Soil organic carbon stocks in southeast Germany(Bavaria) as affected by land use, soil type and samplingdepthMART IN WIE SME I ER * , P ETER SP ORLE IN † , UWE GEUß † , EDZARD HANGEN † ,

S T EPHAN HAUG ‡ , ARTHUR RE I SCHL † , B ERND SCH ILL ING † , MARG IT V ON L U TZOW*

and INGRID KOGEL-KNABNER*§

*Chair of Soil Science, Department of Ecology and Ecosystem Management, Technische Universitat Munchen, D-85350, Freising-

Weihenstephan, Germany, †Bayerisches Landesamt fur Umwelt (LfU), D-95030, Hof, Germany, ‡TUM|Stat, Zentrum

Mathematik, Technische Universitat Munchen, D-85748, Garching, Germany, §Institute for Advanced Study, Technische

Universitat Munchen, Lichtenbergstr. 2a, D-85748, Garching, Germany

Abstract

Precise estimations of soil organic carbon (SOC) stocks are of decided importance for the detection of C sequestration

or emission potential induced by land use changes. For Germany, a comprehensive, land use–specific SOC data set

has not yet been compiled. We evaluated a unique data set of 1460 soil profiles in southeast Germany in order to cal-

culate representative SOC stocks to a depth of 1 m for the main land use types. The results showed that grassland

soils stored the highest amount of SOC, with a median value of 11.8 kg m�2, whereas considerably lower stocks of

9.8 and 9.0 kg m�2 were found for forest and cropland soils, respectively. However, the differences between exten-

sively used land (grassland, forest) and cropland were much lower compared with results from other studies in cen-

tral European countries. The depth distribution of SOC showed that despite low SOC concentrations in A horizons of

cropland soils, their stocks were not considerably lower compared with other land uses. This was due to a deepening

of the topsoil compared with grassland soils. Higher grassland SOC stocks were caused by an accumulation of SOC

in the B horizon which was attributable to a high proportion of C-rich Gleysols within grassland soils. This demon-

strates the relevance of pedogenetic SOC inventories instead of solely land use–based approaches. Our study indi-

cated that cultivation-induced SOC depletion was probably often overestimated since most studies use fixed depth

increments. Moreover, the application of modelled parameters in SOC inventories is questioned because a calculation

of SOC stocks using different pedotransfer functions revealed considerably biased results. We recommend SOC

stocks be determined by horizon for the entire soil profile in order to estimate the impact of land use changes

precisely and to evaluate C sequestration potentials more accurately.

Keywords: carbon sequestration, land use change, pedotransfer function, soil organic matter, topsoil deepening

Received 22 February 2012; revised version received 22 February 2012 and accepted 16 March 2012

Introduction

Soil organic matter is not only the key component for

sustainable soil fertility and productivity but has become

increasingly important as the largest terrestrial C pool in

view of climate change. Accurate estimation of soil

organic carbon (SOC) stocks is thus essential for the cal-

culation of future climate scenarios and to detect the

potential for C sequestration or emission. In this context,

land use changes are of vital importance as a conversion

of natural or extensively used areas (e.g. grassland, for-

est) to intensively used croplands (e.g. for production of

food and energy crops) is reported to be followed by a

strong decline of SOC and vice versa (Guo & Gifford,

2002; Murty et al., 2002; Houghton & Goodale, 2004;

Eglin et al., 2010; Poeplau et al., 2011). The importance of

a precise estimation of SOC stocks is recognized by most

countries and several institutions: for instance, the

European Union (Stolbovoy et al., 2007) and the Intergov-

ernmental Panel on Climate Change (IPCC, 2000, 2003).

However, most attempts to calculate SOC stocks have

failed to include the entire soil profile. Although the

amount of SOC stored in subsoils (below the top

30 cm) exceeds 50% in most ecosystems (Batjes, 1996;

Jobbagy & Jackson, 2000; Harrison et al., 2011; Rumpel

& Kogel-Knabner, 2011), the majority of SOC studies

and inventories only incorporate the first 30 cm of the

soils (e.g. IPCC, 2003; Janssens et al., 2005; Jones et al.,

2005; Smith et al., 2005; Stolbovoy et al., 2007). From

another point of view, subsoil SOC stocks become addi-

tionally important because they primarily constitute theCorrespondence: Martin Wiesmeier, tel. + 49 0 8161 71 3679,

fax + 49 0 8161 71 4466, e-mail: wiesmeier@wzw.tum.de

© 2012 Blackwell Publishing Ltd 1

Global Change Biology (2012), doi: 10.1111/j.1365-2486.2012.02699.x

intermediate and passive soil organic matter pool (von

Lutzow et al., 2008). Under global warming, the tem-

perature dependence of these pools will be decisive for

SOC stock changes (von Lutzow & Kogel-Knabner,

2009). Hence, an accurate estimate of the size of these

pools under different land uses is required.

A further shortcoming of many SOC inventories is

that soil parameters, which are necessary to calculate

SOC stocks (SOC concentration, bulk density, stone

content), are often estimated (using pedotransfer func-

tions and regression analyses) and not measured

because large-scale soil inventories are time-consuming

and hence are conducted with soil augers instead of by

sampling soil profiles. Due to the high variability of

these parameters, direct measurements are essential for

unbiased SOC estimations (Schrumpf et al., 2011).

For Germany, only regional SOC inventories with a

limited amount and quality of data have been available

until now. In their review of the status quo of organic car-

bon in soils of Germany, Prechtel et al. (2009) also point to

a lack of SOCdata, particularly for agricultural soils. They

further suggest incorporating subsoil SOC stocks in

future soil inventories. The accuracy of available studies

is limited because only mean values for important soil

parameters were used (Neufeldt, 2005) or soil samples

were taken within constant depth increments (Baritz &

Strich, 2000) instead of sampling soil horizons that may

lead to inaccurate results if only few depth increments

with high thicknesses are used (Gruneberg et al., 2010).

