Methodological comparison for quantitative analysis of fossil and
recently derived carbon in mine soils with high content of aliphatic
, David Sebag
, Guillaume Cailleau
, Jan Frouz
Institute for Environmental Studies, Faculty of Sciences, Charles University, Benátská 2, Prague 120 28, Czech Republic
Laboratoire M2C, UMR 6143 CNRS, Université de Rouen, 76130 Mont-Saint-Aignan, France
Laboratoire HydroSciences Montpellier, UR 050 IRD, Université de Ngaoundéré, Cameroon
Institute of Earth Surface Dynamics, University of Lausanne, Geopolis, CH-1015 Lausanne, Switzerland
Institute of Macromolecular Chemistry, Academy of Sciences of the Czech Republic, Heyrovského nám. 2, Prague 162 06, Czech Republic
Received 23 June 2015
Received in revised form 18 September
Accepted 1 October 2015
Available online 9 October 2015
Soil organic matter
Sedimentary organic matter
In mine soil, quantiﬁcation of soil organic carbon (OC) derived recently from biomass decomposition is
complicated by the presence of fossil (geogenic) C derived from coal, oil shale, or similar material in
the overburden. The only reliable method for such measurement is
C analysis (i.e. radiocarbon dating)
using instrumentation such as accelerator mass spectrometry, which is too expensive for routine labora-
tory analysis. We tested two previously used and two new methods for recent C quantiﬁcation and com-
pared them with
C AMS radiocarbon dating as a reference using a set of soil samples (n= 14) from
Sokolov, Czech Republic: (i)
C isotope ratio composition, (ii) cross polarization magic angle spinning
C nuclear magnetic resonance (CPMAS
C NMR) spectroscopy, (iii) near infrared spectroscopy (NIRS)
coupled with partial least squares regression and (iv) Rock–Eval pyrolysis. Conventional methods for
OC determination (dry combustion, wet dichromate oxidation, loss-on-ignition) were also compared to
quantify any bias connected with their use. All the methods provided acceptable recent carbon estimates
in the presence of mostly aliphatic fossil C from kerogen. However, the most accurate predictions were
obtained with two approaches using Rock–Eval pyrolysis parameters as predictors, namely (i) S
components and (ii) oxygen index (OI). The S
curve approach is based on the lower thermal stability
of recent vs. fossil organic matter. The OI approach corresponded well with
C NMR spectra, which
showed that samples rich in recent C were richer in carboxyl C and O-alkyl C. These two methods showed
the greatest potential as routine methods for recent C quantiﬁcation.
Ó2015 Elsevier Ltd. All rights reserved.
Soil organic matter (SOM) has a signiﬁcant positive impact on
soil functioning by providing chemical energy and essential nutri-
ents (Tiessen et al., 1994). It also contributes to soil cation
exchange capacity, buffering capacity and inﬂuences soil structural
properties, such as stability, water retention and thermal proper-
ties. Different pools of OM contribute differently to these functions
(Baldock and Skjemstad, 2000).
In mine soil (Vindušková and Frouz, 2013) and some natural
soils (Graz et al., 2011; Clouard et al., 2014), OM derived recently
from vegetation is present with fossil OM in coal or kerogen. The
contribution of such geogenic OC (fossil C hereafter) content to
the above soil properties is poorly understood. The main reason
is that quantiﬁcation of fossil C remains a methodological
Moreover, an increase in soil C storage in mine soil has been
studied as a potentially signiﬁcant sink for atmospheric CO
Sperow, 2006; Vindušková and Frouz, 2013). To assess soil quality
development and the sequestration potential of mine soil properly,
a reliable method for quantitative analysis of recently derived C in
the presence of fossil C is needed. Also, the amount of fossil C rein-
troduced into the modern C cycle remains poorly quantiﬁed
(Butman et al., 2014), again due to a lack of a reliable method for
fossil C quantiﬁcation in modern environments.
Different methods based on radiocarbon dating, spectroscopy
and reactivity have been proposed, with a focus on coal C
(reviewed by Ussiri et al., 2014). The only accurate and widely
accepted method is
C analysis, such as accelerator mass
0146-6380/Ó2015 Elsevier Ltd. All rights reserved.
Corresponding author. Tel.: +420 732518457.
E-mail address: email@example.com (O. Vindušková).
Organic Geochemistry 89-90 (2015) 14–22
Contents lists available at ScienceDirect
journal homepage: www.elsevier.com/locate/orggeochem
spectrometry (AMS) radiocarbon dating; however, it is too expen-
sive for routine use. The method is considered to be accurate for
young mine soils developing from overburden material; analysis
of mine soils on old stored topsoil material is more problematic.
The contribution of fossil C to total organic carbon (TOC) content
of mine soil can be substantial. Rumpel et al. (2001) reported that
lignite-derived fossil C accounted for 13–96% of TOC in recultivated
mine soil afforested without topsoil application. In mine soil in
eastern Ohio, the fossil C contribution to TOC ranges between
10% and 17% with topsoil application (Jacinthe and Lal, 2007).
The situation in our study area – Sokolov, Czech Republic – is
more complex since, in addition to small amounts of coal, the over-
burden contains a signiﬁcant amount of sedimentary OM, most of
which is kerogen of algal origin and is predominantly aliphatic,
unlike aromatic coal (Kr
ˇíbek et al., 1998; Frouz et al., 2011). This
kerogen C is ﬁnely dispersed in the parent rock and consequently
also in the soil, so cannot be hand picked, in contrast to fragments
of coal. It has different chemical and physical properties from coal
C and so requires different treatment. This was shown for example
by Hammes et al. (2007), who found that part of both coal and
kerogen C can interfere with various black carbon quantiﬁcation
methods but each material interacts differently with each type of
method. The amount of fossil C in our study area has been esti-
mated as the carbon content of the C horizon (Frouz et al., 2009).
The approach has been used elsewhere (Reintam et al., 2002),
but the accuracy of such estimates is unknown.
Rock–Eval analysis is a pyrolysis method standardized for
source rock characterization and evaluation in oil and gas explo-
ration (Espitalié et al., 1977). Increasingly, it is being applied to soil
(Disnar et al., 2003; Sebag et al., 2006; Hetenyi and Nyilas, 2014)
and recent sediments (Marchand et al., 2008; Delarue et al.,
2013; Siavalas et al., 2013) and has been proposed as a cost effec-
tive tool for bulk characterization of SOM, as well as for determina-
tion of OC source (Carrie et al., 2012) and biogeochemical stability
(Saenger et al., 2013). However, it has not been applied for quanti-
tative analysis of fossil and recent C.
In a previous short communication, we introduced the potential
of near infrared spectroscopy (NIRS) for the quantiﬁcation of recent
and fossil C (Vindušková et al., 2014). The aims of the present study
(i) Quantify the bias connected with soil OC determination
methods (dry combustion, wet dichromate oxidation, loss-
on-ignition) and their correction by way of subtraction of
subsoil C content from the topsoil content.
