are located far south of the southern permafrost border, and are
included as areas for comparison (analogue to future warming in current foresttundra). The results from the taiga SOC inventories will be upscaled to full areal
coverage and integrated with phytomass inventories performed by scientists from
University of Helsinki, Finland.
In two taiga peatlands we have done transect based inventories of peat depth and basal
peat ages (determined using 14C dating) to recreate late Holocene and current rates of
paludification.
Subarctic alpine tundra, Taavavuoma, NW Sweden
Fieldwork in Tavvavuoma has been carried out at four occasions, two autumn trips
and two winter trips. A third and final autumn trip is planned for September, 2009.
In Tavvavuoma we have a wide research focus, encompassing SOC storage and
partitioning, soil mapping, permafrost mapping and vegetation mapping. Several soil
pits have been sampled and described following FAO terminology (1998, figure 9a).
Nine complete thermokarst lake sediment cores have been collected during winter
sampling. With the help of Russian scientists from Fundamentproject, Moscow, we
have drilled deep boreholes for permafrost monitoring in two palsa complexes at
different altitudes (to 10 and 6 m depth at elevations of 450 and 750 m.a.s.l., figure
9b). We are exploring the possibilities of doing permafrost mapping in the area by
measuring the basal temperature of snow during late winter (King 1983, Heginbottom
2002). We are planning to conduct vegetation mapping at several spatial scales to
17
describe the effects of upscaling proxy resolution and accuracy (vegetation data from
Landsat 7 ETM+, IKONOS and digital aerial imagery from Lantmäteriet).
Figure 9. Photograph to the left (a) shows a soil pit excavated through sorted circle in alpine tundra of
the Tsåktso plateau, Tavvavuoma. Gradual movement of rocks through freeze-thaw cycles sorts rocks
(accumulation of stones in the left part of the pit). Photograph to the right (b) shows the high elevation
palsa mire where one of the deep permafrost boreholes has been drilled.
High Arctic tundra in alpine terrain, Zackenberg, NE Greenland
Zackenberg Research Station (74º28’ N, 20º34’ W) is located in continuous
permafrost terrain on north-eastern Greenland. The research station is located at the
bottom of a glacial valley in an alpine setting.
A ten day field campaign is planned to the Zackenberg research station in August,
2009. The goal is to carry out a distributed sampling program that can subsequently be
upscaled using available proxy data from the Zackenberg station (maps of vegetation,
soil and geomorphology). We will determine total storage and landscape distribution
of SOC. Detailed geochemical analyses and 14C dating will provide information on
the quality and age of stored SOM.
18
Acknowledgements
First and foremost I would like to thank my supervisor Peter Kuhry. You are always
interested and enthusiastic about the research and you are a great support.
Galina Mazhitova, you were a great help and companion in the field and you taught
me what I know about permafrost soils and the WRB-system. You passed away
before your time and you will be missed.
With the help of my co-supervisors Joyanto Routh and Patrick Crill I have gotten
some insight into the world of organic chemistry. It is, at times, very unfamiliar but it
is also interesting and a lot of fun! Maj-Liz Nordberg and Wolter Arnberg have taught
me geomatics at our department, and have provided great help as co-supervisors.
In addition to being nice colleagues, Steffen Holzkämper, Britta Sannel and Päivi
Kaislahti have also provided good company on travels and help in the field. I would
like to thank my Finnish colleagues, Tarmo, Sanna and Malin, for nice company and
cooperation during two long Russian summers. Nathalie Pluchon and Ylva Sjöberg
were great field assistants in Russia. Heike Siegmund is always helpful and has done
excellent work with C/N/H analyses. Charles Tarnocai has provided valuable
comments and insight on soils and cryoturbation in the Canadian Arctic. Simon
Wallgren has been an enthusiastic and resilient help during winter fieldwork in
Tavvavuoma. Lars Labba has provided invaluable support during fieldwork in
Tavvavuoma, both with logistics and good times!
I would like to thank all my colleagues at the Department of Physical Geography and
Quaternary Geology for providing a stimulating and pleasant work environment.
And, finally, a great big thank you to my family and friends, for your support and for
keeping me busy doing other interesting things, so that I don’t turn into a complete
SOC fanatic.