Due to these reasons, precise, land use–specific estima-

tions of total SOC stocks are missing for Germany as well

as formany other European countries.

For this study, a comprehensive data set of 1460 soil

profiles within the state of Bavaria in southeast Ger-

many was used to calculate SOC stocks for the preva-

lent land uses (Fig. 1). This unique data set was

compiled by the Bavarian Environment Agency and the

Bavarian State Institute for Forestry. Each soil profile

was sampled by horizon down to a depth of 1 m or to

the parent material and analyzed for the relevant soil

parameters. The objectives of the study were to

1 Determine representative SOC stocks for cropland,

grassland and forest soils as well as for different soil

classes.

2 Elucidate the vertical distribution of SOC within the

soil profiles and related pedogenetic processes that

impact SOC distribution with depth.

3 Test the hypothesis that land use is the dominant fac-

tor controlling SOC storage and that cultivation of

soils is associated with a pronounced decline of SOC.

Materials and methods

Study area

The state of Bavaria, with an area of 70 550 km2, is located

in southeast Germany and comprises various landscapes.

Fig. 1 Land use map of Bavaria with the location of the soil profiles.

© 2012 Blackwell Publishing Ltd, Global Change Biology, doi: 10.1111/j.1365-2486.2012.02699.x

2 M. WIESMEIER et al.

The northwestern part of Bavaria is dominated by the south-

ern German escarpment landscape that adjoins low mountain

ranges of the Bohemian Massif in the east. Southwards the

Molasse basin ascends to the mountain range of the Alps at

the southern border of Bavaria. Elevation ranges between 107

and 2962 m above sea level. Due to its location in central

Europe, Bavaria exhibits a suboceanic climate that is charac-

terized by a transitional situation between a maritime climate

in the northwest and subcontinental influences in the east.

Mean annual temperature and precipitation from the escarp-

ment landscape in the northwest to the Alps in the south

range between 9° and 4° C and 550 and 2500 mm, respec-

tively. Dominant soil classes are soils with well-developed B

horizons (Cambisols) at 45% of the total area, soils with initial

soil formation (Leptosols, Regosols) with 14% and soils with

water stagnation (Stagnosols, Albeluvisols, Planosols) with

13% according to the German soil systematic (AD-HOC AG

Boden, 2005) and the equivalent Reference Soil Groups of the

WRB system (IUSS Working Group WRB, 2006). Cropland

accounts for 34% of the total land area of Bavaria, grassland

for 16%, forest for 35% and other uses for 15% (Fig. 1).

Selection of SOC data

Available data from different soil surveys and permanent soil

observation sites in Bavaria overseen by the Bavarian Environ-

ment Agency (LfU) and the Bavarian State Institute for

Forestry (LWF) were screened in order to compile a represen-

tative SOC data set. Sampling locations were incorporated

only where soil profiles were sampled by horizon down to the

parent material or to least 1 m depth and then analyzed for

SOC, bulk density (BD) and stone content (SC). The minimum

requirement for SOC analysis was a determination by dry

combustion using a CN elemental analyzer. Generally, only

soil data collected after 1990 were included in order to reduce

temporal SOC changes.

The selected sampling locations that fulfilled all require-

ments comprised a data set of 1460 soil profiles (Fig. 1). The

main land uses were adequately represented, with 384 locations

(26% of the data) as cropland (34% of the total area), 333 loca-

tions (23%) as grassland (16%) and 596 locations (41%) as forest

(35%). The remaining locations were composed of 68 and 79

sites (10%) for bogs and other land uses (15%), respectively.

The main part of the data constituted a grid sampling with

1077 sites (Joneck et al., 2006). Between 2000 and 2004, soil

profiles were sampled using grids of 8 9 8 km within

Bavaria. For each soil profile a representative location was

selected within a radius of 1500 m around the grid node to

achieve a homogeneous sampling area in terms of vegetation,

relief, soil type and parent material and a central position in

the particular land use type. Anthropogenic disturbances in

the subsoil were excluded in a pre-exploratory survey using a

soil auger. For top- and subsoil horizons down to a depth of

about 35 cm, soil material was collected as a composite sam-

ple from eight sub-locations around the main soil profile in

order to cover the small-scale heterogeneity of the soils.

Samples below 35 cm were taken from the main soil profile.

The remaining 383 soil profiles were from permanent soil

monitoring sites (Schubert, 2002) and other regional soil

surveys. With regard to the total area of Bavaria, a sampling

density of one location per 48 km2 was obtained. The main

land uses cropland, grassland and forest had sampling densi-

ties of one location per 63, 34 and 41 km2, respectively.

Determination of SOC stocks and other soil properties

For the calculation of SOC stocks, BD of soil horizons was

quantified in all soil surveys from the mass of the oven-dry

soil (105° C) divided by the volume of the soil cores (Hartge &

Horn, 1989). For 1.8% of organic and subsoil horizons, BD

could not be analyzed directly due to high amounts of stones

in C horizons or low thickness of O horizons. In these cases,

BD was estimated based on BD values from adjacent sites with

similar land use, soil type and parent material. Stone contents

were estimated visually in the soil profiles according to

(AD-HOC AG Boden, 2005). All SOC concentrations were

determined by dry combustion with a CN elemental analyzer.

Samples that contained CaCO3 were heated to 500° C for 4 h

to remove organic carbon, and the concentration of inorganic

C of the residual material was determined by dry combustion.

The SOC content was calculated by subtracting inorganic

C from the total C concentration of the untreated material.