(ii) Apply methods of recent and fossil C quantiﬁcation proposed
in the literature in a new situation where fossil C is domi-
nantly aliphatic, i.e.
C isotope ratio method (Chabbi et al.,
2006; Ussiri and Lal, 2008) and cross-polarization magic
C nuclear magnetic resonance (CPMAS
NMR) spectroscopy (Rumpel et al., 1998) and test Rock–Eval
pyrolysis for this purpose for the ﬁrst time.
(iii) Compare the approaches, including near NIRS.
While a number of studies have focused on coal C, our study
may be useful for situations where other sources of aliphatic fossil
C similar to kerogen can be found. Such sites could typically be oil
shale mines (Karu et al., 2009), coal mines with overburden con-
taining sedimentary OM, and also soils, and ﬂuvial or estuarine
sediments formed by weathering of C rich rocks (Graz et al., 2011).
2. Geological setting
The study area consists of 7 sites at Podkrušnohorská mine spoil
heap near Sokolov, Czech Republic (50°14
09 N, 12°39
05 E) cover-
ing reclaimed and unreclaimed sites 10–51 yr old. No topsoil had
been applied after heaping the overburden; therefore, the overbur-
den represents the parent substrate for all of the mine soils. In the
study area, the overburden consists of tertiary clays rich in kerogen
of algal origin (Types I and II) which is characteristic of a high ali-
phatic C content. They contain 2–10%, sometimes even 15% OC
ˇíbek et al., 1998).
3. Material and methods
Soil samples (n= 14) were collected from 7 sites and two depths
– topsoil (0–10 cm) and subsoil (40–50 cm). In addition, model
material (two types of claystone differing in TOC content, sub-
bituminous coal and O
material sampled from the fermentation
layer at one of the sites) were also sampled and analysed to pro-
vide insight into the character of the fossil and recent OM. Details
of the study site, sampling, sample preparation, TOC analysis and
C AMS analysis pretreatment are described by Vindušková et al.
(2014). Brieﬂy, soil samples were acid-washed to remove carbon-
ate prior to both
3.1. Total carbon (TC) and total nitrogen (TN)
TOC was calculated as the difference between TC and total inor-
ganic C (TIC), both determined from dry combustion and dry com-
bustion with a TIC module, respectively. TN was determined on dry
samples using an elemental CN analyzer (EA 1108, Carlo Erba
C AMS radiocarbon dating
C AMS radiocarbon dating, the content of recent C in soils
was calculated as follows:
/TOC is the proportion of recent C in the OC pool, calcu-
lated from Eq. (2;Rumpel et al., 2003):
C activity and pMC
C activity of
recent OM. For pMC
, an average value of 115 pMC was used as
per earlier studies (Rumpel et al., 1999; Morgenroth et al., 2004;
Fettweis et al., 2005).
3.3. Loss-on-ignition (LOI)
LOI is a widespread method due to its low cost and simplicity
and allow larger amounts of samples to be tested than with the
dry combustion method. However, its accuracy and OM to C con-
version factors are controversial (Pribyl, 2010). It was performed
by heating four 2 g sub-samples in a mufﬂe furnace for 5 h, after
which the weight of the residual ash was recorded. LOI was calcu-
lated as the wt% lost on ignition.
LOI for the four materials was measured at several tempera-
tures (250, 350, 450, 550, 650, 750 and 850 °C). Soil samples were
then analysed at lower temperature (150, 200, 250, 350 and
550 °C) since LOI of the model materials showed that the ideal
temperature might lie near 250 °C. The mass of material lost was
then converted to % carbon via a conversion factor 1.724 (Pribyl,
3.4. Wet dichromate oxidation (C
(modiﬁed Tyurin method) is a wet chemistry method used
in Central and Eastern Europe. It is similar to the Walkley Black
O. Vindušková et al. /Organic Geochemistry 89-90 (2015) 14–22 15
method (Walkley and Black, 1934), but includes 45 min heating at
125 °C to facilitate the digestion of OC with dichromate. The excess
dichromate is then measured titrimetrically by adding ferrous
ammonium sulfate [Fe(NH
O]. Soil samples, model
materials and coal were analyzed in triplicate, quadruplicate and
C isotope ratio method
C ratio was measured using a stable isotope
ratio mass spectrometer and expressed as:
where the standard is Peedee belemnite. The error was < 0.1‰.
3.6. Solid state
C NMR spectra were measured at 11.7 T using a
Bruker Avance III HD 500 US/WB NMR spectrometer (Karlsruhe,
Germany, 2013) in 7 mm ZrO
rotors at a spinning frequency of
5 kHz. The
C CPMAS spectra were measured with a CP contact
time of 2 ms, TOSS (total suppression of spinning sidebands)
sequence consisting of a set of 180°
C pulse of 8 s, repetition
delay of 5 s, and the number of scans ranged from 2 to 15 k to reach
acceptable signal/noise ratio. During the detection of the
signal the high power dipolar decoupling SPINAL-64 was applied.
The frictional heating of the spinning samples (Brus, 2000) was
mitigated by active cooling, and temperature calibration was per-
formed with Pb(NO
. An external standard (glycine) was used
to calibrate the
C scale (176.03 ppm – low ﬁeld carbonyl signal).
For quantiﬁcation, the spectra were integrated using Origin 8.5
software. The chemical shift regions 0–45, 14–110, 110–160, 160–
220 ppm were assigned to alkyl C, O-alkyl C, aromatic C and car-
boxyl C, respectively (Wilson, 1987).
Schmidt et al. (1996) proposed the following signal
intensity ratio as a ﬁngerprint for brown coal particles in soil
A¼alkyl C þaromatic CðÞ
O-alkyl C þcarboxylic CðÞ
We used an inverse ratio A
that would be positively related to
recent OM in our samples and, since fossil C in our area is assumed
to be dominantly aliphatic (Kr
ˇíbek et al., 1998), we also calculated
a modiﬁed ratio A
leaving out the aromatic C term:
¼O-alkyl C þcarboxylic CðÞ
alkyl C þaromatic CðÞ
¼O-alkyl C þcarboxylic CðÞ
Recent carbon relative content was then regressed against A
Details of NIRS measurements are described by Vindušková
et al. (2014). Brieﬂy, spectra were acquired from 14 soil samples
and 125 artiﬁcial mixtures of overburden, coal and O
Two intensities of grinding were tested (coarse < 2 mm,
ﬁne < 0.125 mm). Partial least squares regression analysis was per-
formed (n= 125 or n= 139) and calibrated to recent C content (wt
%). Recent C values predicted from NIRS included in method com-
parison were obtained by leave-one-out cross validation.
Rock–Eval pyrolysis was performed using a Rock Eval 6 (‘Turbo
model’, Vinci Technologies, France). A detailed description of the
method is given by Espitalié et al. (1977), Lafargue et al. (1998)
and Disnar et al. (2003). Brieﬂy, it involves two successive steps.
