Stable carbon isotope depth profiles and soil organic carbon ...

Geoderma 131 (2006) 89–109
www.elsevier.com/locate/geoderma
Stable carbon isotope depth profiles and soil organic carbon
dynamics in the lower Mississippi Basin
Jonathan G. WynnT, Jennifer W. Harden 1 , Terry L. Fries 1
US Geological Survey, Menlo Park, CA, 94025 USA
Received 28 June 2004; received in revised form 9 March 2005; accepted 10 March 2005
Available online 21 April 2005
Abstract
Analysis of depth trends of 13 C abundance in soil organic matter and of 13 C abundance from soil-respired CO 2 provides
useful indications of the dynamics of the terrestrial carbon cycle and of paleoecological change. We measured depth trends of
13 C abundance from cropland and control pairs of soils in the lower Mississippi Basin, as well as the 13 C abundance of soilrespired
CO 2 produced during approximately 1-year soil incubation, to determine the role of several candidate processes on the
13 C depth profile of soil organic matter. Depth profiles of 13 C from uncultivated control soils show a strong relationship
between the natural logarithm of soil organic carbon concentration and its isotopic composition, consistent with a model
Rayleigh distillation of 13 C in decomposing soil due to kinetic fractionation during decomposition. Laboratory incubations
showed that initially respired CO 2 had a relatively constant 13 C content, despite large differences in the 13 C content of bulk soil
organic matter. Initially respired CO 2 was consistently 13 C-depleted with respect to bulk soil and became increasingly 13 C-
depleted during 1-year, consistent with the hypothesis of accumulation of 13 C in the products of microbial decomposition, but
showing increasing decomposition of 13 C-depleted stable organic components during decomposition without input of fresh
biomass. We use the difference between 13 C/ 12 C ratios (calculated as d-values) between respired CO 2 and bulk soil organic
carbon as an index of the degree of decomposition of soil, showing trends which are consistent with trends of 14 C activity, and
with results of a two-pooled kinetic decomposition rate model describing CO 2 production data recorded during 1 year of
incubation. We also observed inconsistencies with the Rayleigh distillation model in paired cropland soils and reasons for these
inconsistencies are discussed.
D 2005 Elsevier B.V. All rights reserved.
Keywords: Soil organic carbon; Carbon isotope; Carbon cycle; Decomposition; Depth
T Corresponding author. Current address. School of Geography
and Geosciences, University of St. Andrews, St. Andrews, Fife
KY16 9AL, United Kingdom. Tel.: +44 1334 463936; fax: +1 650
329 4920.
E-mail addresses: jonathan.wynn@st-andrews.ac.uk
(J.G. Wynn), jharden@usgs.gov (J.W. Harden),
tfries@usgs.gov (T.L. Fries).
1 Fax: +1 650 329 4920.
1. Introduction
Spatial and temporal trends of the carbon isotopic
composition of soil organic matter (SOM) provide one
of the main tools used to understand component
processes of the terrestrial carbon cycle (Bird and
0016-7061/$ - see front matter D 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.geoderma.2005.03.005

90
J.G. Wynn et al. / Geoderma 131 (2006) 89–109
Pousai, 1997; Bird et al., 2001; Ehleringer et al.,
2000; Powers and Schlesinger, 2002). Trends of
carbon isotopes with depth in soils have been
primarily investigated for two purposes: (1) to understand
below-ground dynamics of the carbon cycle
through natural and synthetic 13 C-labeling studies
with a well documented or controlled change in the
isotopic composition of biomass (Balesdent and
Mariotti, 1996) and (2) to reconstruct the nature and
timing of naturally induced changes between biomass
with different isotopic composition (usually a change
between C3 and C4 plants which have very different
isotopic ratios). This second class of 13 C-depth related
research relies heavily on natural 13 C trends because it
applies similar methods and assumptions to detect
vegetation changes which occurred during unrecorded
history or in poorly documented locations (Ambrose
and Sikes, 1991; Boutton, 1996; Boutton et al., 1998;
Kelly et al., 1991). Paleovegetation studies make the
simplification that
13 C/ 12 C fractionation during
decomposition of SOM does not occur, or at least
can be regarded as negligible, and thereby apply a
mixing equation for the relative contribution of C3
and C4 plants:
F C4 ¼ d13 C SOC d 13 C C3
d 13 C C4 d 13 ð1Þ
C C3
( F C4 is the fractional contribution of C4 plants to
SOC. d 13 C SOC , d 13 C C3 and d 13 C C4 indicate the
measured isotopic composition of SOC and average
values for C3 and C4 biomass, respectively, expressed
in d-notation as per mil (x) deviation from a
standard, PeeDee Belemnite, PDB).
However, fractionation during decomposition
either negates or complicates these simple mixing
models, resulting in a bnaturalQ depth profile of
13 C/ 12 C ratios without vegetation change. In order
to apply a mixing model correctly, one must either
demonstrate that fractionation does not occur during
decomposition, or account for any fractionation
during decomposition with control soils such as
bbenchmarkQ profiles under demonstrably stable
vegetation, preferably with a narrow range of 13 C
composition (pure C3 or C4). Robust tests for such
fractionation are needed and, if it can be demonstrated
to occur, quantification of the degree of fractionation
with depth under different soil forming conditions is
needed to constrain paleovegetation work. Because
most well drained soil profiles exhibit 13 C-enrichment
with depth (Bird et al., 2001) and because SOC in
deeper horizons is commonly older (Trumbore, 2000),
we are challenged to address the effects of both burial
and aging on the 13 C/ 12 C ratio of SOC.
The purpose of this paper is to examine how depth
trends of d 13 C SOC are affected by natural decomposition
processes in uncultivated soils and then to examine
the roles of anthropogenic impacts of cultivation and
erosion on these natural depth trends, using paired
cropland and control sites studied as part of the
Mississippi Basin Carbon Project (MBCP). The MBCP
study area in the Yazoo Basin, Mississippi includes
paired cultivated and uncultivated soils, with well
documented history since 1870, in which soil profiles
were measured and collected from ridgetops, upper and
lower slopes, and valley bottoms each with a variety of
replicates and depth intervals sampled (Harden et al.,
1999a). Control sites within this framework of paired
sites allow us to address the roles of natural processes
such as carbon isotope fractionation during decomposition,
mixing of organic matter with different
isotopic composition and preferential decomposition
of variably stable components of SOM (introduced in
Section 1.1.). Given an understanding of these natural
variations, we then move on to examine the role of
well documented agricultural activities on the changes
of vertical distribution of SOC concentration and
d 13 C SOC (introduced in Section 1.2).
1.1. Natural depth trends of stable carbon isotope
ratios
Although there is frequently a strong relationship
between d 13 C SOC of fresh organic litter at the soil
surface and that of overlying biomass at a variety of
spatial scales (Bird and Pousai, 1997; Bird et al.,
2001), vertical trends of d 13 C SOC in well drained
mineral soils frequently show increasing 13 C with
depth, by as much as several per mil within the top
meter of soil. Such depth-increasing trends are evident
in a number of environments, which have been
evidently stable C3 ecosystems through the Pleistocene
and in some cases to the Miocene (Bird et al.,
2003; Krull et al., 2002; Krull and Skjemstad, 2002;
Powers and Schlesinger, 2002; Wynn et al., in press),
implying that ecological change cannot account for

J.G. Wynn et al. / Geoderma 131 (2006) 89–109 91
the depth increases observed in these settings.
Balesdent et al. (1993) and Ehleringer et al. (2000)
provide reviews of a number of hypotheses which
have been developed over the past several decades to
explain the commonly observed depth-enriched
d 13 C SOC profiles in such well drained mineral soils.
For this discussion, we group these hypotheses into 3
categories: (1) those involving mixing of SOM with
differing 13 C content, (2) those involving preferential
decomposition of SOM components with differing
13 C content, and (3) those involving kinetic fractionation
of carbon isotopes during the maturation of
SOM.
1.1.1. Hypothesis group 1: input mixing of organic
matter with differing isotopic composition
Hypotheses 1.1.1.1 through 1.1.1.3 are grouped
due to similar predicted depth trends of SOC
concentration and d 13 C SOC , all of which result from
physical mixing of bulk SOM derived from biomass
with different isotopic composition.
1.1.1.1. Terrestrial Suess effect. Friedli et al. (1986)
demonstrated that the 13 C content of the current
atmosphere has been diluted by up to 1.4x since the
beginning of the Industrial Revolution, due predominantly
to burning of 13 C-depleted fossil fuels (Suess,
1955). Because mean age of SOM increases with
depth, one would expect that mixing of SOM derived
from biomass input to soil during decreasing atmospheric
d 13 C CO2 would produce an increasing depth
d 13 C SOC trend (older more 13 C-enriched SOC at
depth, younger more 13 C-depleted SOC at the surface).
If the mixing can be approximated by two pools
with end-member compositions, SOC would follow a
mixing line between pre- and post-Industrial Revolution
biomass (Fig. 1A), with an overall magnitude of
isotopic difference equal to the magnitude of the
change in atmospheric d 13 C CO2 .
1.1.1.2. Surface litter and root-derived SOM. Many
studies of isotope ratios in plant biomass have
demonstrated that roots are generally more 13 C-
enriched than above ground biomass (particularly
leaves) from the same plant (Brugnoli and Farquhar,
2000; Schweizer et al., 1999; Wedin et al., 1995). This
affects the isotopic composition of SOC through
mixing of inputs to the SOC pool from predominantly
leaf and stem derived litter at the surface, and belowground
litter derived from roots. The resulting trends
of d 13 C SOC should follow a mixing-line relationship,
similar to that above, with depleted surface values
giving way to enriched root-derived SOC at depth
(Fig. 1B), with a maximum total difference of about
1.5x, the mean difference between above- and
below-ground biomass.
1.1.1.3. Variable biomass composition and old versus
new SOM. A third mixing relationship that may
potentially affect the depth profile of d 13 C SOC is
mixing between SOM derived from two different
sources of biomass. Although the magnitude and sign
of isotopic trends may vary, the extreme example of
this would be a change from one photosynthetic
pathway to another, the basis for most natural labeling
studies (magnitude of ~15x). However, within the
C3 plant group, d 13 C ranges from about 25x to
32x and depends on a number of ecological factors
such as water availability (Ehleringer et al., 1993),
salinity (Sandquist and Ehleringer, 1995), topographic
position (Balesdent et al., 1993), and the degree of
recycling of respired CO 2 which is 13 C-depleted
(Vogel, 1978). Mixing of pools of SOM derived from
different C3 biomass sources due to changes in these
factors would similarly follow a mixing line between
the two end-member isotopic compositions (Fig. 1C).
1.1.2. Hypothesis group 2: preferential decomposition
or stabilization of components with different isotopic
composition
Organic components comprising SOM are well
known to have a range of d 13 C due to variation of
metabolic fractionations within plants. Lipids, lignin
and cellulose are generally 13 C-depleted with respect
to the whole plant, while sugars, amino acids, hemicellulose
and pectin are 13 C-enriched with respect to
the whole plant (Boutton, 1996; Deines, 1980). Due to
their relative chemical stability,
13 C-depleted lipids
and lignin can be preferentially accumulated during
the initial stages of SOM decomposition and their
concentration in some cases increases with depth and
with SOM age (Benner et al., 1987; Wedin et al.,
1995). This trend would leave an accumulating pool
of SOM depleted in 13 C and result in a depth- 13 C-
depleted d 13 C SOC profile. Such preferential decomposition
of
13 C-enriched compounds (non-lignin)