SOC stocks for each soil profile were calculated using the

following equation:

SOChz ¼ Rhzz SOCi � BDi � hi � 1� SCi

100

� �; ð1Þ

where SOC_{h_{z} is the total SOC stock (kg m�2) of all soil

horizons hz SOCi is the SOC concentration (mg g�1) of the fine

earth of horizon i,BDi is the BD (g cm�3) of the fine earth of

horizon i, hi is the thickness (cm) of horizon i and SCi is the volu-

metric fraction of rock fragments > 2 mm (%) of horizon i.

In order to evaluate the accuracy of SOC stock estimations

using pedotransfer functions (PTFs), SOC stocks were addi-

tionally calculated with estimated values for BD. Bulk density

was calculated as a function of organic matter (OM) content

and soil texture according to the PTF of Rawls (1983):

PTFBDRawls ¼ 100

OMBDOM

� �þ 100�OM

BDMM

� � ; ð2Þ

where OM is the content of organic matter (%; SOC con-

tent 9 1.72), BDOM is the BD of the OM (0.244 g cm�3) and

BDMM is the BD of the mineral matter (g cm�3). BDMM was

determined for different soil texture classes according to

Rawls (1983). Alternatively, BD was calculated with Eqn (2)

by estimating BDMM as a function of sand content and depth

as proposed by Tranter et al. (2007) (PTFBD Tranter):

BDMM ¼ 1:35þ 0:0045� Sþ 6� 10�5ð44:7� SÞ2 þ 0:06� logðDÞ; ð3Þ

where S is the sand content (%) and D is the depth (cm) of the

soil horizon. Furthermore, a PTF solely based on OM from

Adams (1973) that was modified by De Vos et al. (2005) was

used to calculate BD (PTFBD DE Vos). This PTF is identical to

the approach of Rawls (1983) (Eqn 2) but instead of a texture-

dependent BDMM a constant BD of 1.661 g cm�3 was used

and a different value for BDOM (0.312 g cm�3) was applied.

© 2012 Blackwell Publishing Ltd, Global Change Biology, doi: 10.1111/j.1365-2486.2012.02699.x

SOC STOCKS IN SOUTHEAST GERMANY 3

For all PTF-based SOC stock calculations, the BDPTF of organic

horizons corresponds to the respective BDOM (0.244 g cm�3 or

0.312 g cm�3). SOC stocks were then calculated according to

Eqn (1) using the different values for BDPTF and additionally

without consideration of SC. For an evaluation of the accuracy

of these PTF-based approaches, estimated SOC stocks were

compared to calculated stocks, and mean errors (MEs) as well

as root mean squared errors (RMSEs) were calculated.

Soil texture was determined by wet sieving and sedimenta-

tion according to the method of Kohn (Gee & Bauder, 1986).

Soil pH values were measured in 0.01 M CaCl2 at a soil/solu-

tion ratio of 1 : 2.5 at room temperature. In order to compare

soil horizon–specific properties and to gain insight the distribu-

tion of SOC stocks with depth, soil data from single soil hori-

zons were averaged (BD, SC, SOC concentration, soil texture,

pH) or added (SOC stock, thickness of soil horizon) to master

horizons (O, H = organic horizons; A = topsoil mineral

horizon; B = subsoil mineral horizon; C = parent material).

Statistical analysis

Descriptive statistics were applied to describe the SOC data

set and included mean, minimum/maximum values, median,

interquartile range, extremes and outliers, variance, skewness

and kurtosis. The main aim of the statistical analysis was to

investigate the influence of land use on soil properties. This

was done in the framework of a linear model, where the type

of land use is used as a categorical covariate affecting the

mean level of our data. Due to the spatial dependence in our

data, the model cannot be estimated by ordinary least squares,

as e.g. explained in Lark & Cullis (2004). Therefore we need to

explicitly model the continuous spatial variation by assuming

a basic structure for our variable of interest, denoted by

fSðxÞ : x 2 A � R2g in the following. It is assumed that

fSðxÞ : x 2 A � R2g is a random spatial process over the study

area A (Bavaria). Since it is unrealistic to obtain direct mea-

surements from the signal S, we assume that our measure-

ments, denoted by yi in the following, are realizations from

the model.

Yi ¼ SðxiÞ þ Zi; i ¼ 1; . . .; n; ð4ÞWith Zi being independent and identically distributed N(0, τ2)random variables. The measured response values yik are cate-

gorized into 5 different groups due to land use at location xik,

k = 1, ···, 5, i = 1, ···, nk. The locations are selected in a non ran-

dom way to build a lattice covering the observation region A.

The spatial process S is assumed to be a stationary Gaussian

process with

EðSðxÞÞ ¼ lðxÞ; ð5Þ

VarðSðxÞÞ ¼ r2; ð6Þ

CorrðSðxÞ; SðxÞÞ ¼ qðuÞ;wherejjx� xjj ¼ u; ð7ÞIn our application the Matern correlation function

qðu;/; kÞ ¼ ð2k�1CðkÞÞ�1ðu=/ÞkKkðu=/Þ; ð8Þwhere Г(·) is the usual Gamma function and Kk(·) is a modi-

fied Bessel function of order k = 1.5 showed the best fit to our

data. The spatially varying mean l(xik) of S(xik) is given by the

equation

lðxikÞ ¼ lþ ak0 ; ð9Þwhere l is an overall mean and ak is the effect due to land use

k. Further it is assumed that the conditional distribution of Y

(xik) given S is Gaussian with mean S(xik) and variance τ2

Hence the Yik, k = 1, ···, 5, i = 1, ···, nk are mutually indepen-

dent, conditionally on S. Unfortunately our data does not

fulfill the assumption of having a Gaussian distribution,

conditionally on S. Therefore the Box-Cox transformation Hk

YðkÞðxikÞ :¼ HkðYðxikÞÞ ¼ YkðxikÞ � 1Þk

; ð10Þ

With k = �0.083 is applied to the data to stabilize the variance.