The sample is ﬁrst subjected to pyrolysis under an inert (N
sphere with a temperature increase of 30 °C/min in a 200–650 °C
range. The pyrolysis products are measured continuously – the free
) and hydrocarbons released during pyrolysis (S
are measured with ﬂame ionization detection (FID; mg HC/g) and
the oxygenated compounds (S
CO fraction, mg CO
mg CO/g, respectively) are measured with an IR detector. In the
second step, the remaining sample is heated under an O
sphere to oxidize the residual carbon over a range of 400–850 °C.
The evolved CO and CO
are measured (S
CO and S
) and, when
integrated, represent the residual carbon (RC, wt%) fraction. Inte-
gration of S
CO curves gives pyrolysable carbon (PC,
wt%). TOC is calculated as RC + PC. The temperature at the maxi-
mum in the S
curve is called T
The hydrogen index (HI; mg hydrocarbon/g TOC) is calculated
as the amount of hydrocarbons generated during pyrolysis normal-
ized to the amount of OC (Lafargue et al., 1998). Three oxygen
indices are calculated as OICO = S
CO/TOC 100, OICO
TOC 100, OIRE6 = [(16/26 OICO) + (32/44 OICO
)]. Here, OI
refers to OIRE6.
curves were integrated using R software, following the
procedure described by Gillespie et al. (2014). The temperature
regions 205–280, 280–340, 340–400, 400–460, 460–550 and
550–650 °C were assigned to F1a, F1b, F2, F3, F4 and F5, respec-
tively. These indices are comparable to standard indices calculated
from deconvolution used in previous studies (Sebag et al., 2006). A
sum of the ﬁrst four components of the S
curve (F1a + F1b + F2
+ F3) was calculated and used to estimate recent C relative content
using linear regression.
3.9. Method comparison
For the comparison of different methods for recent and fossil C
quantiﬁcation, we converted all approaches to recent C values (wt
%) and compared them with recent C calculated from
carbon dating. The fossil C could be simply calculated as TOC –
recent C values (wt%). Whenever regression was used to predict
C isotope ratio, NMR, Rock–Eval), leave-one-out cross
validation was used in order to simulate the usage of the pro-
posed regression on an unknown sample and not to overestimate
the potential of the method. Negative predictions were converted
to zero. Where any outlier was excluded from the calibration (
isotope ratio method), it was included again in biplots of mea-
sured and predicted values as well as in the t-test and root mean
square error (RMSE) calculation in order not to overrate the
The calculated C
/TOC value was then used to calculate recent
C content using Eq. 1. RMSE was used to assess accuracy of a
are values obtained from
C dating, ^
using the tested method and nis the number of samples.
Furthermore, a paired t-test was also performed for the values
of tested and reference method to assess if there was any consis-
tent bias (constant difference) between the methods (Smith
et al., 1997).
16 O. Vindušková et al. /Organic Geochemistry 89-90 (2015) 14–22
4. Results and discussion
4.1. Total, recent and fossil carbon content, C:N ratio
Radiocarbon analysis showed that fossil C content varied among
sites (Fig. 1). It accounted for 26–99% of TOC in the soils. In topsoil
and subsoil samples it ranged from 26% to 99%, and from 93% to
99% of TOC, respectively. The absolute fossil C content for topsoil
was 2.1–6.0 wt% C and 2.6–5.6 wt% for subsoil. However, fossil C
content did not differ signiﬁcantly between topsoil and subsoil at
the same site (paired t-test, p> 0.05). This conﬁrms previous
assumptions that fossil C in topsoil can be estimated from the OC
content of subsoil (Frouz et al., 2009). The accuracy of the approach
is discussed in Section 4.8.
A negative non-signiﬁcant correlation was found between C:N
ratio (Table 1) and recent C relative content. Ussiri and Lal
(2008) suggested that C:N could be an estimator for coal C content.
They found that soils without coal C content had C:N ranging from
9.6 to 10.6, whereas the tested coal itself had a value of 50.7. Arti-
ﬁcial mixing of coal with the soil caused a corresponding increase
in the value. C:N could not be used as an estimator of relative
recent C content of our samples, since it differed signiﬁcantly
between coal and claystone and even between the two types of
LOI of model materials at different temperature is presented in
Fig. 2. This conﬁrmed that, when it was performed at a tempera-
ture conventionally used for agricultural soils in the Czech Repub-
lic (550 °C; Zbíral, 2002), the weight of both coal and claystone
changed considerably. The changes are most likely related to the
oxidation of OM in both materials. Koide et al. (2011) also recog-
nized that, with 550 °C, the organic portion of biochar is thermally
oxidized and they used this fact to adapt LOI for biochar quantiﬁ-
cation in ﬁeld soils. Moreover, the use of LOI is generally not rec-
ommended for clay-rich soil (Schumacher, 2002), because of the
loss of structurally bound water from claystone during ignition.
Moreover, loss of carbonate from the sample may also add to LOI
values. However, for our samples, loss on ignition from coal and
claystone was reduced signiﬁcantly by decreasing the temperature
to 250 °C.
Amichev (2007) suggested that bituminous coal and recent OM
can be distinguished by heating at 375 °C for 24 h (which oxidizes
recent OM but not coal). Our results show that similar conclusions
can be only site-speciﬁc and that thermal separation from recent
OM is difﬁcult for coals of low rank.
Measurement of LOI for soils at 150, 200, 250, 350 and 550 °C
(Supplementary Fig. S1), showed that the closest estimate of recent
C was obtained at 200 °C. However at this temperature, C-poor
samples were overestimated and C-rich samples underestimated.
Use of the conventional temperature (550 °C) led to an overestima-
tion of recent C by > 8 wt%. Our ﬁndings correspond with a thermo-
gravimetric study of Pallasser et al. (2013), who concluded that LOI
is best constrained to temperatures from 200 to 430 °C, especially
where clay content is high. When using LOI to separate recent and
fossil C, neoformation of thermally resistant aromatic C from
recent OM can be expected due to incomplete combustion, leading
to overestimation of fossil C.
4.3. Wet dichromate oxidation (C
Results from wet dichromate oxidation (modiﬁed Tyurin
method) are presented in Table 1. For the model materials, C
ues were similar to values from dry combustion (TOC), with some-
what higher C
values for O
material and coal; however, this
could be an effect of high error for C
with these samples. For soil
samples, a paired t-test on the C
values and TOC measured from
dry combustion showed that there was no signiﬁcant difference in
the means of the two sets of results (p> 0.05). However, for topsoil
samples alone, paired t-test indicated that C
values were signiﬁ-
cantly lower than TOC (mean difference 0.65). For subsoil samples
was insigniﬁcantly lower. These results indicate that the
wet dichromate oxidation probably did not react with the whole
range of recent OM in the samples; this corresponds with a previ-
ous comparison of titrimetric Tyurin method and dry combustion
(Jankauskas et al., 2006).