92
J.G. Wynn et al. / Geoderma 131 (2006) 89–109
A. B.
F post-IR post-Industrial
1
Revolution
0.5
Z
0.1
0.05
~19.2‰
discrimination
0.01
0.005
pre-Industrial
Revolution
-29 -28 -27 -26 -25 -9 -8 -7 -6 -5
δ 13 C biomass δ 13 Catm.
F root
1
0.5
0.1
0.05
0.01
0.005
C. D.
leaf litter derived
Z
13 C-enriched
root-derived SOC
-28 -26 -24 -22 -20
δ 13 C
F new
1
0.5
0.1
0.05
Z
fresh SOC
Z
F SOC
1
0.5
0.1
0.05
δ i
“fresh” SOC
Z
α = 0.997
0.01
0.005
“old”
13
C-depleted
SOC
“old”
13
C-enriched
SOC
0.01
0.005
α = 0.999
13 C-distilled
SOC
-28 -26 -24 -22 -20
δ 13 C
-28 -26 -24
δ 13 C
-22 -20
Fig. 1. Model plots of the fraction of remaining soil organic carbon ( F SOC ) with respect to stable carbon isotopic composition of SOC
(d 13 C SOC ). In all plots z indicates the direction of increasing depth and, by assumption, increasing SOC age. A, B, and C represent mixing lines
between pools of SOC with different isotopic composition (corresponding to Hypotheses 1.1.1.1, 1.1.1.2, and 1.1.1.3 in the text; concave
downward in linear–log space). A. Mixing of SOC derived from the modern atmosphere versus that derived from a pre-Industrial Revolution
atmosphere. B. Mixing of leaf litter-derived SOC and root-derived SOC. C. Mixing of boldQ and bnewQ SOC formed under two different
vegetation communities (slope could vary from positive to negative depending on direction of shift). D. 13 C distillation during decomposing
SOM (corresponding to Hypothesis 1.1.3 in the text; linear in linear–log space). The gray lines show the model with varying fractionation
factors from 0.997 to 0.999 (discrimination from ~1x to 3x).
during the initial stages of SOM decomposition
presumably works against the processes contributing
to depth-enrichment observed within most well
drained soils.
1.1.3. Hypothesis group 3: kinetic fractionation
during maturation of SOC
A number of studies have hypothesized that kinetic
fractionation of carbon isotopes occurs during maturation
of SOM and that this fractionation is established
by relationships between SOC concentration
and isotopic composition (Balesdent et al., 1993;
Balesdent and Mariotti, 1996; Högberg and Ekblad,
1996; Natelhoffer and Fry, 1988; Powers and Schlesinger,
2002; Schwartz et al., 1992). This relationship
is nonlinear and best fit by log SOC concentration as a
linear function of isotopic composition, often understood
in terms of Rayleigh distillation described
below. Kinetic fractionation has been suggested to
occur through microbial respiration of CO 2 from a
decomposing pool of SOM (Huntington et al., 1998;
Mary et al., 1992), as respiration processes typically
involve kinetic fractionation whereby 12 C is preferentially
respired. Despite this evidence, a number of
laboratory studies of short-term controlled biomass
decomposition have not produced consistent evidence

J.G. Wynn et al. / Geoderma 131 (2006) 89–109 93
for microbial kinetic fractionation, especially of the
magnitude and sign needed to explain depth trends of
d 13 C SOC in soils (Lin and Ehleringer, 1997; Wedin et
al., 1995). However, in a novel study, Šantrůčková et
al. (2000) measured the isotopic composition of
physically separated microbial biomass (C mic ) and
observed that 13 C-enrichment of C mic is offset by 13 C-
depletion in respired CO 2 , resulting in respired CO 2
with an isotopic composition similar to that of SOC.
These authors then hypothesized that preferential
protection and stabilization of these 13 C-enriched
decomposition products (C mic ) produces the commonly
observed 13 C-enrichment with depth in soil,
despite the fact that respired CO 2 has a similar
isotopic composition to that of SOC. Differential
stabilization and sorption of organic components of
SOM (Kaiser et al., 2001; Rumpel et al., 2002) can
also be grouped with this category, as these processes
produce similar trends of d 13 C SOC through preferential
preservation of the 13 C-enriched decomposition
products of microbial transformation.
1.2. Anthropogenic effects on depth trends of stable
carbon isotope ratios
The effects of human activities on depth trends of
d 13 C SOC must be viewed as superimposed on any
natural trends discussed in the previous section. As
discussed above, most of this work to date focuses on
interpretation of ecological change between C3 and
C4 vegetation. Because C3 and C4 plants differ by
approximately 15x (Deines, 1980) and the precision
limits of most analytical techniques are well within a
few tenths of a per mil, stable carbon isotopes can be a
powerful isotopic tracer of ecological and paleoecological
change. Such trends clearly appear as depthincreasing
profiles of d 13 C SOC in well-documented
woody vegetation thickening settings and can be
further detailed with additional analyses separating
SOM into a pool structure and 14 C measurements
(Krull et al., in press). The effects of enhanced
decomposition of SOM and erosion are less well
studied using stable isotope techniques than by
radioisotopic measurements (Harden et al., 2002,
1999b) and, hence, their effects on depth trends of
d 13 C SOC are modeled in more detail in the discussion
of this paper. In most simplified terms, both erosion
and enhanced decomposition of SOM can be expected
to bflattenQ the depth profile of d 13 C SOC . In the case of
erosion, surficial horizons are removed, leaving a
truncated profile of lower concentration and more
highly decomposed SOM, thereby lacking the typically
abrupt 13 C-depletion in surface horizons where
SOC concentration increases exponentially. In the
case of enhanced decomposition, the most labile
fractions of SOM are removed preferentially from
surface horizons, leaving a similarly flattened trend of
d 13 C SOC with a more homogenous distribution of the
stable fractions of SOM.
2. Materials and methods
2.1. Site selection and descriptions
The MBCP study area lies within the Yazoo River
basin of the lower Mississippi River Valley, a region
historically known for its agricultural production
(generally cotton) and soil erosion (Sundquist et al.,
1998). The soils are developed on very erodable siltrich
Peoria Loess and have well-developed fragipans
(soil horizons at depth with relatively high bulk
density and which have a high mechanical strength
when dry but are moderately brittle when moist).
Fragipans are good indicators of intensive stable
chemical weathering and their depth can be used as
an indicator of the degree of increased erosion since
the period of stable weathering. The Goodwin Creek
(GC) control site is an area of 20 acres with mixed
hardwood forest (dominantly Post Oak, Quercus
stellata) that is documented to be at least 80 years
old. Soils at Goodwin Creek are described as Loring
series (fine-silty, mixed, thermic Typic Fragiudalfs).
Local mean annual precipitation is approximately
1400 mm, while mean annual temperature is 17.2 8C
(Sharpe et al., 1998). The depth to fragipan within this
area is 50–60 cm, suggesting minimal local erosion.
The cropland site at Nelson Farm (NF) is part of an
experimental station run by the U.S. Department of
Agriculture (USDA) Agricultural Research Service
(ARS), which was first farmed by the Nelson family
in 1870. Soils at Nelson Farm are described as eroded
Memphis silt loam (Typic Hapludalf) on ridges and
severely eroded Grenada silt loam (Glossic Fragiudalfs)
on hillslopes (Sharpe et al., 1998). The
agricultural stations’ cropping history is well docu-