The Gaussian model then is assumed to hold for the trans-

formed observations. For more details on the Gaussian model

for geostatistical data we refer to Diggle & Riberio (2007). All

calculations were performed using R, which is a free software

environment for statistical computing (R Development Core

Team, 2011). The Gaussian linear model was fitted to the data

by restricted maximum likelihood estimation as e.g. explained

in Diggle & Riberio (2007), Chapter 5.4. In R this was done by

using routines from the contributed R package geoR (Riberio

& Diggle, 2001). The residual analysis showed a satisfactory fit

to the data. In a second step all pairwise comparisons for the

different levels of land use were computed. The null hypothe-

ses of no difference between two levels of land use, were

tested using a Wald statistic as e.g. explained in Lark & Cullis

(2004). The p-values were adjusted for multiple comparisons

by a simple Bonferroni correction.

Results

Basic soil characteristic of land use types

In order to characterize the selected soil profiles of the

main land use types, basic soil properties (soil texture,

pH, soil thickness) were calculated (Table 1). No signif-

icant differences were found for the total thickness of

the soil profiles between the different land use types

with depths from 84 to 96 cm. For cropland, grassland

and other land uses soil texture was comparable along

all horizons with proportions of 31–42% for sand, 30–46% for silt and 23–29% for clay. Sand contents

increased slightly with soil depth, whereas silt contents

decreased accordingly. Forest sites show slightly higher

sand contents of 42–47% and lower proportions of clay

(15–21%). Soil pH values ranged from 6.0 to 6.8 for

cropland and other land uses, whereas considerably

lower values of 3.7–5.8 were detected in organic as well

as top- and subsoil mineral horizons for grassland and

forest soils.

All investigated soil profiles were classified into soil

classes according to the AD-HOC AG Boden (2005);

Table 2). In general, soils with well-developed B hori-

zons (Cambisols) were the most abundant soil class

© 2012 Blackwell Publishing Ltd, Global Change Biology, doi: 10.1111/j.1365-2486.2012.02699.x

4 M. WIESMEIER et al.

within cropland, grassland and forest sites with pro-

portions of 37–55%. At sites with other land uses,

soils with initial soil formation (Leptosols, Regosols)

and Cambisols dominated, and natural bogs and fens

(Histosols) and groundwater soils (Gleysols) were the

most abundant soil classes for bog areas. Despite the

dominance of Cambisols, Gleysols played an impor-

tant role at grassland sites, accounting for 19% of all

investigated soil profiles. For cropland and forest,

soils with initial soil formation (Leptosols, Regosols)

and soils with water stagnation (Stagnosols, Albeluvi-

sols, Planosols) had considerable proportions of 20%

and 17%, respectively.

Calculated and estimated SOC stocks for different landuses

Soil organic carbon stocks down to the parent material

for different land uses are shown in Fig. 2. Grassland

soils showed significantly (P < 0.01) higher SOC stocks

of 11.8 kg m�2 compared with forest and cropland

soils. For forest soils, slightly but significantly

(P < 0.05) higher SOC stocks of 9.8 kg m�2 were deter-

mined compared to cropland soils with 9.0 kg m�2.

The highest amounts of SOC were found for bogs with

51.1 kg m�2 and for other uses SOC stocks of

8.5 kg m�2 were determined.

Table 1 Basic properties (soil texture, pH, soil thickness) of soils under different land use (mean values with standard deviation)

Land use n Horizon Sand (%) Silt (%) Clay (%) pH (CaCl2) Depth (cm)

Cropland (C) 384 A 30.7 ± 23.0 46.0 ± 18.4 23.3 ± 11.4 6.3 ± 0.8 93 ± 22

B 31.4 ± 25.4 39.9 ± 17.2 28.7 ± 15.5 6.2 ± 0.9

C 32.3 ± 28.6 43.2 ± 24.4 24.6 ± 16.5 6.7 ± 1.1

Grassland (G) 333 A 34.6 ± 20.1 40.2 ± 13.3 25.1 ± 13.0 5.7 ± 0.9 94 ± 19

B 34.9 ± 22.2 37.5 ± 13.9 27.7 ± 16.2 5.8 ± 1.0

C 41.2 ± 28.5 30.2 ± 15.0 28.6 ± 18.8 6.6 ± 1.2

Forest (F) 596 O – – – 3.7 ± 0.9 93 ± 25

A 43.0 ± 22.7 41.8 ± 18.2 15.2 ± 8.5 3.8 ± 1.1

B 41.8 ± 22.2 36.9 ± 14.7 21.4 ± 12.5 4.4 ± 0.9

C 46.8 ± 26.1 32.0 ± 14.6 21.4 ± 17.7 5.6 ± 1.6

Other uses (O) 79 A 33.1 ± 25.9 41.0 ± 17.1 26.2 ± 16.5 6.2 ± 1.1 84 ± 27

B 37.9 ± 28.2 34.4 ± 15.6 27.9 ± 17.9 6.0 ± 1.1

C 41.8 ± 25.6 34.6 ± 16.2 23.9 ± 15.0 6.8 ± 1.1

Bog (B) 68 H – – – 5.5 ± 1.7 96 ± 20

A 14.3 ± 2.8 51.5 ± 9.3 34.1 ± 10.9 6.5 ± 1.2

B 39.9 ± 26.5 40.3 ± 19.6 19.7 ± 18.9 6.8 ± 1.3

C – – – 7.5 ± 0.1

Table 2 Number and relative percentage of soil classes under different land uses according to the German soil systematic

(AD-HOC AG Boden, 2005) and the equivalent Reference Soil Groups of the WRB (IUSS Working Group WRB, 2006): C = soils

from limestone weathering residues (Cambisols, Luvisols); S = soils with water stagnation (Stagnosols, Albeluvisols, Planosols);

L = soils with clay migration (Luvisols); B = soils with well-developed B horizons (Cambisols); D = clay-rich soils (Vertisols, Vertic