C isotope ratio method
C isotope ratio values are presented in Table 1. The variability
between C-rich and C-poor claystone was quite high, but fell into
the range (35‰to 25‰) for kerogens of Type I reported by
Whiticar (1996;35‰to 25‰). The d
C values for kerogens
depend on the OM source and do not change from the time of their
formation (Tissot and Welte, 1978; Meyers, 1994). The value for
the coal (26.9‰) fell into the range reported for coals (22‰to
Even though it seems that C horizons are generally more
C than A horizons (an effect of dominance of lipid-
rich Types I and II kerogen), the correlation between
and recent C relative content for all samples was rather poor (R
0.41). Most likely,
C ratio of fossil C in soil samples was variable
among sites, just like it was between the two claystones as differ-
ent sites were graded using different layers of the Sokolov forma-
tion. This is a source of variability which is not related to recent
OM content and thus precludes the use of the d
C method for
recent OM quantiﬁcation only from the d
C value of topsoil
An improvement in correlation was observed if a ratio of d
for topsoil and of subsoil was used instead. This standardization
removed the inﬂuence of variability among sites and brought
quite good correlation with recent C relative content (C
TOC = 1135.1 (d
C of topsoil/d
C of subsoil) + 1146.1, adj.
0.82, p< 0.05) after removal of one outlier (site U10) indicated
by Cook’s distance > 1. When this relationship was used and ﬁnal
recent C content calculated, quite satisfying predictions could be
obtained from full cross validation (Section 4.8). However, since
Fig. 1. Fossil and recent carbon (wt%) in mine soils near Sokolov, Czech Republic.
Each column represents a soil sample. Samples were collected from seven sites at
two depths – topsoil (T, 0–10 cm) and subsoil (S, 40–50 cm). Site names indicate
reclaimed (R) and unreclaimed (U) sites and their age (yr since reclamation or
overburden heaping at unreclaimed sites).
O. Vindušková et al. /Organic Geochemistry 89-90 (2015) 14–22 17
there is no good explanation for the outlying value of U10 site,
the future applicability of the proposed model is questionable.
An opposite correlation between d
C ratio and
C activity was
found in studies focused on the distinction of recent C and coal C
(Chabbi et al., 2006; Ussiri and Lal, 2008). Chabbi et al. (2006)
found a positive relationship between d
C and lignite content
0.95) and Ussiri and Lal (2008) found a similar, but less precise
0.84) for bituminous coal. The opposite trend in our
samples is most likely due to the different sources of fossil C in our
study and the above studies. Fossil C from coal is relatively
C compared with recent OM, whereas fossil C in
our study area has similar or lower
C content than recent OM
due to its Type I kerogen origin.
Spectra from soil samples are shown in Supplementary Fig. S2
along with the recent C relative content of the respective samples.
The distinct peak in the 0–50 ppm region, representing aliphatic
carbon species was observed in all the spectra. This is in accord
with a study of the claystones of the Sokolov Formation (Kr
et al., 1998), which demonstrated a dominating peak in the alkyl
C region, a smaller peak in the aromatic carbon region and negligi-
ble peaks in the O-alkyl and carboxyl C regions.
Additionally, some of the samples showed a signal at 72 ppm,
indicating the presence of polysaccharides. Furthermore, the signal
at 56 ppm for samples 11 and 3 is characteristic of lignin. Both
polysaccharides and lignin are typical plant litter compounds, indi-
cating recently-derived SOM. The aromatic C region (110–
160 ppm) did not show any clear peaks compared with spectra
from lignite-rich soils (Rumpel et al., 1998; Clouard et al., 2014)
and the relative intensity of aromatic C (Table 2) was even lower
than reported for a claystone sample from the study area (29%;
ˇíbek et al. 1998). This is most likely the effect of a greater relative
importance of the other chemical shift regions. Clearly the relative
intensity of the O-alkyl C and carboxyl C regions increases with
recent C relative content. The effect was conﬁrmed when inverse
A ratios were calculated (Table 2). Considering the unclear signal
in the aromatic C region, an adjusted A
ratio was calculated by
leaving out the aromatic C term.
Both ratios showed comparable correlation with recent relative
C, with p< 0.05 (C
/TOC = 45.61 A
TOC = 82.82 A
0.83). The correlation is even slightly
stronger than a similar correlation found by Rumpel et al. (2000)
between lignite contribution to TOC and the Aratio (R
of the two ratios for recent C prediction is compared with other
approaches in Section 4.8.
Rumpel et al. (1998) reported that the Asignal intensity ratio
was 1.4 and 1.5 for lignite and similar ratios were characteristic
of soils samples dominated by lignite, whereas values for the forest
ﬂoor material were well below 1. The corresponding ratio for our
soils ranged from 1.0 to 6.2 and from 2.4 to 8.8 for the topsoil
and subsoil, respectively. This corresponds well with the assump-
tion that fossil C in our study area is dominated by kerogen of ali-
phatic character (Kr
ˇíbek et al., 1998).
Detailed results from NIRS are described by Vindušková et al.
(2014). Brieﬂy, all models showed a close relationship between
measured and predicted values (R> 0.9). Models calibrated using
both mixtures and soils provided better predictions for soils than
models based only on mixtures. For recent C content, coarsely
ground samples (< 2 mm) produced better estimates than ﬁnely
Sample description and results from dry combustion,
C isotope analysis and wet dichromate oxidation (C
). Age – yr since reclamation or overburden heaping at
unreclaimed sites; R – reclaimed, U – unreclaimed sites.
id Age (yr) Type Depth (cm) TC (%) TN (%) TIC (%) pMC
/TOC (%) C
C(‰) C:N C
(%) C from LOI 200 °C (%)
1 28 R 0–10 9.09 0.55 0.00 48.39 42.08 3.83 29.5 16.6 8.29 1.58
2 28 R 40–50 6.62 0.27 0.08 2.16 1.88 0.12 30.6 24.6 4.70 0.84
3 37 R 0–10 12.6 0.78 0.00 74.52 64.80 8.17 28.1 16.2 12.46 1.95
4 37 R 40–50 4.71 0.21 0.22 2.31 2.01 0.09 29.2 22.4 3.96 0.71
5 10 U 0–10 6.23 0.33 0.45 1.68 1.46 0.08 30.7 19.2 5.30 1.02
6 10 U 40–50 6.84 0.33 0.44 0.92 0.80 0.05 30.8 20.9 5.52 0.91
7 10 R 0–10 4.13 0.13 0.02 10.95 9.52 0.39 28.3 31.7 3.16 1.05
8 10 R 40–50 4.95 0.26 2.12 1.9 1.65 0.05 30.5 19.4 3.83 0.98
9 28 U 0–10 7.29 0.42 0.06 15.26 13.27 0.96 30.5 17.2 6.46 0.97
10 28 U 40–50 4.98 0.25 1.82 1.96 1.70 0.05 30.5 20.3 4.89 0.64
11 19 U 0–10 7.82 0.51 0.00 45.91 39.92 3.12 29.0 15.4 6.91 1.51
12 19 U 40–50 2.78 0.14 0.26 7.59 6.60 0.17 29.0 20.3 2.58 0.86
13 51 U 0–10 14.4 0.94 0.00 84.96 73.88 10.62 28.6 15.2 13.90 2.75
14 51 U 40–50 7.53 0.44 0.15 1.75 1.52 0.11 30.8 16.9 7.30 0.68
1.84 0 104 35.82 29.4 21.6 43.57
Claystone (C-rich) 13.35
0.59 0.07 32.4 23.3 12.76
Claystone (C-poor) 4.63
3.77 0.13 27.8 13.2 3.61
0.80 0 0.12 0.07 26.9 78.2 78.27
Data from Vindušková et al. (2014).