94
J.G. Wynn et al. / Geoderma 131 (2006) 89–109
mented, consisting predominantly of cotton and
soybean (C3) cultivation, with a brief period of
sorghum and corn (Harden et al., 1999b). At both
Goodwin Creek and Nelson Farm, soils were sampled
at a range of slope positions: ridgetop (slope b3%),
upper midslope (eroding, slopes 5–10%), lower
toeslope (depositional, slopes b3%), and alluvial
valley bottoms (0–3% slope). Samples were collected
in 1 in diameter stainless steel tubing from a variety of
depth increments and replicates for solid phase
measurements are described below (see results section
and Harden et al. (1999a) for depth intervals,
replication and raw data). Triplicate samples from
20 cm length 1 in diameter stainless steel tubing were
collected for laboratory incubations using the procedure
described below. Roots were not removed or
bpickedQ from any samples prior to analysis.
2.2. Methodology
2.2.1. Solid-phase carbon concentration and stable
isotope ratios
Dry mass concentration of SOC (discussed here as
fractional mass, f C ) and its stable carbon isotope
composition (d 13 C SOC ) were measured on duplicate
samples using elemental analysis mass spectrometry
(EA-MS) at the U.S. Geological Survey (USGS) in
Menlo Park, California. Dry mass concentration was
converted to dry volume concentration (reported here
in grams of carbon per cubic centimeter, g C/cm 3 )
using dry bulk density measurements, on these
duplicates (mass concentration of dry solids in g/
cm 3 ). In all samples, no effort was made to remove
roots from soil prior to analysis. pH field measurements
made on soils from both the Goodwin Creek
and Nelson Farm sites indicated that no inorganic
carbon should be present and inorganic carbon
determination on a selected set of samples indicated
concentrations were less than 0.005%. Hence no lab
procedure was used for pretreatment removal of soil
carbonate. All solid phase measurements and much
ancillary data for the sites are reported in a USGS
open-file report (Harden et al., 1999a).
2.2.2. Laboratory incubations
Samples for incubation were selected to examine
differences in decomposition rate related to cultivation,
erosional regime, and depth and thus include two
depth increments from upper and lower slope
positions in both cropland and control sites. Separate
samples for laboratory incubations were taken during
spring or fall and incubated in undisturbed sampling
tubing to best simulate field conditions of soil
respiration. Three replicate samples 20 cm in length
were incubated in ~2000 mL closed vessels with
swagelok fittings at constant laboratory temperature
(F2 8C), beginning at field moisture conditions, with
minimal water loss during intervals when CO 2
samples were collected and head space flushed with
dry CO 2 -free air. Because the incubations were
carried out simultaneously, minor changes in temperature
and moisture content affected all samples
equally. The vessel head space was flushed frequently
to avoid accumulation of CO 2 concentrations greater
than 5% by volume. Incubation measurements at time
series reported here were carried out following an
initial bdisturbanceQ peak of high CO 2 flux from
samples due to continuation of root respiration. CO 2
concentration of gas samples taken from the incubation
vessel was measured during flushing intervals
using a gas chromatograph (GC) at the USGS in
Menlo Park. Time series measurements of the amount
of SOC remaining during incubation experiments
were made at frequent intervals over the course of
approximately 1-year incubations to describe the
kinetic rates of SOC decomposition. Concentration
measurements were converted to mass flux of CO 2
using the volume of head space in the sample vessel,
thickness, bulk density and carbon concentration of
the sample and the ideal gas law. Measurements from
the beginning (following the initial disturbance peak)
of these laboratory incubations are comparable to flux
measured using gas chambers in the field (Harden et
al., 1999b; Huntington et al., 1998). Slope of a simple
linear regression of lab versus field flux is 1.08
(R 2 =0.74). After 1000 h of incubation the laboratory
flux is lowered to a slope of 0.90 times the field flux
(R 2 =0.81), demonstrating that the incubations simulate
field conditions and that decomposition rates
slowed during the course of the experiment, likely
due to preferential consumption of the labile pool
during 1 year without fresh biomass input.
2.2.3. Isotopic composition of respired CO 2
Respired CO 2 for isotopic analyses was collected
after an initial spike of rapid respiration after which

J.G. Wynn et al. / Geoderma 131 (2006) 89–109 95
CO 2 flux returned to relatively stable values. Samples
of soil-respired CO 2 were collected from the incubation
vessels using a 1 L evacuated stainless steel
vessel attached with a swagelok fitting, providing at
least 1 mg C for 14 C analysis and at least 100 Amol C
for 13 C analysis. These gas samples were cryogenically
purified for 13 C and 14 C analyses on a vacuum
extraction line in Menlo Park. Isotopic composition
was measured on samples incubated over the course
of approximately 1 year, with samples taken for
isotope analysis at three intervals (after approximately
66, 139 and 371 days of incubation). d 13 C of respired
CO 2 was measured by dual-inlet mass spectrometry
(DI-MS) at the USGS in Menlo Park. For analysis of
d 13 C data, we use the difference in stable isotope
composition of bulk SOC before respiration and the
cumulative respired CO 2 as an bapparent discriminationQ
between solid and gas phases:
Dd 13 C ¼ d 13 C CO2 d 13 C SOC : ð2Þ
Graphite targets were prepared from respired CO 2
at the University of California-Irvine and were ionized
and measured for 14 C activity by accelerator mass
spectrometry (AMS) at the Lawrence Livermore
Center for AMS. Radiocarbon activity is shown here
in fraction modern units (FM), which are equivalent to
percent modern (pM) but expressed as a fraction.
3. MBCP SOC data and discussion
3.1. Concentration profiles of SOC
Soil organic carbon concentration in undisturbed
soils typically follows an exponentially decreasing
function with depth (Elzein and Balesdent, 1995;
Jobbágy and Jackson, 2000). We model SOC concentration
profiles using a simplified solution to the
diffusion–production equation because this derivation
empirically fits the concentration profiles of SOC
well and conforms to theoretical constraints of
bbiodiffusive transportQ of SOC (Elzein and Balesdent,
1995; O’Brien and Stout, 1978). Our use of this
equation assumes that heterotrophic production of
soil-respired CO 2 (expressed as a negative production
term) is in balance with biodiffusive transport. Thus,
this transport function describes the physical movement
and input of carbon down the soil profile,
which is also in balance with the diffusion process of
respiration of CO 2 from the soil, assuming steady
state conditions. The solution of SOC concentration
with depth, C(z), follows the steady state diffusion–
production equation:
D B2 C
Bz
¼
Pz ðÞ
ð3Þ
where C is mass/volume concentration (g/cm 3 ), z is
depth (cm), P(z) is a term describing production of
CO 2 from SOC (g C/cm 3 /s), and D is the biodiffusion
coefficient for SOC transport. This equation has an
analytical solution:
Pz ðÞ¼P s e z=¯z C
ð4Þ
where P s is the production at the surface and z¯C is
the e-folding depth (characteristic production depth,
where P = P s /e). This exponential production function
accounts for realistic depth variation in the rate
of input of organic matter from root biomass, which
mimics root density depth functions that follows
from mathematical models of root growth (Newson,
1995). Thus the steady-state solution of SOC
concentration is:
Cz ðÞ¼C s þ P s¯z 2
D 1 e z=¯z C
ð5Þ
where C s is the concentration at the surface. From
Fick’s Law, we use:
J s ¼ P s ¯z C ð6Þ
(where J s is the flux at the surface, g C/cm 2 /s) to
simplify the steady-state solution to:
Cz ðÞ¼C s
J
s¯z C
D 1 e z=¯z C
: ð7Þ
For each of the landscape positions (ridgetop,
upper slope, lower slope, and valley) at both the
control and cropland sites, we fit the carbon concentration
profiles, C(z), to this exponential function
(Table 1; Fig. 2), with independent parameters C s , J s /
D, and z¯C optimized by least-squares regression.
Concentration data for groups of profiles in similar
landscape positions show reasonably good fit to the
model (R 2 N0.87; note that ridgetop and valley landscape
positions are represented by a single profile

96
J.G. Wynn et al. / Geoderma 131 (2006) 89–109
Table 1
Parameters of empirical curve-fit for concentration profiles, C(z),
and isotope depth profiles y 13 C(z), according to Eqs. (7) and (8),
shown in Fig. 2
SOC concentration profiles
(site and profile type)
Cz ðÞ¼C s
(Eq. (7))
C s
(g/cm 3 )
Js¯zC
D
1 e z=¯zC
J s /D
(g/cm 4 )
z¯C
(cm)
Goodwin Creek
Valley (1 profile) 0.0641 0.0119 4.93 0.96
Lower slope (3 profiles) 0.0310 0.0030 8.87 0.88
Upper slope (3 profiles) 0.0328 0.0023 13.1 0.88
Ridgetop (1 profile) 0.0263 0.0010 24.6 0.89
Nelson Farm
Valley (1 profile) 0.0334 0.00355 7.70 0.90
Lower slope (4 profiles) 0.0179 0.00049 34.3 0.87
Upper slope (4 profiles) 0.0159 0.00103 14.0 0.87
Ridgetop (1 profile) 0.0194 0.00077 22.9 0.96
SOC stable isotope profiles
site and profile type
d 13 Cz ðÞ¼d 13 C s þ k 1
(Eq. (8))
d 13 C s
(x)
k
(x)
e z=¯z d
z¯d
(cm)
Goodwin Creek
Valley (1 profile) 27.2 2.75 8.54 0.46
Lower slope (3 profiles) 27.4 4.83 25.1 0.78
Upper slope (3 profiles) 28.0 4.37 14.5 0.88
Ridgetop (1 profile) 27.8 4.96 28.2 0.95
while upper and lower slope positions are represented
by three replicate profiles). Good fit to this concentration
model is best expressed at the control sites,
where the model is most appropriately applied (carbon
balances are roughly in steady state). The C(z)
function can also be applied to describe the concentration
data at the cropland sites with reasonable fit
parameters (Table 1).
SOC concentration at the surface, C s , is lower by
about a factor of two in the cropland sites compared to
R 2
R 2
control sites, although characteristic production depth,
z¯C, shows no consistent difference between cropland
and control pairs. These trends most likely reflect loss
of labile SOC, which is most abundant at the surface,
through some combination of erosion and agricultural
enhancement of decomposition rate. At both control
and cropland sites, the characteristic production depth,
z¯C becomes more shallow from ridgetop to valley,
while SOC concentration at the surface, C s , increases
from ridgetop to valley (i.e., C(z) profiles from
ridgetops are broadly curved, while those from valleys
are sharply kinked; Fig. 2). An index of root density
from described soil profiles reflects these differences,
where fine and very fine roots penetrate deeper in
midslope positions than in the valley profile (data in
Harden et al., 1999a). Several reasons may exist for
these trends, including (1) the removal of labile
carbon from surface horizons of upper slope positions
and its deposition at the surface in downslope profiles
and (2) shallower seasonal water table and rooting
depths in lowland positions. There may also be a trend
of decreasing porosity downslope allowing biodiffusion
of SOC to shallower depths in the lower slope
positions, although this trend is not demonstrated by
bulk density trends, which can be used to at least
approximate trends in porosity (data in Harden et al.,
1999a).
3.2. Stable carbon isotope depth profiles
d 13 C SOC depth profiles of the control sites (Goodwin
Creek) generally increase by 2–3x within the
first 20 cm of the soil surface and by an additional 2–
3x below about 20 cm to a depth of 1 m sampled.
Stable isotope depth profiles at the cropland site
(Nelson Farm) increase with depth in the first 20 cm
by approximately 2–3x, similar to the control sites.
However, below about 20–30 cm, d 13 C(z) frequently
becomes more depleted with depth in the upper and
Fig. 2. Soil organic carbon concentration and isotope depth profiles of 4 slope settings from control (Goodwin Creek) and cropland (Nelson
Farm) sites. Depth is plotted as depth below the mineral surface. Top row shows carbon concentration profiles, C(z), fit to Eq. (7) in the text
(data from replicate profiles: closed circles, solid line=bulk SOC control site; open circles, dashed line=bulk SOC cropland site; parameters of fit
in Table 1). Second row shows stable carbon isotopic composition, d 13 C, of solid and gas phase. Crosses show d 13 C of incubated samples with
horizontal lines indicating discrimination to time series of stable isotope measurements of CO 2 samples produced by incubation (black
numbers=respired CO 2 control sites; gray numbers=respired CO 2 cropland sites). These time series data are shown in more detail in Fig. 6.
Third row shows radiocarbon composition, FM, of solid phase bulk SOM. Crosses show FM of incubated samples prior to incubation for upper
and lower slope settings.