Cambisols); P = Podzols; R = soils with initial soil formation (Leptosols, Regosols); Y = anthropogenic soils (Anthrosols, Cambi-

sols); A = alluvial soils (Fluvisols, Gleysols and Reference Soil Groups in combination with Gleyic qualifier); G = groundwater soils

(Gleysols); H = natural bogs and fens (Histosols)

Land use C S L B D P R Y A G H Total

Cropland (n) 5 39 26 171 15 1 76 20 21 10 – 384

(%) 1.3 10.2 6.8 44.5 3.9 0.3 19.8 5.2 5.5 2.6 – 100

Grassland (n) 3 35 2 122 13 – 47 12 35 64 – 333

(%) 0.9 10.5 0.6 36.6 3.9 – 14.1 3.6 10.5 19.2 – 100

Forest (n) 9 103 30 329 9 38 51 – 13 14 – 596

(%) 1.5 17.3 5.0 55.2 1.5 6.4 8.6 – 2.2 2.3 – 100

Other uses (n) 1 5 – 29 2 – 30 3 2 7 – 79

(%) 1.3 6.3 – 36.7 2.5 – 38.0 3.8 2.5 8.9 – 100

Bog (n) – 1 – – – – 3 – 1 28 35 68

(%) – 1.5 – – – – 4.4 – 1.5 41.2 51.5 100

© 2012 Blackwell Publishing Ltd, Global Change Biology, doi: 10.1111/j.1365-2486.2012.02699.x

SOC STOCKS IN SOUTHEAST GERMANY 5

SOC stocks were also calculated using different

PTFs for BD and without consideration of SC

(Table 3). For cropland and grassland sites, SOC

stocks which were calculated with estimated values

for BD were slightly under- or overestimated (�10%

to +12%) compared with the calculation with the mea-

sured values for BD. A comparison between calcu-

lated and estimated SOC stocks for cropland and

grassland revealed MEs of �2.07 to 0.92 and RMSEs

of 1.28–3.58 (Fig. 3). In contrast, a higher deviation

was found for forest sites. Generally, the application

of PTFs for BD led to a considerable overestimation of

SOC stocks of 20–50%; accordingly, values for ME

and RMSE were higher with ranges of �5.29 to �2.00

and 3.63–6.47, respectively. SOC stocks that were cal-

culated without consideration of SC showed an

overestimation of 9%, 18% and 12% for cropland,

grassland and forest sites, respectively.

The storage of SOC in relation to soil depth and soil class

In order to gain insight into the vertical distribution of

SOC stocks within the soil profiles of the dominant land

use types, relevant soil data of single soil horizons were

aggregated to the master horizons O, A, B and C

(Fig. 4). In B and C horizons from all three land use

types, relatively low SOC concentrations < 10 mg g�1

were calculated without any significant differences. In

the topsoil (A horizon), grassland and forest soils

showed significantly (P < 0.01) higher concentrations

of 30 and 48 mg g�1, respectively, compared with crop-

land soils with 15 mg g�1. Bulk density generally

increased with depth in all soils, but considerable dif-

ferences were found between the land use types, partic-

ularly in the A horizon. The topsoils of cropland had

significantly (P < 0.05) higher BD of 1.39 g cm�3 com-

pared with grassland (1.20 g cm�3) and forest (0.91 g

cm�3). Also B horizons of cropland soils were denser

than grassland and forest subsoils. For C horizons, no

significant differences were found. Stone content was

relatively low (3.5%) in the A horizons of all land use

types. For B horizons, slightly higher contributions of

stones (7.5%) were calculated for grassland and forest,

whereas in C horizons, SC of 25–30% was determined.

In cropland soils, SCs did not increase with depth. The

thickness of the master horizons varied widely between

the land use types. A horizons of cropland soils (30 cm)

were much thicker compared with grassland (20 cm)

and forest soils (6 cm). In contrast, grasslands soils

exhibited bulky B horizons with thicknesses of 57 cm,

whereas forest and cropland soils had shallower B hori-

zons of 51 and 42 cm, respectively. The thickness of the

C horizons was comparable between the land use

types.

Based on the differences in SOC concentration, BD

and thickness of the horizons, the distribution of SOC

along the master horizons varied considerably between

cropland, grassland and forest soils. Similar SOC stocks

were found for the A horizons of cropland and grass-

land (5.8 and 6.3 kg m�2) but much lower amounts

were found in forest topsoils (2.4 kg m�2). Within B

horizons, grassland soils stored almost twice the

amount of SOC (4.3 kg m�2) compared with cropland

(2.2 kg m�2). For B horizons of forest soils, SOC stocks

of 3.6 kg m�2 were calculated. Relatively low amounts

(0.8–1.1 kg m�2) were found for C horizons. At forest

sites, additional SOC amounts of 2.5 kg m�2 were

stored in the O horizon.

In addition to the stratification by land use, soil class

–specific SOC stocks were calculated (Fig. 5). The

Fig. 2 Boxplots of soil organic carbon stocks from entire soil

profiles down to the C horizon for cropland (C), grassland (G),

forest (F), other land uses (O) and bogs (B). Note the separate

scale on the right (Lines within the boxes give the median, boxes

the 25th and 75th percentile, whiskers the lowest and highest

values excluding outliers, circles (1.5–3.0 interquartile range)

and triangles (more than 3.0 interquartile ranges) represent out-

liers and extremes).