Fig. 2. LOI for model materials at different temperatures.
18 O. Vindušková et al. /Organic Geochemistry 89-90 (2015) 14–22
ground samples (< 0.125 mm). Accuracy of this approach is dis-
cussed further in Section 4.8.
The Rock–Eval parameters are presented in Table 3. Plotting HI
vs. OI allows interpretation of maturity and the origin of the OM
(Fig. 3). Whereas claystone samples were characteristic with high
HI and low OI values, the O
material had low HI and high OI val-
ues. Soil samples then lay on a ‘‘mixing” curve between the clay-
stone and O
material regions and their position correlated with
their recent C relative content. This may be interpreted as a transi-
tion between lipid-rich fossil OM to carbohydrate and lignin-rich
recent OM – see Fig. 4 for comparison of our data with data from
Carrie et al. (2012) who tested Rock–Eval 6 on a set of pure bio-
chemicals and biological standards. It should be noted that this is
a result of a site-speciﬁc situation where the source of fossil OM
is mainly aquatic (algal lipid-rich) whereas the source of recent
OM is terrigenous. Also, the coal had lower OI than the O
in accord with the transformation undergone by OM during
Relative intensity distribution (%) in solid state
C NMR spectra of soil samples and calculated A ratios. C
/TOC – recent C relative content measured with
id ppm Alkyl C O-alkyl C Aromatic C Carboxyl C A
0–45 45–110 110–160 160–220
1 47 25 14 13 1.61 0.81 0.62 42.08
2 65 10 13 13 3.48 0.34 0.29 1.88
3 27 38 23 12 1.00 1.82 1.00 64.80
4 65 9 20 6 5.54 0.24 0.18 2.01
5 70 9 16 5 6.21 0.20 0.16 1.46
6 58 20 12 10 2.37 0.51 0.42 0.80
7 43 25 22 11 1.83 0.82 0.55 9.52
8 58 16 18 8 3.14 0.42 0.32 1.65
9 51 22 16 12 1.97 0.66 0.51 13.27
10 61 14 10 14 2.55 0.46 0.39 1.70
11 37 25 25 13 1.63 1.04 0.61 39.92
12 81 5 9 5 8.83 0.13 0.11 6.60
13 30 40 19 11 0.97 1.67 1.03 73.88
14 60 14 16 9 3.22 0.39 0.31 1.52
(alkyl C + aromatic C)/(O-alkyl C + carboxyl C).
(O-alkyl C + carboxyl C)/(alkyl C + aromatic C).
(O-alkyl C + carboxyl C)/alkyl C.
Rock–Eval pyrolysis parameters.
Sample PC (%) RC (%) TOC (%) MINC (%) HI (mg HC/g TOC) OI (mg CO
/g TOC) T
(°C) F1a + F1b + F2 + F3
1 2.54 5.18 7.72 1.03 348 146 434 42
2 2.10 3.00 5.09 0.88 481 46 434 32
3 2.62 8.05 10.66 1.28 236 181 429 58
4 1.11 2.11 3.22 1.23 393 63 432 40
5 1.91 2.12 4.04 1.91 556 43 434 34
6 2.00 2.51 4.51 2.07 522 37 434 34
7 0.61 2.34 2.95 1.23 205 132 434 39
8 1.15 1.88 3.03 1.79 441 44 436 31
9 2.56 3.15 5.71 1.39 513 79 434 33
10 1.15 1.99 3.13 1.67 426 45 433 32
11 2.09 4.70 6.79 0.80 318 162 432 46
12 0.58 1.58 2.16 0.56 293 94 433 38
13 3.37 10.07 13.44 0.90 240 188 430 59
14 2.93 2.95 5.87 1.00 588 37 434 33
material 9.48 27.64 37.12 1.51 250 174 321 76
Claystone (C-rich) 7.64 3.06 10.70 1.24 856 13 438 29
Claystone (C-poor) 1.10 1.28 2.38 2.04 546 30 434 42
Coal 19.53 41.93 61.46 0.81 371 29 412 60
Fig. 3. HI vs. OI diagram for samples (O
, fermentation layer sample). Bubbles
represent soil samples and their size represents recent C relative content of soil
O. Vindušková et al. /Organic Geochemistry 89-90 (2015) 14–22 19
coaliﬁcation. Diagenesis leads to loss of functional groups and an
increase in aromaticity (Tissot and Welte, 1978).
Fig. 3 highlights the possibility of using OI and HI for the esti-
mation of recent C relative content. Testing of OI and HI as single
or combined predictors showed that OI is a sufﬁcient predictor of
relative recent C and HI does not bring further signiﬁcant improve-
ment of the prediction. Recent C relative content could be well
predicted from OI by exponential regression (C
/TOC = 1.041
0.95, p< 0.05). This approach is discussed further in
Section 4.8. It corresponds well with ﬁndings of Carrie et al.
(2012), who concluded that the S
) are the best discriminants for the source of OM
as terrigenous plant matter typically has much higher OICO values
than aquatic biota, proteins and lipids.
A signiﬁcant negative correlation (Spearman’s r0.59, p0.025)
was found between T
and recent C relative content. This corre-
sponds with the lower T
(320 °C) for the O
material and higher
values measured for the C-poor and C-rich claystone (434 and
438 °C, respectively) and generally with the interpretation of T
as an indicator of maturity (Disnar et al., 2003).
for the O
material was at the lower limit of the range
reported (Disnar et al., 2003) for Ol, Of and some of Oh horizons
(320–390 °C) and even lower when compared with Sebag et al.
(2006) who found a T
of 350 °C for forest litter horizons. They
observed a T
between 380 °C and 400 °C for humic (Oh, O/A)
layers and above 400 °C for organo-mineral (A) horizon samples.
for the coal lay, however, somewhere between those of O
material and claystone and may cause the variability that impaired
the prediction of recent C relative content of soil samples from
The relative contribution of six components to the S
given in Supplementary Table S1. A strong correlation was found
between recent C relative content and a sum of the ﬁrst four compo-
nents given in Table 3 (C
/TOC = 2.6 (F1a + F1b + F2 + F3) 83.9;
0.88). Use of this relationship for recent C quantiﬁcation is dis-
cussed in Section 4.8.