My PhD project is supported by the EU 6th Framework CARBO-North project
(contract 036993) and a grant from the Swedish Research council for the project
“Landscape patterns of soil organic matter quantity and lability in permafrost terrain”,
with additional funding from the Department of Physical Geography and Quaternary
Geology, Stockholm University. The Göran Gustafsson Foundation and the Swedish
Society for Anthropology and Geography kindly provide support for the work in
Tavvavuoma. The Bert Bolin Centre for Climate Change has kindly given me support
to attend conferences in both Fairbanks and in Vienna. The Mannerfelt foundation
provided support for travels and a very rewarding fieldtrip in Alaska. Thanks to The
Wallenberg Jubileumsfond for supporting my upcoming travels to Greenland.
19
References
ACIA, Arctic Climate Impact Assesment (2004) Impacts of a warming arctic. Cambridge University
Press, Cambridge, United Kingdom
Batjes, N.H. (1996) Total carbon and nitrogen in the soils of the world. European Journal of Soil
Science 47: 151-163
Batjes, N.H. (2002) A Homogenized Soil Profile Dataset for Global and Regional Environmental
Research (WISE, Version 1.1) Report 2002/01, International Soil Reference and Information Center
Report, Wageningen, Netherlands
Bockheim J.G. (2007), Importance of cryoturbation in redistributing organic carbon in permafrostaffected soils. Soil Science Society of America Journal, 71, 1335-1342.
Borcard D, Legendre P. and Drapeau P. (1992) Partialling out the spatial component of ecological
variation. Ecology 73 (3): 1045-1055
Brown, J. Ferrians, O.J. Jr., Heginbottom, J.A. and Melnikov, E.S. (1997) Circum-Arctic map of
permafrost and ground-ice conditions, 1:10 000 000, Map CP-45, United States Geological Survey,
International Permafrost Association.
Christensen, J.H., B. Hewitson, A. Busuioc, A. Chen, X. Gao, I. Held, R. Jones, R.K. Kolli, W.-T.
Kwon, R. Laprise, V. Magaña Rueda, L. Mearns, C.G. Menéndez, J. Räisänen, A. Rinke, A. Sarr and
P. Whetton, (2007), Regional Climate Projections. In: Climate Change 2007: The Physical Science
Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental
Panel on Climate Change, [Solomon, S., D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M.
Tignor and H.L. Miller (eds.)], Cambridge University Press, Cambridge, United Kingdom and New
York, NY, USA.
Christensen, T.R., Friborg, T. and Johansson, M. (2008) Trace Gas Budgets of High Arctic Permafrost
Regions, Ninth International Conference on Permafrost, Proceedings. Vol 1: 251-256
Christensen, T.R., T. Johanson, J. Åkerman and M. Mastepanov (2004) Thawing sub-arctic permafrost:
Effects on vegetation and methane emissions. Geophys. Res. Lett., 31, L04501,
doi:10.1029/2003GL018680
Davidson E.A. and I.A. Janssens (2006) Temperature sensitivity of soil carbon decomposition and
feedbacks to climate change, Nature, 440, 165-173
Darwin C. (1881) The Formation of Vegetable Mould Through the Action of Worms, With
Observations on Their Habits. John Murray, London, United Kingdom. 326 pp.
Dokuchaev V.V. (1898) Writings (in Russian), Vol 6, 381 pp. Akad. Nauk, Moscow
FAO (1993) WORLD SOIL RESOURCES, An explanatory note on the FAO World Soil Resources
Map at 1:25 000 000 scale, World Soil Resources Report 66 Rev. 1, Rome: FAO
FAO (1995) Global and National Soils and Terrain Digital Databases (Soter) World Soil Resources
Report 74 Rev. 1, Rome: FAO
FAO (1998), World Reference Base for Soil Resources, World Soil Resource Reports, 84, FAO, Rome
Forster, P., V. Ramaswamy, P. Artaxo, T. Berntsen, R. Betts, D.W. Fahey, J. Haywood, J. Lean, D.C.
Lowe, G. Myhre, J. Nganga, R. Prinn, G. Raga, M. Schulz and R. Van Dorland, 2007: Changes in
Atmospheric Constituents and in Radiative Forcing. In: Climate Change 2007:The Physical Science
Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental
Panel on Climate Change, [Solomon, S., D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt,
M.Tignor and H.L. Miller (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New
York, NY, USA.