Table 3 SOC stocks for cropland (C), grassland (G) and for-

est (F) calculated using all parameters (SOC content, BD, SC,

depth); by pedotransfer functions for BD (PTFBD) according to

Rawls (1983), De Vos et al. (2005) and Tranter et al. (2007); and

without consideration of SC

SOC stock calculation

SOC stocks (kg m�2)

C G F

All parameters 9.0 11.8 9.8

PTFBD (Rawls, 1983) 8.1 11.3 11.8

PTFBD (Tranter et al., 2007) 8.9 12.3 12.4

PTFBD (De Vos et al., 2005) 9.6 13.2 14.7

Without SC 9.8 13.9 11.0

© 2012 Blackwell Publishing Ltd, Global Change Biology, doi: 10.1111/j.1365-2486.2012.02699.x

6 M. WIESMEIER et al.

classes of soils from limestone weathering residues

(Cambisols, Luvisols), soils with water stagnation

(Stagnosols, Albeluvisols, Planosols), soils with clay

migration (Luvisols), Cambisols, clay-rich soils (Verti-

sols, Vertic Cambisols), Podzols and soils with initial

soil formation (Leptosols, Regosols) showed compara-

ble SOC stocks of 8.9–10.0 kg m�2. For anthropogenic

soils (Anthrosols, Cambisols), alluvial soils (Fluvisols,

Gleysols and Reference Soil Groups with Gleyic quali-

fier) and Gleysols, significantly (P < 0.05) higher stocks

of 13.5, 15.1 and 18.2 kg m�2 were calculated, respec-

tively. The highest amounts of SOC were stored in

natural bogs and fens (Histosols) with 69.1 kg m�2.

Discussion

SOC stocks as influenced by land use and sampling depth

The calculation of SOC stocks down to 1 m or to the

parent material for Bavarian soils seems to confirm our

hypothesis that the storage of SOC is strongly con-

trolled by land use. Except for bogs, where the accumu-

lation of SOC is caused by permanent water saturation,

land use clearly affects the storage of SOC within the

soil profile. Since the mean thickness of the soil profiles

was similar for all land use types (Table 1), an effect of

different total soil depths on SOC storage could be

Fig. 3 Comparison between calculated and estimated SOC stocks for cropland (C), grassland (G) and forest (F) using pedotransfer

functions (PTFBD) according to Rawls (1983), De Vos et al. (2005) and Tranter et al. (2007) with mean errors (ME) and root mean squared

errors (RMSE).

© 2012 Blackwell Publishing Ltd, Global Change Biology, doi: 10.1111/j.1365-2486.2012.02699.x

SOC STOCKS IN SOUTHEAST GERMANY 7

excluded. The storage of SOC increased in the order

cropland < forest < grassland (Fig. 2) as was found in

numerous other studies. However, the differences of

SOC stocks between the main land uses were consider-

ably smaller than in other soil inventories in compara-

ble climatic regions. Table 4 compiles land-use specific

SOC stocks for central European countries derived

from large-scale surveys for a soil depth of 1 m or

down to the parent material as well as for the first

30 cm of the soils. The sampling densities were highly

variable ranging between 10 and 311 km2 per sampling

location. On average, considerably higher stocks were

estimated in SOC inventories to 1 m depth or to the

parent material for central European grasslands

(13.4 kg m�2) and forests (12.5 kg m�2) compared with

our results of 11.8 and 9.8 kg m�2 for grassland and

forest, respectively. In contrast, a similar range of SOC

stocks was found for cropland soils (8.7 and

9.0 kg m�2). Based on SOC data from the cited refer-

ences (1 m depth or parent material), SOC stocks of

cropland soils were 35% and 30% lower compared with

stocks in grassland and forest soils. The loss of SOC in

cultivated soils has been investigated in numerous

studies and can be linked to the removal of crop

residues, reduced vegetation cover, enhanced minerali-

zation of OM, erosion and deteriorated soil aggregation

(e.g. Gregorich et al., 1998; Six et al., 1998; Balesdent

et al., 2000; Stoate et al., 2001; Hamza & Anderson,

2005). Besides the cited SOC inventories at the national

scale, meta analyses of land use changes at the regional

scale and modelling approaches revealed a cultivation-

induced SOC decline of 30–80% when forest and grass-

land were converted to cropland (Guo & Gifford, 2002;

Murty et al., 2002; Houghton & Goodale, 2004; Janssens

et al., 2005; Sleutel et al., 2007; Poeplau et al., 2011).

In contrast, our results showed much lower SOC

losses for cropland soils, with decreases of only 24%

and 8% compared with grassland and forest, respec-

tively. This discrepancy can primarily be related to the

consideration of different soil depths. Since only the

first 30 cm of the soil are directly affected by cultiva-

tion, most studies dealing with land use changes are

restricted to this depth or less (e.g. Murty et al., 2002;

Poeplau et al., 2011). Also most soil inventories are

restricted to a depth of 30 cm, resulting in distinct dif-

ferences of SOC stocks between cropland and grass-

land/forest. For example, soil inventories for central

Europe which are restricted to this depth (Table 4)

Fig. 4 Boxplots of SOC concentration, bulk density (BD), stone content (SC), thickness and SOC stocks of O, A, B and C horizons for

cropland (C), grassland (G) and forest (F) soils (Lines within the boxes give the median, boxes the 25th and 75th percentile, whiskers

the lowest and highest values excluding outliers, circles (1.5–3.0 interquartile range) and triangles (more than 3.0 interquartile ranges)

represent outliers and extremes).

© 2012 Blackwell Publishing Ltd, Global Change Biology, doi: 10.1111/j.1365-2486.2012.02699.x

8 M. WIESMEIER et al.

show on average 40% and 43% lower SOC stocks for

cropland compared with grassland and forest, respec-

tively. As these large differences cannot be assigned to

subsoils, the depletion of SOC in cropland soils is rela-

tively lower compared with grassland/forest when

considering total SOC stocks down to the parent mate-

rial as was indicated in this study. Harrison et al. (2011)

also pointed out that shallow soil sampling could lead

to a major underestimation of total SOC stocks and thus

to an incorrect assessment of the impact of controlling

parameters on SOC storage.