Carrie et al. (2012) suggested that the shape of the S
(S2a/S2b ratio) could be used as an indicator of the dominant OM
source in fresh material with higher values (> 2) indicative of aqua-
tic biota and smaller values (< 0.3) indicative of terrigenous plant
material. Conversely, in our study, the mean S2a/S2b ratio [calcu-
lated as (F1a + F2a + F3)/(F4 + F5)] for C and A horizons was 0.08
and 0.27, respectively, suggesting that terrigenous OM in our soils
led to a higher ratio, i.e. relatively more hydrocarbons were
released during pyrolysis below 400 °C. This may be explained by
the greater thermal stability of the OM present as kerogen than
recent soil OM. We also conclude that factors other than OM source
affect the shape of the S
pyrogram and conclusions from the anal-
ysis of fresh material should be transferred to soils and sediments
with caution. Copard et al. (2006) found that the F1 and F6 clusters
deconvolution were present in bedrock but absent from lit-
ter and proposed their use as markers of fossil OM (FOM) in mod-
ern environments. The integration method used in their study was
slightly different (F1 and F6 were deﬁned as clusters with T
and 550 °C); however, we found no signal from a labile FOM frac-
tion corresponding to their F1 cluster in our samples of unweath-
ered claystone. On the other hand, their proposed refractory F6
cluster was in our study represented by F4 and F5 (460–550 °C
and 550–650 °C, respectively). The strong correlation between
the sum of the F1 to F3 components and recent C relative content
fully complements the conclusion that F4 and F5 are indicative of
FOM. Again, this is due to the fact that relatively more hydrocar-
bons are released during thermal cracking at lower temperatures
(< 460 °C) from recent (less-altered) OM than from FOM, as can
be seen also in the F4 + F5 values for the O
material and claystones
(pure recent and fossil OM, respectively). Also, according to previ-
ous studies, the latter two components are relatively scarce in soils
compared with fossil sediments (Sebag et al., 2006).
We have shown that, apart from the OI, S
ters can also be used for quantitative estimation of recent (or fossil)
OM in the study soils.
4.8. Method comparison
Estimates obtained from twelve methods are plotted against
recent carbon obtained from
C dating in Fig. 5. The lines in the
biplots are lines of equality on which all points would lie if the
tested method gave exactly the same results as
Biplots for TOC, C
and LOI 550 °C indicate that all of these
three conventional methods used for soil OC determination in agri-
cultural soil lead to overestimation, with the highest error for LOI
550 °C (RMSE 8.70). Its extent is clearly evident when plotted as
the mean difference between the tested method and reference
(called bias when signiﬁcant; Supplementary Fig. S3). The mean
bias was 8.23, 4.75 and 4.39% C for LOI, TOC and C
However, as shown in the second column of biplots, a correction
for these three methods with acceptable level of accuracy was
obtained by subtracting subsoil content from topsoil content. This
time, corrected LOI 550 °C showed the greatest accuracy (RMSE
1.21) followed by C
(RMSE 1.62) and TOC (RMSE 2.07). The mean
difference was reduced to 0.06, 0.16 and 0.37 for LOI, TOC and
, respectively, and was then not signiﬁcant for any of the three
methods (paired t-test, p< 0.05). However, laborious and costly,
sampling and analysis of the subsoil is required for such correction
with subsoil content.
Another way to adapt the conventional methods is to decrease
the heating temperature for LOI to 200 °C. Such a modiﬁcation pro-
duces acceptable but less accurate results (RMSE 2.64) vs. subsoil
corrected methods; however, sampling of subsoil is not required.
Regression based on d
C topsoil/subsoil ratio gave estimates with
accuracy similar to the subsoil corrected conventional methods
(RMSE 1.73). Since application of this method also requires subsoil
Fig. 4. HI vs. OI diagram for samples measured here and samples of pure
biochemicals and biological standards measured by Carrie et al. (2012). Biostan-
dards: 1, copepods; 2, phytoplankton; 3, needles; 4, bark.
20 O. Vindušková et al. /Organic Geochemistry 89-90 (2015) 14–22
sampling and – in addition – is more costly, we cannot recommend
it as an alternative to conventional methods.
NIRS of coarsely ground samples produced estimates with accu-
racy similar to LOI 200 °C (RMSE 2.70). On the other hand, NIRS of
ﬁnely ground samples leads to signiﬁcant overestimation (paired t-
test, p< 0.05, mean difference 1.48, see Supplementary Fig. S3).
Like LOI 200 °C, NIRS does not require subsoil sampling and may
be a cost effective tool for analysis of a large number of samples.
NMR was used to estimate recent carbon as follows. Recent C
relative content was calculated from inverse Aratios using linear
relationships described in Section 4.5 and then used to produce
estimates of recent C from Eq. 1. Prediction using the A
does not account for an aromatic C term yielded more accurate
results (RMSE 0.77; Fig. 5) than the A
ratio (RMSE 0.89, data not
presented). This may indicate that the aliphatic structure of kero-
gen is the major source of fossil C in our samples and the contribu-
tion from coal is not that signiﬁcant.
Rock–Eval OI was used to estimate recent C as follows. Recent C
relative content was calculated from OI using the exponential rela-
tionship described in Section 4.7 and then used to produce esti-
mates of recent carbon from Eq. 1. This approach produced the
second most accurate predictions (RMSE 0.69).
Similarly, a sum of ﬁrst four components of the S
+ F1b + F2 + F3) was used to estimate recent C, ﬁrst by using a lin-
ear relationship (Section 4.7) to calculate recent C relative content
and then estimating recent carbon from Eq. 1. This approach pro-
duced the most accurate predictions (RMSE 0.60).
Based both on t-test results and RMSE, both Rock–Eval
approaches were the most accurate. Even though NMR gave also
very precise predictions, it is a laborious and costly method,
requires long measurement times and highly specialized instru-
mentation and is therefore not suited to routine analysis of mine
soil samples. Rock–Eval on the other hand is relatively simple
and rapid. Use of the OI as an indicator of recent OM is in accord
with the NMR analysis, which conﬁrmed that recent OM is rich
in carbohydrates and lignin, while fossil OM contains predomi-
nantly aliphatic and aromatic structures. Use of the Rock–Eval S
curve is based on lower thermal stability of recent OM than fossil
Finally, we would like to highlight that, due to lack of data from
additional sites, truly independent evaluation was not possible. It
is also important to note that the regression equations used for
NMR and Rock–Eval are speciﬁc to the type of fossil OM in our
study area. Therefore, to apply such an approach in other areas, a
site-speciﬁc equation would need to be developed.
Fossil carbon accounted for 13–99% of TOC in study soil sam-
ples, which is equivalent to 2–6% C content. Subtracting the C con-
tent of subsoil (50 cm deep) from the topsoil content provided an
acceptable estimate of recent C. This could be applied to C content
from dry combustion, wet dichromate oxidation and LOI (550 °C).
In the study area, it is inappropriate to use these methods for soil
quality or C sequestration assessment without such correction.
Of all of methods, Rock–Eval proved to most accurately deter-
mine the actual recent C content of each soil sample over a wide
range of recent C concentration. It is far simpler than other meth-
ods such as
C dating or NMR, which require more sophisticated
Financial support provided by the The Charles University Grant
Agency (Grant No. 922513) is gratefully acknowledged. We wish to
thank T. Adatte for Rock–Eval analysis. We thank two anonymous
reviewers for constructive suggestions.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.orggeochem.