J.G. Wynn et al. / Geoderma 131 (2006) 89–109 97
Valley Lower Slope Upper Slope Ridgetop
C (z)
C (z)
C (z)
C (z)
0 0.01 0.03 0.05 0 0.01 0.03 0.05 0 0.01 0.03 0.05 0 0.01 0.03 0.05
0
10
20
depth (cm)
30
40
50
60
70
80
90
Goodwin Creek,
control site
Nelson Farm,
cropland site
100
0
10
20
3 2 1
3
2 12
depth (cm)
30
40
50
60
32
1
3
221
31
70
80
90
100
Goodwin Creek,
control site
Nelson Farm,
cropland site
-30 -26 -22
δ 13 C
-30 -26 -22
δ 13 C
-30 -26 -22
δ 13 C
-30 -26 -22
δ 13 C
0
10
bulk
soil
incub.
sample
Goodwin Creek,
control site
Nelson Farm,
cropland site
depth (cm)
20
30
40
50
60
70
80
90
100
0.85 0.95 1.05 1.15
FM
0.85 0.95 1.05 1.15
FM
0.85 0.95 1.05 1.15
FM

98
J.G. Wynn et al. / Geoderma 131 (2006) 89–109
lower slope cropland sites. Although there is much
variability within replicates of d 13 C SOC measurements,
depth trends can be characterized by an
equation similar to that used for SOC concentration:
d 13 Cz ðÞ¼d 13 C s þ k 1 e z=¯z d
: ð8Þ
An empirical constant, k, describes the magnitude
of isotopic shift from surface to maximum depth.
d 13 C s is the isotopic composition at the surface and z¯d
is the e-folding depth at which d 13 C=d 13 C s / e.
Modeled d 13 C s values are typical of local C3 organic
matter. E-folding depths for d 13 C(z) are consistently
deeper than those determined for C(z).
3.3. Estimation of the fraction of remaining SOC
Each of the model SOC concentration profiles
calculated in Section 3.1 were used to estimate the
degree of decomposition of soil organic carbon at a
given depth, f SOC (z), which is normalized to a value
of 1 at the surface:
f SOC ðÞ¼ z
Cz ðÞ
C s
ð9Þ
Substituting from Eq. (4) for C(z), this gives:
¯z 2 C
f SOC ðÞ¼1 z
P
ð s Pz ðÞÞ
C s D
"
¯z 2 C
¼ 1
P #
s 1 e z=¯z C
ð10Þ
C s D
showing the inverse relationship of f SOC (z) to production
(heterotrophic respiration) and characteristic
production depth and the positive relationship of
f SOC (z) with the biodiffusion coefficient, D, and the
normalized to SOC concentration at the surface, C s .
In our subsequent analysis of depth functions of
d 13 C SOC , f SOC is used as a proxy for the degree of
decomposition of SOC and analyzed with respect to
d 13 C SOC as a tracer of processes such as organic
matter mixing, kinetic fractionation during humification
and changes in substrate quality during humification
as shown in the models of Fig. 1. Using f SOC as
a proxy for the degree of decomposition of SOC
assumes that organic matter at the surface is bfresh,Q
( f SOC =1).
3.4. Rayleigh distillation of SOC from the Goodwin
Creek control site
Comparison of the SOC concentration and isotope
data from the control site at Goodwin Creek to the
models shown in Fig. 1 and outlined in Sections 1.1.1,
1.1.2, and 1.1.3 above suggests that depth trends at
this control site are best explained by kinetic
fractionation during SOM humification, which produces
depth-increasing d 13 C SOC with a nonlinear
relationship between f SOC and d 13 C SOC .(Figs. 2 and
3). The magnitude of isotopic difference between
surface and deep samples (generally ~4.5x at Goodwin
Creek) argues against any of the hypotheses
solely involving mixing of SOM of different isotopic
composition (hypotheses grouped in Section 1.1.1),
which cannot explain depth trends exceeding a
maximum of 3x, as does the nonlinear relationship.
Preferential decomposition of more labile components
of SOM with depth (Hypothesis 1.1.2) presumably
works against the observed depth increasing trends
(i.e. organic matter with residually accumulated stable
components would typically show more depleted
d 13 C SOC values at depth). Therefore, in accord with
the hypothesis of kinetic fractionation during decomposition
(Hypothesis 1.1.3), we model depth trends of
d 13 C SOC at the control site using a nonlinear equation
describing Rayleigh distillation in terms of the
fraction of remaining SOC and stable isotope ratios,
taking into account factors describing the efficiency of
microbial assimilation, e, and the fraction of assimilated
carbon retained by a stabilized pool of SOM, t
(Wynn et al., in press):
F ¼
" d 13 C f
1000 þ 1
# h 1
a 1þe ð
½a e 1 ð
d 13 C i
1000 þ 1
ð Þ t 1Þ
ð Þ t 1ÞŠþt
i
1
ð11Þ
for each group of control soil profiles (ridgetop, upper
slope, lower slope, and valley; d 13 C f is the isotopic
composition of SOC when sampled, d 13 C i is the
composition of input from biomass, a is the fractionation
factor between SOC and respired CO 2 , and F is
the fraction of remaining SOC, here approximated by
the calculated value of f SOC ). Nonlinear least-squares
regression of the data from control sites at Goodwin
Creek to this equation show that all profile types fit

J.G. Wynn et al. / Geoderma 131 (2006) 89–109 99
the model of Rayleigh distillation well (the Rayleigh
distillation equation occurs as a line in log–linear
space (Fig. 3; Table 2). Our use of the Rayleigh
distillation equation assumes that SOC decomposes in
an open system (CO 2 diffuses towards the atmosphere)
and that all components of SOC decompose
and contribute to soil-respired CO 2 at the same rate
with depth in the profile (SOC is treated in bulk,
without separation into a pool structure). This
application of the Rayleigh equation is also based
on the assumption that f SOC approximates the degree
of decomposition of SOC ( F), the controlling variable
in the Rayleigh distillation equation (i.e. it assumes
that f SOC iF). Given these assumptions, in all profile
types of the control site there appears to be a
consistent 1–2x difference in the rates of respiration
of 13 Cand 12 C. This fractionation is further supported
by isotopic measurements of respired CO 2 from
incubation studies discussed below.
In order to refine this model, we repeat this analysis
removing potential effects of mixing of SOC with
different isotopic composition (Hypothesis 1.1.1;
linear relationship between f SOC and d 13 C SOC ) to
examine the effects of distillation alone. We recalculate
the regression to the Rayleigh distillation equation
after subtraction of a synthetic linear function (Fig. 3):
d 13 C corrected ¼ d 13 C raw
d 13 C correction
¼ d 13 C raw kf SOC ð12Þ
using a value of an isotopic difference of 1.5x (k) to
represent the entire difference between litter-derived
and modern SOC at the surface ( f SOC =1) and rootderived
and old SOC at depth where f SOC approaches
0. The effect of this subtraction is to increase the
negative slope in Fig. 3 and decrease the calculated
fractionation, a in the Rayleigh equation (Fig. 3).
Therefore, the kinetic fractionation interpreted for the
control site is above and beyond any trend that might
be observed from mixing of above- and below-ground
pools, which follows linear trends of f SOC to d 13 C SOC .
Although our data from the control site are
consistent with the model of kinetic fractionation
during SOM humification, data from the agricultural
site are difficult, if not impossible, to interpret in this
context (Fig. 4). Only the valley profile from Nelson
Farm shows a consistent trend of increasing d 13 C SOC
with increasingly decomposed SOM ( f SOC Y 0).
f SOC
1
0.5
GC-Ridgetop
f SOC
1
0.5
GC-Upper Slope
0.1
0.05
0.01
0.005
prior to removal
of mixing line.
after removal of
mixing line.
R 2 = 0.63; p = 9.4
0.1
0.05
0.01
0.005
R 2 = 0.85; p < 0.05
-28 -26 -24 -22 -20 δ 13 C
-28 -26 -24 -22 -20 δ 13 C
f SOC
1
0.5
GC-Lower Slope
f SOC
1
0.5
GC-Valley
0.1
0.05
R 2 = 0.77; p < 0.05
0.1
0.05
0.01
0.005
0.01
0.005
R 2 = 0.85; p = 0.2
-28 -26 -24 -22 -20 δ 13 C
-28 -26 -24 -22 -20 δ 13 C
Fig. 3. Regression analysis of the degree of decomposition of soil organic carbon ( f SOC ) to its stable isotopic composition (d 13 C SOC ) for the
various erosional regimes of the MBPC control sites (non-agricultural, Goodwin Creek). The Rayleigh distillation relationship following Eq.
(11) is shown prior to and following subtraction of a linear f SOC to d 13 C SOC relationship (a mixing line), as is discussed in the text. Parameters of
fit are found in Table 2.