Moreover, SOC stock calculations are often based

on estimated rather than measured soil parameters

using PTFs and regression analyses. SOC studies

down to 1 m or to the parent material compiled in

Table 4 are either based on mean values for SOC

concentration and BD (Leifeld et al., 2005; Neufeldt,

2005; Bolliger et al., 2008; Mestdagh et al., 2009) or on

estimated stone contents and BD (Lettens et al.,

2005b) or even on typical carbon densities for differ-

ent land use types (Gingrich et al., 2007). In the stud-

ies of Baritz & Strich (2000) and Schubert (2010), soil

samples were taken within constant soil depths

instead of sampling soil horizons, which resulted in a

more accurate estimation of SOC stocks (Gruneberg

et al., 2010). All these shortcomings probably lead to

biased SOC estimations.

In order to evaluate the accuracy of SOC stock calcu-

lations that are partly based on estimated parameters,

SOC stocks were additionally calculated using common

PTFs for BD and without consideration of SCs

(Table 3). The well-established PTF of Rawls (1983) was

used to estimate BD as a function of OM content and

texture-dependent mineral BD that was used in several

SOC inventories (Lettens et al., 2005a; Goidts & van

Wesemael, 2007; Stevens & van Wesemael, 2008). Tran-

ter et al. (2007) modified this method by replacing min-

eral BD as a function of sand content and depth that is

also frequently used in SOC studies (Anderson-Teixeira

et al., 2009). Finally, the well-known PTF of Adams

(1973) that was modified by De Vos et al. (2005) was

used to calculate BD. This approach is solely based on

OM and was applied in several SOC studies (Post &

Kwon, 2000). Since SOC stocks calculated with estimated

values for BD were only slightly under- or overesti-

mated for cropland and grassland sites (approximately

10%), the applied PTFs seem to be only applicable for a

rough estimation of SOC stocks under these land uses.

However, for forest soils PTFs led to considerable over-

estimation of SOC stocks of up to 50%, pointing toward

a systematic bias. A comparison between measured

and estimated SOC stocks demonstrated the bias

because ME values were generally negative for all

applied PTFs at forest sites (Fig. 3). This PTF-induced

Fig. 5 Boxplots of SOC stocks from entire soil profiles down to the C horizon for different soil classes of the German soil systematic

(AD-HOC AG Boden, 2005) and according Reference Soil Groups of the WRB (IUSS Working Group WRB, 2006): C = soils from lime-

stone weathering residues (Cambisols, Luvisols); S = soils with water stagnation (Stagnosols, Albeluvisols, Planosols); L = soils with

clay migration (Luvisols); B = soils with well-developed B horizons (Cambisols); D = clay-rich soils (Vertisols, Vertic Cambisols);

P = Podzols; R = soils with initial soil formation (Leptosols, Regosols); Y = anthropogenic soils (Anthrosols, Cambisols); A = alluvial

soils (Fluvisols, Gleysols and Reference Soil Groups in combination with Gleyic qualifier); G = groundwater soils (Gleysols); H = natu-

ral bogs and fens (Histosols). Note the separate scale on the right (Lines within the boxes give the median, boxes the 25th and 75th per-

centile, whiskers the lowest and highest values excluding outliers, circles (1.5–3.0 interquartile range) and triangles (more than 3.0

interquartile ranges) represent outliers and extremes).

© 2012 Blackwell Publishing Ltd, Global Change Biology, doi: 10.1111/j.1365-2486.2012.02699.x

SOC STOCKS IN SOUTHEAST GERMANY 9

bias is probably responsible for higher forest SOC

stocks reported in several SOC inventories (e.g. studies

in Table 4). Generally, the applicability of PTFs could

probably be improved, if distinct PTFs would be used

for different land uses, soil types and soil depths (top-

soil/subsoil).

To summarize, the long-term effect of different land

uses on SOC storage can only be detected satisfacto-

rily when SOC stocks are determined down to the

parent material and all soil parameters are measured

as was done in this study. Our results indicated that

cultivation was not followed by a strong decline of

SOC in southeast Germany, because only slightly

smaller SOC stocks were detected in cropland soils

compared with forest soils. Moreover, grassland soils

showed considerably higher SOC stocks compared

with forest soils, which indicates in turn an additional

accumulation of SOC.

Distribution of SOC stocks with depth

The distribution of total SOC stocks to master soil hori-

zons provided a more detailed view of the controlling

parameters of SOC storage with depth (Fig. 4). A hori-

zon-based approach was used instead of a designation

of SOC stocks to depth increments since differences of

SOC storage between the land use types are more pro-

nounced and pedogenetic processes that control SOC

storage can be detected (Gruneberg et al., 2010). Gener-

ally, much lower SOC concentrations in the A horizon

Table 4 Compilation of studies that estimated land use–specific SOC stocks for central European countries to a depth of 100 cm/

parent material (A) as well as to a depth of 30 cm (B)

Study area Depth (cm)

SOC stocks (kg m�2)

n SD* (km2) ReferenceCropland Grassland Forest

(A) SOC inventories to a depth of 100 cm/parent material (pm)

Switzerland 100/pm 9.0 9.3–11.7 11.9 808 35 Leifeld et al., 2005;

Bolliger et al., 2008;

Austria pm 6.7 13.5–16.2 11.4 n.a. n.a. Gingrich et al., 2007;

Bavaria, Germany 100 n.a. n.a. 11.7 372 66 Schubert, 2010;

Baden-Wurttemberg,

Germany

100 10.6 15.0 n.a. 395 47 Neufeldt, 2005;

Germany 90(100)† n.a. n.a. 10.9(12.1)† 1800 56 Baritz & Strich, 2000;

Flanders, Belgium 100 n.a. 11.1–16.0 n.a. 1530 1 Mestdagh et al., 2009;