Associate Editor—I. Kögel-Knabner
B.Y., Amichev, 2007. Biogeochemistry of Carbon on Disturbed Forest Landscapes
(Ph.D. dissertation). University Libraries, Virginia Polytechnic Institute and
State University, Department of Forestry, Blacksburg, VA.
Baldock, J., Skjemstad, J., 2000. Role of the soil matrix and minerals in protecting
natural organic materials against biological attack. Organic Geochemistry 31,
Brus, J., 2000. Heating of samples induced by fast magic-angle spinning. Solid State
Nuclear Magnetic Resonance 16, 151–160.
Butman, D.E., Wilson, H.F., Barnes, R.T., Xenopoulos, M.A., Raymond, P.A., 2014.
Increased mobilization of aged carbon to rivers by human disturbance. Nature
Geoscience 8, 112–116.
Carrie, J., Sanei, H., Stern, G., 2012. Standardisation of Rock–Eval pyrolysis for the
analysis of recent sediments and soils. Organic Geochemistry 46, 38–53.
Chabbi, A., Rumpel, C., Grootes, P.M., Mariotti, A., Hüttl, R.F., 2006. Isotopic tracers
for the analysis of vegetation-derived organic matter in lignite-containing soils
and sediments along a transect ranging from a forest soil to submerged lake
sediment. Organic Geochemistry 37, 740–753.
Clouard, M., Criquet, S., Borschneck, D., Ziarelli, F., Marzaioli, F., Balesdent, J., Keller,
C., 2014. Impact of lignite on pedogenetic processes and microbial functions in
Mediterranean soils. Geoderma 232–234, 257–269.
Copard, Y., Di-Giovanni, C., Martaud, T., Albéric, P., Olivier, J.-E., 2006. Using Rock–
Eval 6 pyrolysis for tracking fossil organic carbon in modern environments:
implications for the roles of erosion and weathering. Earth Surface Processes
and Landforms 31, 135–153.
Delarue, F., Disnar, J.-R., Copard, Y., Gogo, S., Jacob, J., Laggoun-Défarge, F., 2013. Can
Rock–Eval pyrolysis assess the biogeochemical composition of organic matter
during peatiﬁcation? Organic Geochemistry 61, 66–72.
Fig. 5. Comparison of C from tested methods with recent C (C
) obtained from
radiocarbon method. TOC, dry combustion; LOI, loss-on-ignition; C
, wet dichro-
mate (modiﬁed Tyurin) method; d
C ratio, regression based on topsoil/subsoil d
ratio; NIRS, near infrared spectroscopy after coarse grinding; cor., correction by
subtraction of subsoil from topsoil; OI, regression based on oxygen index; S
regression based on F1a + F1b + F2 + F3 components of S
curve. Data for NIRS are
from Vindušková et al. (2014).
O. Vindušková et al. /Organic Geochemistry 89-90 (2015) 14–22 21
Disnar, J.R., Guillet, B., Keravis, D., Di-Giovanni, C., Sebag, D., 2003. Soil organic
matter (SOM) characterization by Rock–Eval pyrolysis: scope and limitations.
Organic Geochemistry 34, 327–343.
Espitalié, J., Laporte, J., Madec, M., Marquis, F., Leplat, P., Paulet, J., Boutefeu, A., 1977.
Rapid method for source rocks characterization and for determination of
petroleum potential and degree of evolution. Revue Institute Français du Pétrole
Fettweis, U., Bens, O., Huttl, R.F., 2005. Accumulation and properties of soil organic
carbon at reclaimed sites in the Lusatian lignite mining district afforested with
Pinus sp. Geoderma 129, 81–91.
Frouz, J., Pizl, V., Cienciala, E., Kalcik, J., 2009. Carbon storage in post-mining forest
soil, the role of tree biomass and soil bioturbation. Biogeochemistry 94, 111–
Frouz, J., Cajthaml, T., Kr
ˇíbek, B., Schaeffer, P., Bartuška, M., Galertová, R., Rojík, P.,
˚fek, V., 2011. Deep, subsurface microﬂora after excavation respiration and
biomass and its potential role in degradation of fossil organic matter. Folia
Microbiologica 56, 389–396.
Gillespie, A.W., Sanei, H., Diochon, A., Ellert, B.H., Regier, T.Z., Chevrier, D., Dynes, J.J.,
Tarnocai, C., Gregorich, E.G., 2014. Perennially and annually frozen soil carbon
differ in their susceptibility to decomposition: analysis of subarctic earth
hummocks by bioassay, XANES and pyrolysis. Soil Biology and Biochemistry 68,
Graz, Y., Di-Giovanni, C., Copard, Y., Elie, M., Faure, P., Laggoun Defarge, F., Lévèque,
J., Michels, R., Olivier, J.E., 2011. Occurrence of fossil organic matter in modern
environments: optical, geochemical and isotopic evidence. Applied
Geochemistry 26, 1302–1314.
Hammes, K., Schmidt, M.W.I., Smernik, R.J., Currie, L.A., Ball, W.P., Nguyen, T.H.,
Louchouarn, P., Houel, S., Gustafsson, Ö., Elmquist, M., Cornelissen, G.,
Skjemstad, J.O., Masiello, C.A., Song, J., Peng, P., Mitra, S., Dunn, J.C., Hatcher,
P.G., Hockaday, W.C., Smith, D.M., Hartkopf-Fröder, C., Böhmer, A., Lüer, B.,
Huebert, B.J., Amelung, W., Brodowski, S., Huang, L., Zhang, W., Gschwend, P.M.,
Flores-Cervantes, D.X., Largeau, C., Rouzaud, J.-N., Rumpel, C., Guggenberger, G.,
Kaiser, K., Rodionov, A., Gonzalez-Vila, F.J., Gonzalez-Perez, J.A., de la Rosa, J.M.,
Manning, D.A.C., López-Capél, E., Ding, L., 2007. Comparison of quantiﬁcation
methods to measure ﬁre-derived (black/elemental) carbon in soils and
sediments using reference materials from soil, water, sediment and the
atmosphere. Global Biogeochemical Cycles 21, 3016.
Hetenyi, M., Nyilas, T., 2014. Soil organic matter characterization Using S3 and S4
Signals from Rock–Eval Pyrolysis. Pedosphere 24, 563–574.
Jacinthe, P.-A., Lal, R., 2007. Carbon storage and minesoil properties in relation to
topsoil application techniques. Soil Science Society of America Journal 71,
Jankauskas, B., Slepetiene, A., Jankauskiene, G., Fullen, M.A., Booth, C.A., 2006. A
comparative study of analytical methodologies to determine the soil organic
matter content of Lithuanian Eutric Albeluvisols. Geoderma 136, 763–773.
Karu, H., Szava-Kovats, R., Pensa, M., Kull, O., 2009. Carbon sequestration in a
chronosequence of Scots pine stands in a reclaimed opencast oil shale mine.
Canadian Journal of Forest Research-Revue Canadienne De Recherche Forestiere
Koide, R.T., Petprakob, K., Peoples, M., 2011. Quantitative analysis of biochar in ﬁeld
soil. Soil Biology and Biochemistry 43, 1563–1568.