100
J.G. Wynn et al. / Geoderma 131 (2006) 89–109
Table 2
Parameters for Rayleigh distillation model curve-fit for kinetic
fractionation with depth in the soil profile (Eq. (11))
Site and profile type
h 1 i
F ¼
d 13 C f
1000 þ1
d 13 C i
1000 þ1
Hence, data from the cropland site (Nelson Farm)
cannot fit the Rayleigh distillation equation. This
probably occurs because of the known disturbances to
both isotope ratios and concentration profiles. The
reasons for misfit of the data from the cropland site to
að1þeÞðt 1Þ
½aðe 1Þðt 1ÞŠþt
(Eq.(11))
a d 13 C i (x) R 2
Goodwin Creek control site (after subtraction of mixing line)
Valley (1 profile) 0.9989 (D=1.1x) 27.3 0.85
Lower slope (3 profiles) 0.9980 (D=2.0x) 26.3 0.77
Upper slope (3 profiles) 0.9986 (D=1.4x) 26.3 0.85
Ridgetop (1 profile) 0.9982 (D=1.8x) 26.3 0.63
Nelson Farm cropland site (after subtraction of mixing line)
Valley (1 profile) 0.9975 26.6 0.75
Values of e and t used are each 0.5, which follow from Šantrůčková
et al. (2000) and Wynn et al. (in press).
1
the distillation model are addressed in a later
discussion section of anthropogenic effects on
d 13 C SOC trends and SOM quality.
Kinetic fractionation during microbial decomposition
of SOM has previously been interpreted from
depth-increasing soil profiles of d 13 C SOC (Balesdent
et al., 1993; Balesdent and Mariotti, 1996; Natelhoffer
and Fry, 1988), modeled by continuous-quality theory
(Ågren et al., 1996) and examined by laboratory
incubation studies of biomass (Fernandez et al., 2003;
Lin and Ehleringer, 1997; Melillo et al., 1989, 1982;
Wedin et al., 1995). However, isotope studies of CO 2
respired by incubation of undisturbed soil, as reported
here, are relatively poorly represented in the literature.
Incubation of plant biomass does not take into account
potential organic–mineral interactions (Huang and
Schnitzer, 1986) or the potential of kinetic isotope
fractionation in clay-dominated soils (Bird et al.,
2003; Bird and Pousai, 1997).
3.5. Kinetics of SOC decomposition
Total SOC decomposed during the length of our 1-
year experiments ranges from about 4–14% of the
f SOC NF-Ridgetop
-28 -26 -24 -22 -20 δ 13 C
1
0.5
f SOC
1
0.5
NF-Upper Slope
0.1
0.05
0.01
0.005
1
0.5
prior to removal
of mixing line.
after removal of
mixing line.
0.01
0.005
-28 -26 -24 -22 -20 δ 13 C
0.1
0.05
f SOC
1
0.5
NF-Valley
f SOC NF-Lower Slope
-28 -26 -24 -22 -20 δ 13 C
0.1
0.05
0.1
0.05
R 2 = 0.85;
p < 0.05
0.01
0.005
0.01
0.005
-28 -26 -24 -22 -20 δ 13 C
Fig. 4. Regression analysis of the degree of decomposition of soil organic carbon to its stable isotopic composition for the various erosional
regimes of the MBPC cropland sites (Nelson Farm). Distillation relationship for the corresponding control sites (Goodwin Creek) from Fig. 3
are shown in gray for comparison. Only the valley profile shows a significant correlation to the distillation model although it is different from
the valley setting at the control site. Uncorrected and corrected relationships for this profile are shown in black lines as in Fig. 3.

J.G. Wynn et al. / Geoderma 131 (2006) 89–109 101
initial SOC pool. To describe the kinetic rate of
decomposition, we follow Fernandez et al. (2003) and
use model of first-order decomposition that uses a
simplified two-pooled structure (labile and refractory)
with separate decomposition rates for each pool, as
developed by Andrén and Paustian (1987). Following
this model, time series data of the fraction of carbon
remaining were fit to a double exponential equation of
the form:
F SOC remaining ¼ f L e k Lt þ ð1
f L Þe k Rt
ð13Þ
with three independent model fit parameters ( f L , k L
and k R ) optimized by least-squares regression. F is the
fraction of SOC remaining at time t, f L is the
fractional quantity of the labile SOC pool and k L
and k R are first-order decomposition constants for
labile and refractory pools of SOC. Least-squares
regression model fits to this data are extremely good
(R 2 N0.996; Table 3). A number of other incubation
studies have addressed organic matter decomposition
rates and isotope fractionation during decomposition
of organic matter at similar incubation time scales to
those measured here (Fernandez et al., 2003; Schweizer
et al., 1999; Wedin et al., 1995). Decomposition
rates of both the labile and refractory pools of SOC in
the Mississippi Basin are significantly slower (each by
about an order of magnitude) than those determined
by similar kinetic models for biomass decomposition
(Fernandez et al., 2003) but consistent with models of
SOM pools from the Century Model parameterization
at these sites for which 1/k of SOM1D =1 year and of
SOM2=32 years (Sharpe et al., 1998).
For all soils in this study, f L of all shallow samples
(0–20 cm) is greater than that of deeper samples (20–
40 cm). f L is also greater from the control site
(Goodwin Creek) than from the cropland site in
equivalent slope positions (Table 3). Similarly, f L is
also greater in ridgetop slope positions than in upper
slope positions in the cropland agricultural sites
(Nelson Farm). These trends indicate that the initial
quality of organic matter (represented by fraction of
labile SOC) decreases with depth in all cases, is lower
in both slope settings at the cropland site than the
control site, and lower in eroded settings (upper and
lower slope positions) than in stable ridgetop positions.
This in turn can be interpreted in the context of
erosional regimes within the cropland sites, where loss
due to erosion removes labile SOC from the surface
and effectively uplifts more stable SOC from greater
depths.
3.6. Carbon isotopes of laboratory respired CO 2
Through the course of the 1-year incubations, all
samples show decreasing 14 C activity (FM CO2 ) and
increasingly negative apparent
13 C discrimination,
(Dd 13 C; Fig. 6). These trends indicate increasing
utilization of old (low 14 C) and 13 C-depleted (more
negative d 13 C) SOC during the course of 1-year
Table 3
Parameters for kinetic model curve-fit of carbon loss during incubation, following first-order decay with a two-pooled structure
Site and profile type F ¼ f L e kLt þ ð1 f L Þe kRt
(Eq. (13))
f L k L (d 1 ) 1/k L (year) k R (d 1 ) 1/k R (yr) R 2 1/k B (year)
Goodwin Creek
Upper slope 20 cm 0.073 0.0114 0.24 19.710 5 13.9 0.9995 12.9
Upper slope 40 cm 0.062 0.0155 0.17 9.3810 5 29.2 0.9972 27.4
Nelson Farm
Lower slope 20 cm 0.041 0.0154 0.18 13.610 5 20.1 0.9996 19.3
Lower slope 40 cm 0.028 0.0086 0.31 3.0110 5 91.0 0.9967 88.5
Upper slope 20 cm 0.040 0.0199 0.14 12.810 5 21.4 0.9991 20.5
Upper slope 40 cm 0.027 0.0155 0.18 9.3810 5 29.2 0.9972 28.4
Ridgetop 20 cm 0.083 0.0058 0.47 8.0810 5 33.9 0.9985 31.1
Ridgetop 40 cm 0.043 0.0131 0.21 17.010 5 16.1 0.9988 15.4
f L is fractional quantity of the labile SOC pool; k L , k R and k B are the decay constants for the labile and refractory pools of SOC and bulk SOC,
respectively; and R 2 is the correlation coefficient of the model fit.