20
GLC 2000 (2002) GLC 2000: Global Land Cover Mapping for the year 2000, project status November
2002 [E. Bartholome, E., A. S. Belward, F. Achard, S. Bartalev, C. Carmonamoreno,
H. Eva, S. Fritz, J-M. Gregoire, P. Mayaux, and H-J. Stibig (eds.)] European Commission. available at:
http://www.tem.jrc.it/glc2000/Publications/GLC2000%20EUR20524_report_v2.pdf
Global Land Cover Facility (2009)
Gorham E. (1991) Northern peatlands: Role in the Carbon cycle and probable responses to Climatic
Warming, Ecological applications, 1(2), 182-195
Government of Canada (1998) The State of Canada's Ecosystems in Maps, EcoATLAS, Ottawa,
Ontario, Canada. Available at: http://geogratis.cgdi.gc.ca/Ecosystem/ecosystem.html
Gruber N., P. Friedlingstein, C.B. Field, R. Valentini, M. Heimann, J.E. Richey, P. Romero-Lankao, D.
Schulze, and C.T.A. Chen (2004) The vulnerability of the carbon cycle in the 21st century: An
assessment of carbon-climate-human interactions, in The Global Carbon Cycle: Integrating Humans,
Climate, and the Natural World, edited by Field C. and M. Raupach, pp. 45-76, Island Press,
Washington, DC.
Heginbottom J. A. (2002) Permafrost mapping: a review. Progress in Physical Geography 26 (4): 623642
Heikkinen, J. E. P., T. Virtanen, J. T. Huttunen, V. Elsakov and P. J. Martikainen (2004), Carbon
balance in East European tundra, Global Biogeochemical Cycles, 18, GB1023,
doi:10,1029/2003GB002054
Ikeya, K., and Watanabe, A. (2003) Direct expression of an index for the degree of humification of
humic acids using organic carbon concentration. Soil Science and Plant Nutrition 49, 47-53.
Jenny H. (1930) A study on the influence of climate upon the nitrogen and organic matter content of
the soils. Univ. Missouri Agric. Exp. Stn. Res. Bull. 162: 66 pp.
Jenny H. (1941) Factors of soil formation. McGraw-Hill, New York
Jenny H. (1980) The Soil Resource: Origin and Behavior. Springer, New York
Jobbagy, E. G. and R. B. Jackson (2000), The Vertical Distribution of Soil Organic Carbon and Its
Relation to Climate and Vegetation, Ecological Applications, 10(2), 423-436.
Jongman R. H. G., Ter Braak C. J. F. and Van Tongeren O. F. R. (1995) Data analysis in community
and landscape ecology. 299 pp., Cambridge University Press, Cambridge, UK
Kent M. (2006) Numerical classification and ordination methods in biogeography. Progress in Physical
Geography 30 (3): 399-408
King L. (1983) high mountain permafrost in Scandinavia. Permafrost, Fourth International
Conference, Proceedings: 612-617
Kuhry P. and Vitt D.H. (1996) Fossil Carbon/Nitrogen ratios as a measure of peat decomposition.
Ecology 77 (1), 271-275
Kuhry P., G.G. Mazhitova, P.-A. Forest, S.V. Deneva, T. Virtanen and S. Kultti (2002), Upscaling soil
organic carbon estimates for the Usa Basin (Northeast European Russia) using GIS-based landcover
and soil classification schemes, Danish Journal of Geography, 102, 11-25
Lemke P, Ren J, Alley RB, Allison I, Carrasco J, Flato G, Fujii Y, Kaser G, Mote P, Thomas RH,
Zhang T. (2007) Observations: Changes in snow, ice and frozen ground, in Climate Change 2007: The
Physical Science Basis. Contribution of Working group I to the Fourth Assessment Report of the
Intergovernmental Panel on Climate Change, [Solomon, S., D. Qin, M. Manning, Z. Chen, M.
21
Marquis, K.B. Averyt, M. Tignor, H.L. Miller (eds,)], pp. 337-384, Cambridge University Press:
Cambridge, United Kingdom and New York, NY, USA
Mazhitova G.G., V.G. Kazakov, E.V. Lopatin and T. Virtanen (2003) Geographic Information System
and Soil Carbon Estimates for the Usa River Basin, Komi Republic, Eurasian Soil Science, 36(2), 133144
NASA (2007), MODIS 16-day Composite MOD44C, latlon.na.2004289b3, Collection 4, The Global
Land Cover Facility, University of Maryland, College Park, Maryland, Day 289, 2004.
Oechel W.C, S.J. Hastings, G. Vourlitis, M. Jenkins, G. Riechers and N. Grulke (1993), Recent change
of Arctic tundra ecosystems from a net carbon dioxide sink to a source, Nature, 361, 520-523
Peres-Neto P.R., P. Legendre, S. Dray, and D. Borcard (2006) Variation partitioning of species data
matrices: estimation and comparison of fractions, Ecology, 87(10), 2614-2625
Ping, C. L., G.J. Michaelson, M.T, Jorgenson, J.M. Kimble, H. Epstein, V.E. Romanovsky and D.A.