ˇíbek, B., Strnad, M., Bohác
ˇek, Z., Sy
´korová, I., C
ˇejka, J., Sobalík, Z., 1998.
Geochemistry of Miocene lacustrine sediments from the Sokolov Coal Basin
(Czech Republic). International Journal of Coal Geology 37, 207–233.
Lafargue, E., Marquis, F., Pillot, D., 1998. Rock–Eval 6 applications in hydrocarbon
exploration, production, and soil contamination studies. Revue Institute
Français du Pétrole 53, 421–437.
Marchand, C., Lallier-Vergès, E., Disnar, J.-R., Kéravis, D., 2008. Organic carbon
sources and transformations in mangrove sediments: a Rock–Eval pyrolysis
approach. Organic Geochemistry 39, 408–421.
Meyers, P.A., 1994. Preservation of elemental and isotopic source identiﬁcation of
sedimentary organic matter. Chemical Geology 114, 289–302.
Morgenroth, G., Kretschmer, W., Scharf, A., Uhl, T., Fettweis, U., Bens, O., Huttl, R.F.,
C measurement of soil in post-mining landscapes. Nuclear Instruments
& Methods in Physics Research Section B: Beam Interactions with Materials and
Atoms 223, 568–572.
Pallasser, R., Minasny, B., McBratney, A.B., 2013. Soil carbon determination by
thermogravimetrics. PeerJ 1, e6.
Pribyl, D.W., 2010. A critical review of the conventional SOC to SOM conversion
factor. Geoderma 156, 75–83.
Reintam, L., Kaar, E., Rooma, I., 2002. Development of soil organic matter under pine
on quarry detritus of open-cast oil-shale mining. Forest Ecology and
Management 171, 191–198.
Rumpel, C., Skjemstad, J.O., Knicker, H., Kögel-Knabner, I., Hüttl, R.F., 2000.
Techniques for the differentiation of carbon types present in lignite-rich mine
soils. Organic Geochemistry 31, 543–551.
Rumpel, C., Balesdent, J., Grootes, P., Weber, E., Kogel-Knabner, I., 2003.
Quantiﬁcation of lignite- and vegetation-derived soil carbon using
measurements in a forested chronosequence. Geoderma 112, 155–166.
Rumpel, C., Janik, L.J., Skjemstad, J.O., Kogel-Knabner, I., 2001. Quantiﬁcation of
carbon derived from lignite in soils using mid-infrared spectroscopy and partial
least squares. Organic Geochemistry 32, 831–839.
Rumpel, C., Knicker, H., Kögel-Knabner, I., Skjemstad, J.O.O., Huttl, R.F., 1998. Types
and chemical composition of organic matter in reforested lignite-rich mine
soils. Geoderma 86, 123–142.
Rumpel, C., Kögel-Knabner, I., Hüttl, R., 1999. Organic matter composition and
degree of humiﬁcation in lignite-rich mine soils under a chronosequence of
pine. Plant and Soil 213, 161–168.
Saenger, A., Cécillon, L., Sebag, D., Brun, J.-J., 2013. Soil organic carbon quantity,
chemistry and thermal stability in a mountainous landscape: a Rock–Eval
pyrolysis survey. Organic Geochemistry 54, 101–114.
Schmidt, M.W.I., Knicker, H., Hatcher, P.G., Kögel-Knabner, I., 1996. Impact of brown
coal dust on the organic matter in particle-size fractions of a Mollisol. Organic
Geochemistry 25, 29–39.
Schumacher, B.A., 2002. Methods for the Determination of Total Organic Carbon
(TOC) in Soils and Sediments. Ecological Risk Assessment Support Center Ofﬁce
of Research and Development US, Environmental Protection Agency (ERASC),
Sebag, D., Disnar, J.R., Guillet, B., Di Giovanni, C., Verrecchia, E.P., Durand, A., 2006.
Monitoring organic matter dynamics in soil proﬁles by ‘‘Rock–Eval pyrolysis”:
bulk characterization and quantiﬁcation of degradation. European Journal of
Soil Science 57, 344–355.
Siavalas, G., Werner, D., Karapanagioti, H.K., Bowler, B.F.J., Manning, D.A.C.,
Christanis, K., 2013. Comparison of methods for the characterization and
quantiﬁcation of carbon forms in estuarine and marine sediments from coal
mining regions. Organic Geochemistry 59, 61–74.
Smith, P., Smith, J.U., Powlson, D.S., McGill, W.B., Arah, J.R.M., Chertov, O.G.,
Coleman, K., Franko, U., Frolking, S., Jenkinson, D.S., Jensen, L.S., Kelly, R.H.,
Klein-Gunnewiek, H., Komarov, A.S., Li, C., Molina, J.A.E., Mueller, T., Parton, W.J.,
Thornley, J.H.M., Whitmore, A.P., 1997. A comparison of the performance of nine
soil organic matter models using datasets from seven long-term experiments.
Geoderma 81, 153–225.
Sperow, M., 2006. Carbon sequestration potential in reclaimed mine sites in seven
east-central states. Journal of Environmental Quality 35, 1428–1438.
Tiessen, H., Cuevas, E., Chacon, P., 1994. The role of soil organic matter in sustaining
soil fertility. Nature 371, 783–785.
Tissot, B., Welte, D., 1978. Kerogen: composition and classiﬁcation. In: Petroleum
Formation and Occurrence. Editions Technip, Paris, pp. 123–147.
Ussiri, D.A.N., Jacinthe, P.-A., Lal, R., 2014. Methods for determination of coal carbon
in reclaimed minesoils: A review. Geoderma 214–215, 155–167.
Ussiri, D.A.N., Lal, R., 2008. Method for determining coal carbon in reclaimed
minesoils contaminated with coal. Soil Science Society of America Journal 72,
Vindušková, O., Dvor
ˇek, V., Prohasková, A., Frouz, J., 2014. Distinguishing recent
and fossil organic matter – a critical step in evaluation of post-mining soil
development – using near infrared spectroscopy. Ecological Engineering 73,
Vindušková, O., Frouz, J., 2013. Soil carbon accumulation after open-cast coal and oil
shale mining in Northern Hemisphere: a quantitative review. Environmental
Earth Sciences 69, 1685–1698.
Walkley, A., Black, I.A., 1934. An examination of the Degtjareff method for
determining soil organic matter, and a proposed modiﬁcation of the chromic
acid titration method. Soil Science 37, 29–38.
Whiticar, M.J., 1996. Stable isotope geochemistry of coals, humic kerogens and
related natural gases. International Journal of Coal Geology 32, 191–215.
Wilson, M.A., 1987. NMR Techniques and Applications in Geochemistry and Soil
Chemistry. Pergamon Press, Oxford.
Zbíral, J., 2002. Analy
˚d III: Jednotné Pracovní Postupy. Ústr
ˇední kontrolní a
zkušební ústav zeme
22 O. Vindušková et al. /Organic Geochemistry 89-90 (2015) 14–22