102
J.G. Wynn et al. / Geoderma 131 (2006) 89–109
Cropland Sites (Nelson Farm)
1
Ridgetop
0.98
0-20cm depth
20-40cm depth
0.96
Upper Slope
0.94
0-20cm depth
F 20-40cm depth
SOC remaining
0.92
Lower Slope
0.90
0-20cm depth
0.88
20-40cm depth
Control Sites (Goodwin Creek)
0.86
Upper Slope
0.84 0-20cm depth
0 100 200 300 400
20-40cm depth
Time (days)
Fig. 5. Time series of SOC remaining during incubation experiments fit with a two-pooled kinetic model decomposition (Eq. (13) in the text).
Best fit regression parameters for each sample are found in Table 3.
incubations without the addition of fresh SOC from
biomass. Although all sites demonstrate such trends
in 13 C and 14 C content of respired CO 2 , pronounced
differences in the magnitude of these effects occur
between control and cropland sites, upper and lower
slope settings, and particularly between shallow and
deep samples. Deep samples (20–40 cm) consistently
show greater initial utilization of old and 13 C-
depleted SOC than shallow samples (0–20 cm) in
all settings: agricultural and non-agricultural and
both upper and lower slope positions of agricultural
sites (Figs. 5 and 6). Between control and cropland
1.25 pre-incubated SOC 14 C activities
1.15
1.05
FM CO2
0.95
0.85
0.75
0.00
-1.00
-2.00
∆δ 13 C
-3.00
-4.00
Cropland Sites (Nelson Farm)
Upper Slope
0-20cm depth
20-40cm depth
Lower Slope
0-20cm depth
20-40cm depth
Control Sites (Goodwin Creek)
Upper Slope
0-20cm depth
20-40cm depth
-5.00
-6.00
-7.00
0 50 100 150 200 250 300 350 400
Time (days)
Fig. 6. Carbon-14 activity and carbon-13 isotopic composition of soil-respired CO 2 measured during the course of incubation experiments
shown in Fig. 5. Carbon-14 activity is given in fraction modern (FM) units. The activity of bulk SOC before incubation is shown for reference
(which do not correspond to time values but are spread horizontally to see detail). Carbon-13 isotopic composition is shown as an bapparent
discrimination,Q i.e. the difference of cumulative respired CO 2 from the isotopic composition of bulk soil organic carbon before respiration
(d 13 C respired CO2 d 13 C bulk SOC ; respired CO 2 consistently more 13 C-depleted than bulk SOC).

J.G. Wynn et al. / Geoderma 131 (2006) 89–109 103
pairs, the cropland sites begin by respiring older
and more depleted SOC than their control counterparts.
Within the agricultural pair (upper and lower
slopes), this observation is more pronounced in
lower slope (depositional) than in upper slope
(erosional) position.
These trends are consistent with the interpretations
observed in the decomposition time series during
incubation, during which the deep samples decompose
more slowly than shallow samples (and have a
lower fraction of labile SOC), and cropland samples
decompose more slowly than their control counterparts.
Thus, both the decomposition time series and
the apparent isotopic discrimination indicate paucity
or depletion of labile pools in deep soil layers (relative
to shallow), in cropland soils (relative to nonagricultural
soils), and in eroded soils (relative to
stable landscape positions).
The hypothesis that Rayleigh fractionation during
decomposition of SOM leads to depth-increasing
d 13 C SOC trends at the control site is further supported
by isotope data from laboratory incubations. Apparent
discrimination at the control site begins with a
depletion of about 0.6x in the 0–20 cm depth
sample and about 3.2x in the 20–40 cm depth
sample (lines fit through the cumulative CO 2
Dd 13 C(t) time series have intercepts of 0.6x and
3.2x). In all samples from both control and
cropland sites, respired CO 2 is always initially
depleted with respect to bulk SOC by at least
0.6x but becomes as much as 6.5x in some
of the deep samples. Differences in these apparent
discrimination values (Dd 13 C=d 13 C CO2 d 13 C SOC )
mostly reflect changes in d 13 C SOC ( 20.7x to
25.9x), while d 13 C CO2 of the initially respired
gas in each remains relatively consistent ( 26.1x to
27.1x). Differences in Dd 13 C appear to reflect the
state of decomposition of SOM (more decomposed
samples from deep soil, cropland sites, and eroded
settings are increasingly b 13 C-distilledQ). These differences
also correlate to differences in the fraction
of labile SOC, f L , modeled from the kinetic time
series which show that deep samples, cropland sites,
and eroded settings decompose more slowly through
the course of 1 year ( f L =0.00772 (Dd 13 C) +0.070;
R 2 =0.63). Apparent discrimination also shows a
strong correlation to bulk SOM turnover time, 1/
k B , calculated from carbon loss during the 1-year
incubation experiments (k B =0.0106 (Dd 13 C)+0.076;
R 2 =0.78), indicating that Dd 13 C may prove to be a
useful tool in describing rates of SOM turnover due
to variations in organic matter quality.
3.7. Changes in organic matter quality indicated by
respired d 13 C CO2 time series
Trends of decreasing respired d 13 C CO2 are consistent
among sample types and roughly linear over
the course of the 1-year incubations, although our 1-
year incubations only consumed between 4% and
14% of the original SOC (Figs. 5 and 6). Decreasing
d 13 C CO2 during incubation without addition of fresh
labile SOC presumably results from increasing
utilization of stable compounds such as lignin, lipids
and cellulose that are typically 13 C-depleted (Boutton,
1996; Deines, 1980). The interpretation of
increasing utilization of stable SOM components is
supported by time series of 14 C activity of respired
CO 2 during the incubation, which in all cases
decreases during the incubations (Fig. 6), in some
cases by over 0.2 FM units.
Trends in d 13 C CO2 of cumulative respired CO 2
during incubation have direct consequences for the
isotopic composition of the residual SOC during the
incubation experiments. We calculate d 13 C SOC of the
remaining SOC using simple mass balance between
the known isotopic value at the beginning of the
experiment, the cumulative isotopic composition of
the evolved CO 2 and the fraction of SOC remaining
during the incubation ( F SOC ):
d 13 C SOC ¼ d13 C 0 þ ð1 F SOC Þd 13
C CO2 cum: resp:
:
F SOC
ð14Þ
The incubation time series demonstrate trends of
increasing utilization of 13 C-depleted SOM, which
leaves the residual pool of SOC enriched in 13 C.
Given the decreasing trend of d 13 C CO2 of respired
CO 2 , we model the isotopic trend of the residual
SOC at the control sites, which increases by about
0.2–0.5x due to increasingly preferential respiration
of
13 C-depleted components (Fig. 7). Although
several biomass decomposition studies (Boutton,
1996; Fernandez et al., 2003) have suggested that
residual accumulation of lignin and other stable

104
J.G. Wynn et al. / Geoderma 131 (2006) 89–109
Control Sites (Goodwin Creek)
Upper Slope
-30
0-20cm depth
20-40cm depth
0.8 0.85 0.9 0.95
F SOC
components in a pool of decomposing SOM should
leave humifying soil 13 C-depleted, our SOM decomposition
study indicates that during later stages of
-22
-24
-26
-28
δ 13 C
Fig. 7. Stable carbon isotope composition of respired CO 2 (dashed
lines), and calculated composition of residual SOC (solid lines)
during 1-year incubation experiments of non-agricultural control
sites (Goodwin Creek), using mass balance and measurements of
flux and isotopic composition of respired CO 2 .
humification, heterotrophic respiration begins to
increasingly consume these 13 C-depleted components.
As this trend continues along the course of
SOM decomposition, the residual products of
decomposition are packed away with SOM age,
producing the observed depth trends of SOM age
and 13 C content.
3.8. Anthropogenic effects on depth trends of SOC
concentration and stable isotopes
Cropland and control sites are very different in
depth trends of 14 C activity of respired CO 2 (FM CO2 ),
13 C apparent discrimination (Dd 13 C) and the fraction
of labile carbon ( f L ) calculated from the kinetic
decomposition model (Fig. 8). FM CO2 and f L at the
control site indicate pronounced incorporation of the
bomb-spike-derived SOC into the labile pool of SOM
and indicate that respiration is derived largely from
this decades-old and labile fraction of SOM. Dd 13 Cof
A.
FM CO2
∆δ 13 C
f L
1.04 1.12 1.2
-6 -4 -2 0
0 0.04 0.08
Sample
Depth
(cm)
B.
10
30
1.2
GC-U
NF-L
-4
NF-U
FM FM
CO
CO 2 NF-U
2
-2
GC-U
δ 13 C
∆
NF-L
1.0
0 0.02 0.04 0.06 0.08 0 0.02 0.04 0.06 0.08
f L
f L
-6
0
1.2
GC-U
NF-U
NF-L
1.0
0 -2 -4 -6
∆δ 13 C
Control Sites (Goodwin Creek)
Upper Slope
Cropland Sites (Nelson Farm)
Upper Slope
Lower Slope
respired bulk SOC ( 14 C)
CO 2 0-20, 20-40
respired
CO 2
Control Sites (Goodwin Creek)
Upper Slope
Cropland Sites (Nelson Farm)
respired
CO 2
Upper Slope
Lower Slope
respired CO 2
0-20, 20-40
solid SOC
0-20, 20-40
Fig. 8. Depth trends of calculated 14 C activity (FM CO2 ) and apparent discrimination (Dd 13 C) of respired CO 2 calculated at the onset of the
incubation experiments, with the fraction of labile carbon ( f L ) modeled from kinetic model of SOC decomposition. A. Plots versus depth. Points
showing solid-phase 14 C activity are shown for reference. B. Cross-plots of isotopic composition of respired CO 2 versus f L and cross-plot of 14 C
and 13 C from respired CO 2 .