Walker (2008) High stocks of soil organic carbon in the North American region, Nature Geoscience,
1:615-619
Post W.M., W.R. Emmanuel, P.J. Zinke and A.G. Stangenberger (1982), Soil carbon pools and world
life zones, Nature, 298, 156-159
Schlesinger, W. H. (1977) Carbon balance in terrestrial detritus, Annual Review of Ecology and
Systematics, 8, 51-81
Schlesinger, W. H. (1997) Biogeochemistry, an analysis of global change, 443 pp., Academic Press,
San Diego, California
Schuur, E.A.G., J. Bockheim, J.G. Canadell, E. Euskirchen, C.B. Field, S.V Goryachkin, S. Hagemann,
P. Kuhry, P.M. Lafleur, H. Lee, G. Mazhitova, F.E. Nelson, A. Rinke, V.E. Romanovsky, N.
Shiklomanov, C. Tarnocai, S. Venevsky, J.G. Vogel and S.A. Zimov (2008), Vulnerability of
Permafrost Carbon to Climate Change: Implications for the Global Carbon Cycle, BioScience, 58(8),
701-714, doi:10.1641/B580807
Simbahan G.C. and Dobermann A. (2006) Sampling optimization based on secondary information and
its utilization in soil organic mapping. Geoderma 133: 345-362
Smith L. C., G.M. MacDonald, A.A. Velichko, D.W. Beilman, O.K. Borisova, K.E. Frey, K.V.
Kremenetski and Y. Sheng (2004), Siberian peatlands a net carbon sink and global methane source
since the early Holocene, Science, 303, 353-356
Smith L.C., Y. Sheng, G.M. MacDonald and L.D. Hinzman (2005), Disappearing arctic lakes, Science,
308, 1429
Stendel M. and J.H. Christensen (2002), Impacts of global warming on permafrost conditions in a
coupled GCM, Geophysical Research Letters, 29(13), 1632
Tarnocai, C. (2006) The effect of Climate Change on Carbon in Canadian Peatlands, Global and
Planetry Change, 53, 222-232
Tarnocai, C. and Broll, G. (2008) Soil Organic Carbon Stocks in the Northern Permafrost Region and
Their Role in Climate Change. Ninth International Conference on Permafrost, Proceedings. 1751-1755
Tarnocai, C., Canadell, J., Mazhitova, G., Schuur, E.A.G., Kuhry, P. and Zimov, S. Soil organic
carbon stocks in the northern circumpolar permafrost region. Global Biogeochemical Cycles (in press)
Tarnocai, C. & Lacelle, B. (1996) Soil organic carbon of Canada map. Eastern Cereal and Oilseed
Research Centre, Agriculture and Agri-Food Canada, Research Branch, Ottawa, Ontario.
22
Tarnocai, C., Swanson, D., Kimble, J., and Broll, G. (2007) Northern Circumpolar Soil Carbon
Database. Research Branch, Agriculture and Agri-Food Canada, Ottawa, Canada (digital database)
Ter Braak C.J.F and P. Smilauer (2002), CANOCO Reference manual and CanoDraw for Windows
User’s guide: Software for Canonical Community Ordination (version 4.5), 500 pp., Microcomputer
Power, Ithaca, NY
Yoshikawa K. and L.D. Hinzman (2003), Shrinking Thermokarst ponds and groundwater dynamics in
discontinuous permafrost near Council, Alaska, Permafrost Periglac. Process., 4, 151-160
Wagner H. H. (2003) Spatial covariance in plant communities: integrating ordination, geostatistics and
variance testing. Ecology 84 (8): 1045-1057
Wagner H. H. (2004) Direct multi-scale ordination with Canonical Correspondence Analysis. Ecology
85 (2): 342-351
Wagner, H. H. and Fortin, M. J. (2005). Spatial analysis of landscapes: Concepts and statistics. Ecology
86(8): 1975-1987.
White A., M. G. R. Cannel and A. D. Friend (2000), The high-latitude terrestrial carbon sink: a model
analysis, Global Change Biology, 6, 227-245
Zimov S. A., E. A. G. Schuur and F.S. Chapin (2006), Permafrost and the global carbon budget,
Science, 312, 1612-1613
23
24