J.G. Wynn et al. / Geoderma 131 (2006) 89–109 105
the control site is relatively low at 20 cm and increases
by several per mil with depth, confirming a greater
degree of decomposition in deep samples despite
similarity in the d 13 C CO2 of the respired CO 2 . Both
upper and lower slope positions of the cropland site
show similar slopes of depth trends between 0–20
and 20–40 cm samples to that of the control site
but much greater magnitude of Dd 13 C (simply the
difference between d 13 C SOC and d 13 C CO2 of respired
CO 2 ). These differences are most likely explained
by erosion and removal of the labile fraction from
the surface at the cropland sites, effectively uplifting
older, more stable and more 13 C-distilled SOC from
subsurface horizons to be mixed with the young
and newly forming labile pools. FM CO2 and f L
measurements of the cropland sites are also
consistent with these observations, showing greater
utilization of 14 C-depleted and refractory SOC than
at the control site. These combined results suggest
that Dd 13 C may be a useful indicator of the quality
of SOM (indicated by f L ) but also an indicator of
this mixing process set about by erosion of surface
horizons.
Although d 13 C SOC depth trends from the control
site are consistent with the model of 13 C fractionation
during SOM decomposition and stabilization of
the residually 13 C-enriched SOC in the humifying
SOM (Fig. 9A), the cropland sites demonstrate very
irregular trends, which cannot be explained by the
distillation model (Figs. 2, 4, 9B–E). Potential SOM
changes by agricultural land use that may account
for the lack of a linear relationship between log f SOC
and d 13 C SOC include: changing d 13 C of crop
biomass, removal of labile SOM through enhanced
decomposition or by cropping, and erosion of the
labile organic matter from surface horizons with
rebuilding of a SOC concentration profile over the
eroded profile. Presence of C4 crops or crops with
more enriched d 13 C than the natural C3 woodland
community would shift d 13 C SOC of surface horizons
towards more positive values, particularly in the
most greatly affected surface horizons where SOC
production is highest (Fig. 9B). Although some C4
crops were cultivated at Nelson Farm during 1950–
1954 and this may account for some of the shift
towards more positive d 13 C SOC values at the surface,
we regard this effect as minimal due to its short
duration. In the absence of cropland fertilization,
enhanced removal of labile over recalcitrant SOC
from the surface by preferential decomposition or by
cropping would similarly drive the f SOC to d 13 C SOC
relationship towards more positive d 13 C SOC values
by decreasing the fraction of labile SOC, which has
a d 13 C SOC value similar to that of input from
biomass, d i . This effect is most pronounced in
surface horizons where labile SOC is most abundant
(Fig. 9C). Enhanced removal of labile from the
surface horizon would also produce a bflattenedQ
SOC concentration profile, C(z), by decreasing C s
and effectively increasing z¯, the characteristic production
depth, because the labile pool is more
available to heterotrophic respiration, resulting in
the nonlinear response of f SOC to d 13 C SOC . Removal
of SOC from surface horizons through erosion would
effectively uplift more refractory and 13 C-distilled
carbon from depth to mix with surface f SOC values.
Because the more decomposed SOC at depth has a
13 C-enriched signal, the result is both a shortening
and upward shift of the f SOC to d 13 C SOC relationship
(Fig. 9D). Soil organic matter regeneration following
erosion would shift the deep soil values to the left
(more 13 C-depleted) due to mixing of new fresh
organic matter, with an isotopic composition of d i ,
into what were deeper horizons with highly 13 C-
distilled (enriched) isotopic composition (Fig. 9E).
The negative isotopic shift is smaller in surface
horizons where the effect of dilution with fresh
organic matter is less pronounced.
Data from erosional settings at the cropland site
(ridgetop and both midslope positions) are most
consistent with the modeled effects of erosion and
regeneration, showing a concave-leftward trend of
f SOC to d 13 C SOC (comparison of Figs. 4–8 and 9E).
Although erosion likely accounts for the majority of
the trends observed, the effects of C4 plant cultivation
and removal of labile SOC from surface horizons may
also partially contribute to positive d 13 C SOC shifts,
especially of near-surface samples. Removal of surface
SOC through both cropping and erosion is also
evident from the increase in the apparent f SOC and
flattening of the C(z) concentration profile (Fig. 2).
Erosional removal of SOC in both the upper and
lower slope positions is evident from the return to
more 13 C-depleted values at depth by the admixture of
a more labile component of less decomposed SOM
(Fig. 5).

106
J.G. Wynn et al. / Geoderma 131 (2006) 89–109
A.
1
0.5
δ i
δ f
(z)
f SOC
0.1
0.05
0.01
0.005
surface depth
Kinetic fractionation
during SOC decomposition
-30 -28 -26 -24 -22 -20
δ
13
C
B.
C.
1
0.5
δ i-old
δi-new
1
0.5
f SOC
0.1
0.05
0.01
0.005
Increased δ 13 C of biomass
f SOC
0.1
0.05
0.01
0.005
Enhanced decomposition of
labile SOC
D.
1
0.5
-30 -28 -26 -24 -22 -20
δ i
δ 13 C
E.
0.5
-30 -28 -26 -24 -22 -20
δ 13 C
1 δ i
δ 13 C
f SOC
0.1
0.05
0.01
0.005
Erosion
f SOC
0.1
0.05
0.01
0.005
SOC replacement
after erosion, decomposition
-30 -28 -26 -24 -22 -20
δ 13 C
-30 -28 -26 -24 -22 -20
Fig. 9. Model of agricultural effects on the relationship of SOC concentration to isotopic composition (plots of d 13 C SOC to f SOC as shown in
Figs. 3 and 4 for control and cropland sites). A. Natural trend resulting from kinetic fractionation during decomposition (following the Rayleigh
distillation equation). B. Installation of C4 crops increases d 13 C SOC of surface materials in proportion to new SOC production. C. Enhanced
decomposition of labile SOM due to cropping residually accumulates more decomposed and 13 C-distilled SOC, particularly at the surface. D.
Erosion of surface horizons (without SOC replacement) removes the upper portion of the line in A. Because f SOC is normalized to a value of 1 at
the new surface, the shortened line is also shifted upward. E. Regeneration of a new SOC concentration profile over an eroded profile mixes new
labile 13 C-depleted SOC, diluting the decomposed and 13 C-distilled SOC from the previously eroded soil. Because SOC production is nonlinear
with depth, the lowermost samples are most diluted while surface samples are less diluted, producing a curved relationship.

J.G. Wynn et al. / Geoderma 131 (2006) 89–109 107
4. Conclusions and implications for future work
Stable and radiogenic isotope studies of SOC and
CO 2 respired during laboratory incubation are used
to asses the role of natural processes on the depth
distribution of SOC concentration and its stable
isotopic composition in non-agricultural control sites
and to assess deviations from these natural trends in
paired cropland sites of the Mississippi Basin.
Kinetic fractionation during decomposition of SOM
can be described by a model of Rayleigh distillation
which accounts for stable isotope trends in the
control soil, while the cropland soil deviates strongly
from this relationship due to a variety of effects of
agricultural land use practices, dominated by the
removal of surface SOM by erosion, and partial
replacement with a new profile of fresh labile SOM.
Stable isotope and radiocarbon measurements of
respired CO 2 during incubation time series provides
a means of addressing the role of SOM quality in
determining 13 C and SOC contents and confirms the
hypotheses of fractionation during decomposition by
Rayleigh distillation, during which the 13 C-distilled
products are concentrated in the humifying pool of
SOC. Continuation of this process over time scales
on the order of SOC turnover time produces the
increasing depth trends of stable carbon isotopes,
which are frequently observed in well drained and
productive soils.
Our results have important implications for paleovegetation
studies which attempt to use depth trends
of d 13 C SOC to reconstruct pre-recorded ecological
changes between C3 and C4 ecosystems. Effective
paleovegetation interpretations based on d 13 C SOC
depth profiles must first account for distillative
accumulation of 13 C in humifying SOM and demonstrate
a deviation from the Rayleigh distillation model
of depth-increasing d 13 C SOC , which produces linear
relationships between d 13 C SOC and natural logarithm
of the degree of decomposition of SOM, approximated
by normalized concentration, f SOC .
These results also have implications for isotopic
tracer studies using CO 2 flux from soils. Our
incubation studies suggest that the isotopic signature
of soil-respired CO 2 alone is not very sensitive to land
use, erosion, or the quality of SOM but rather that
apparent discrimination between bulk SOM and
respired CO 2 is a good indicator of the quality of
SOM and of these processes. The degree of decomposition
of SOM can be interpreted from the isotopic
difference between relatively consistent d 13 C CO2 of
respired CO 2 and depth-variable d 13 C SOC (apparent
discrimination, Dd 13 C). This difference increases with
depth (more stabilized decomposition products in
SOM) in all soils studied, including agricultural/nonagricultural
pairs, and a range of slope positions,
reflecting changes in SOM quality with depth.
Apparent discrimination between respired CO 2 and
bulk SOM is also a good indicator of differences in
quality of SOM being sensitive to land use changes
and is higher in agricultural settings and in more
eroded agricultural settings. Apparent discrimination
between respired CO 2 and bulk SOM is also a good
indicator of differences in quality of SOM and is
sensitive to land use changes. A strong correlation
between apparent discrimination, Dd 13 C, and bulk
SOM turnover time, 1/k B , calculated from carbon loss
in 1-year incubation experiments indicates that Dd 13 C
may prove to be a useful tool in describing rates of
SOM turnover (R 2 =0.78).
Acknowledgements
We thank Helaine Markewich, Milan Pavich, Joe
Murphey, and USDA-ARS staff at the National
Sedimentation Laboratory in Oxford, MI for support
and help with site selection. Susan Trumbore, Shihui
Zheng and Doug White provided support with isotope
analyses. Jason Neff, Mark Waldrop and two anonymous
reviewers provided useful comments and
suggestions.
References
2gren, G.J., Bosatta, E., Balesdent, J., 1996. Isotope discrimination
during decomposition of organic carbon: a theoretical analysis.
Soil Sci. Soc. Am. J. 60, 1121–1126.
Ambrose, S.H., Sikes, N.E., 1991. Soil carbon isotope evidence for
Holocene habitat change in the Kenya Rift Valley. Science 253,
1402–1405.
Andrén, O., Paustian, K., 1987. Barley straw decomposition in the
field: a comparison of models. Ecology 68 (5), 1190–1200.
Balesdent, J., Mariotti, A., 1996. Measurement of soil organic
matter turnover using 13 C natural abundance. In: Boutton, T.W.,
Yamasaki, S. (Eds.), Mass Spectrometry of Soils. Marcel
Dekker Inc., New York, pp. 83–102.

108
J.G. Wynn et al. / Geoderma 131 (2006) 89–109
Balesdent, J., Girardin, C., Mariotti, A., 1993. Site-related d 13 Cof
tree leaves and soil organic matter in a temperate forest. Ecology
74 (6), 1713–1721.
Benner, R., Fogel, M.L., Sprague, E.K., Hodson, R.E., 1987.
Depletion of 13 C in lignin and its implications for stable isotope
studies. Nature 327, 708–710.
Bird, M.I., Pousai, P., 1997. Variations of d 13 C in the surface
soil organic carbon pool. Glob. Biogeochem. Cycles 11,
313–322.
Bird, M.I., Šantrlčková, H., Lloyd, J., Veenendaal, E.M., 2001.
Global soil organic carbon pool. In: Schulze, E.-D., et al., (Eds.),
Global Biogeochemical Cycles in the Climate System. Academic
Press, pp. 185–197.
Bird, M.I., Kracht, O., Derrien, D., Zhou, Y., 2003. The effect of
soil texture and roots on the carbon isotope composition of soil
organic carbon. Aust. J. Soil Res. 41, 77–94.
Boutton, T.W., 1996. Stable carbon isotope ratios of soil organic
matter and their use as indicators of vegetation and climate
change. In: Boutton, T.W., Yamasaki, S. (Eds.), Mass Spectrometry
of Soils. Marcel Dekker Inc., New York, pp. 47–82.
Boutton, T.W., Archer, S.R., Midwood, A.J., Zitzer, S.F., Bol, R.,
1998. d 13 C values of soil organic carbon and their use in
documenting vegetation change in a subtropical savanna
ecosystem. Geoderma 82, 5 –41.
Brugnoli, E., Farquhar, G.D., 2000. Photosynthetic fractionation of
carbon isotopes. In: Leegood, S., von Caemmerer, S. (Eds.),
Photosynthesis: Physiology and Metabolism. Kluwer, Dordrecht,
pp. 399–434.
Deines, P., 1980. The isotopic composition of reduced organic
carbon. In: Fritz, P., Fontes, J.C. (Eds.), Handbook of Environmental
Isotope Geochemistry. The Terrestrial Environment vol.
1. Elsevier, New York, pp. 329–406.
Ehleringer, J.R., Hall, A.E., Farquhar, G.D., 1993. Stable
Isotopes and Plant Carbon–Water Relations. Academic Press,
San Diego.
Ehleringer, J.D., Buchmann, N., Flanagan, L.B., 2000. Carbon
isotope ratios in belowground carbon cycle processes. Ecol.
Appli. 10, 412–422.
Elzein, A., Balesdent, J., 1995. Mechanistic simulation of vertical
distribution of carbon concentrations and residence time in soils.
Soil Sci. Soc. Am. J. 59, 1328–1335.
Fernandez, I., Mahieu, N., Cadisch, G., 2003. Carbon isotopic
fractionation during decomposition of plant materials of different
quality. Glob. Biogeochem. Cycles 17 (3), 1075.
Friedli, H., Lftsher, H., Oeschger, H., Siegenthaler, U., Stauffer, B.,
1986. Ice core record of the 13 C/ 12 C ratio of atmospheric CO 2 in
the past two centuries. Nature 324, 237–238.
Harden, J.W., Fries, T.L., Huntington, T.G., 1999a. Mississippi
Basin Carbon Project—upland soil database for sites in Yazoo
Basin, northwestern Mississippi, U.S. Geol. Surv. Open-File
Rep., 99–319.
Harden, J.W., et al., 1999b. Dynamic replacement and loss of soil
carbon on eroding cropland. Glob. Biogeochem. Cycles 13 (4),
885–901.
Harden, J.W., Fries, T.L., Pavich, M.J., 2002. Cycling of beryllium
and carbon through hillslope soils in Iowa. Biogeochem. 60,
317–336.
Hfgberg, P., Ekblad, A., 1996. Substrate-induced respiration
measured in situ in a C3-plant ecosystem using additions of
C4-sucrose. Soil Biol. Biochem. 28, 1131–1138.
Huang, P.M., Schnitzer, M. (Eds.), 1986. Interactions of Soil
Minerals with Natural Organics and Microbes. Soil Sci. Soc.
Am. Spec. Pub. vol. 17. 606 pp.
Huntington, T.G., et al., 1998. Soil, environmental, and watershed
measurements in support of carbon cycling studies in northwestern
Mississippi. Open-File Rep. (U. S. Geol. Surv., 1978),
98–501.
Jobbágy, E.G., Jackson, R.B., 2000. The vertical distribution of soil
organic carbon and its relation to climate and vegetation. Ecol.
Appl. 10 (2), 423–436.
Kaiser, P., Guggenberger, G., Zech, W., 2001. Isotopic fractionation
of dissolved organic carbon in shallow forest soils as affected by
sorption. Eur. J. Soil Sci. 52, 585–597.
Kelly, E.F., Amundson, R.G., Marino, B.D., DeNiro, M.J., 1991.
Stable carbon isotopic composition of carbonate in holocene
grassland soils. Soil Sci. Soc. Am. J. 55, 1651–1658.
Krull, E.S., Skjemstad, J.O., 2002. d 13 C and d 13 N profiles in 14 C-
dated oxisol and vertisols as a function of soil chemistry and
mineralogy. Geoderma 112, 1 – 29.
Krull, E.S., Bestland, E.A., Gates, W.P., 2002. Soil organic
matter decomposition and turnover in a tropical Ultisol:
evidence from d 13 C, d 15 N and geochemistry. Radiocarbon 44,
93–112.
Krull, E.S. et al., in press. Recent vegetation changes in central
Queensland, Australia: evidence from d 13 C and 14 C
analyses of soil organic matter. Geoderma (doi:10.1016/
j.geoderma.2004.09.012).
Lin, G., Ehleringer, J.D., 1997. Carbon isotopic fractionation does
not occur during dark respiration in C3 and C4 plants. Plant
Physiol. 114, 391–394.
Mary, B., Mariotti, A., Morel, J.L., 1992. Use of 13 C variations at
natural abundance for studying the biodegradation of root
mucilage, roots and glucose in soil. Soil Biol. Biochem. 24,
1065–1072.
Melillo, J.M., Aber, J.D., Muratore, J.F., 1982. Nitrogen and lignin
control of hardwood leaf litter decomposition dynamics.
Ecology 63, 621–626.
Melillo, J.M., et al., 1989. Carbon and nitrogen dynamics along the
decay continuum: plant litter to soil organic matter. Plant Soil
115, 189–198.
Natelhoffer, K.J., Fry, B., 1988. Controls on natural nitrogen-15 and
carbon-13 abundances in forest soil organic matter. Soil Sci.
Soc. Am. J. 52, 1633– 1640.
Newson, R., 1995. A canonical model for production and
distribution of root mass in-space and time. J. Math. Biol. 33,
477–488.
O’Brien, B.J., Stout, J.D., 1978. Movement and turnover of soil
organic matter in the study of soil genesis. Soil Sci. Soc. Am.
Proc. 38, 501–506.
Powers, J.S., Schlesinger, W.H., 2002. Geographic and vertical
patterns of stable carbon isotopes in tropical rain forest soils of
Costa Rica. Geoderma 109, 141–160.
Rumpel, C., Kfgel-Knabner, I., Bruhn, F., 2002. Vertical distribution,
age, and chemical composition of organic carbon in

J.G. Wynn et al. / Geoderma 131 (2006) 89–109 109
two forest soils of different pedogenesis. Org. Geochem. 33,
1131–1142.
Sandquist, D.R., Ehleringer, J.D., 1995. Carbon isotope discrimination
in the C4 shrub Atriplex confertifolia (Torr. and Frem.)
Wats. along a salinity gradient. Great Basin Nat. 55, 135–141.
Šantrlčková, H., Bird, M.I., Lloyd, J., 2000. Microbial processes
and carbon-isotope fractionation in tropical and temperate
grassland soils. Funct. Ecol. 14, 108–114.
Schwartz, D., Mariotti, A., Trouve, C., Van Den Borg, K., Guillet,
B., 1992. Étude des profils isotopiques 13 Cet 14 C d’un sol
ferrallitique sableux du littoral congolais: implications sur la
dynamique de la matière organique et l’histoire de la végétation.
C. R. Acad. de Sci., Paris, Séries II 315, 1411 – 1417.
Schweizer, M., Fear, J., Cadisch, G., 1999. Isotopic ( 13 C)
fractionation during plant residue decomposition and its
implications for soil organic matter studies. Rapid Commun.
Mass Spectrom. 13, 1284–1290.
Sharpe, J.M., Harden, J.W., Dabney, S.M., Ojima, D.S., Parton,
W.J., 1998. Implementation of the century ecosystem model for
an eroding hillslope in Mississippi, U.S. Geol. Surv. Open-File
Rep., 98–440.
Sundquist, E.T., et al., 1998. U.S. geological survey Mississippi
Basin Carbon Project science plan. Open-File Rep. (U. S. Geol.
Surv.), 97–177.
Suess, H.E., 1955. Radiocarbon concentration in modern wood.
Science 122, 415–417.
Trumbore, S.E., 2000. Constraints on below-ground carbon cycling
from radiocarbon: the age of soil organic matter and respired
CO 2 . Ecol. Appl. 10, 399–411.
Vogel, J.C., 1978. Recycling of carbon in a forest environment.
Oecol. Plant. 13, 89–94.
Wedin, T.A., Tieszen, L.L., Dewey, B., Pastor, J., 1995. Carbon
isotope dynamics during grass decomposition and soil organic
matter formation. Ecology 76, 1383–1392.
Wynn, J.G., Bird, M.I. and Wong, V., in press. Rayleigh distillation
and the depth profile of 13 C/ 12 C ratios of soil organic carbon
from soils of disparate texture in Iron Range National Park, Far
North Queensland, Australia. Geochim. Cosmochim. Acta.