Belgium 100 8.4 13.0 14.8–15.5 n.a. n.a. Lettens et al., 2005b;

Mean 100 8.7 ± 1.6 13.4 ± 1.8 12.5 ± 1.6

(B) SOC inventories to a depth of 30 cm

Wallonia, Belgium 30 4.1–4.3 8.3–8.9 n.a. 427 22 Goidts & van Wesemael,

2007; Goidts et al., 2009

Belgium 30 5.0 7.9 8.8–9.2 n.a. n.a. Lettens et al., 2005a;

Belgium 30 5.0 8.0 n.a. 629 29 Meersmans et al., 2011

Wallonia, Belgium 30 n.a. n.a. 9.6 127 10 Stevens & van Wesemael,

2008;

France 30 4.9 7.6 7.0 1748 311 Martin et al., 2011

Baden-Wurttemberg,

Germany

30 6.0 9.5 n.a. 395 47 Neufeldt, 2005;

Denmark 20(30)‡ n.a. n.a. 6.2(9.3)‡ n.a. n.a. Liski et al., 2002

Belgium 20(30)‡ n.a. n.a. 4.2(6.3)‡ n.a. n.a.

the Netherlands 20(30)‡ n.a. n.a. 3.8(5.7)‡ n.a. n.a.

Austria 20(30)‡ n.a. n.a. 7.1(10.7)‡ n.a. n.a.

France 20(30)‡ n.a. n.a. 3.5(5.3)‡ n.a. n.a.

Germany 20(30)‡ n.a. n.a. 6.8(10.2)‡ n.a. n.a.

Switzerland 20(30)‡ n.a. n.a. 10.0(15.0)‡ n.a. n.a.

Mean 30 5.0 ± 0.6 8.3 ± 0.8 8.8 ± 2.9

*SD, sampling density (study area in km2/sampling locations).

†extrapolation to 100 cm in parentheses.

‡extrapolation to 30 cm in parentheses.

n.a., not available.

© 2012 Blackwell Publishing Ltd, Global Change Biology, doi: 10.1111/j.1365-2486.2012.02699.x

10 M. WIESMEIER et al.

of cropland soils compared with grassland and forest

soils points at first sight toward a strong depletion of

SOC due to cultivation. However, SOC stocks deter-

mined for cropland A horizons are only slightly lower

(�8%) than grassland stocks and are even more than

twice as high as for forest A horizon. This is due to BD

being considerably higher in cropland A horizons

because cultivation leads to soil compaction, particu-

larly in the plow pan (Hamza & Anderson, 2005).

Moreover, the thickness of cropland A horizons is con-

siderably higher, indicating a vertical deepening of the

A horizon of 10 cm due to tillage. For cropland soils of

northwestern Germany, Nieder & Richter (1986) calcu-

lated that topsoil deepened 10 cm on average, which

coincided with our results, and estimated that a poten-

tial SOC enrichment of 1.4–1.7 kg m�2 could thereby be

achieved. Obviously, soil cultivation is leading not only

to SOC depletion but also to relocation of OM from the

topsoil into deeper soil regions and correspondingly,

the amount of mineralized OM is probably often over-

estimated. In studies with constant depth increments

this ‘dilution effect’ of SOC cannot be detected.

Hence, the differences of SOC stocks between crop-

land and grassland/forest can only partly be related

to cultivation-induced depletion of SOC in the topsoil.

Instead, large differences of SOC stocks were found

for B horizons. Grassland B horizons stored almost

twice the amount of SOC as cropland subsoils,

although BD is even lower in grassland subsoils. A

distinctly higher thickness of grassland B horizons as

well as higher SOC concentrations is responsible for

this strong accumulation of SOC. The total stock dis-

crepancy between grassland and cropland is largely

related to this difference in the subsoil. In forest soils,

small A horizon stocks are compensated for by higher

subsoil SOC stocks and additionally by considerable

amounts in the O horizons.

The importance of soil type–specific SOC storage

The high amount of SOC stored in B horizons of

grassland soils could be related to different processes.

On one hand, grassland species with high root-to-

shoot ratios have a high belowground net primary

productivity associated with a deep OM input in

grassland soils (Jobbagy & Jackson, 2000; Rumpel &

Kogel-Knabner, 2011). On the other hand, Jackson

et al. (1996) showed that in temperate grasslands, 83%

of the root biomass is found in the first 30 cm of the

soils. Pedogenetic processes probably provide further

explanation for the accumulation of SOC in the sub-

soil of grasslands. Despite Cambisols being the most

abundant soil class, Gleysols also played an important

role, with a proportion of almost 20% within

grassland soils (Table 2). In cropland and forest soils,

Gleysols had negligible proportions of 2.6% and 2.3%

due to this groundwater dominated soil class being

only partially usable for cultivation. The permanent

water saturation in the subsoil of Gleysols is associ-

ated with a distinct accumulation of OM. The relation

of SOC stocks to soil classes clearly showed this, since

Gleysols were the most SOC-rich soil class besides

Histosols with stocks of 18.2 kg m�2 (Fig. 5). High

SOC stocks in grassland B horizons can thus be

related to the preferential use of Gleysols as grass-

land, where high groundwater levels lead to an accu-

mulation of OM in the subsoil.

These results cause the importance of land use to be

questioned as the most important factor controlling the

storage of SOC in the soil region studied and highlight

the relevance of pedogenetic properties in terms of soil

classes. However, it should be noted that parameters,

which influence SOC storage, are generally not indepen-

dent from each other and often correlated and thus, their

explanatory power is partly related to other variables.

Since certain soil classes seem to be correlated to specific

land use types, we calculated land use–specific SOC

stocks separately for the most abundant soil class Cam-

bisols (Fig. 6). Within the Cambisol data set, consider-

ably lower grassland SOC stocks (10.3 kg m�2) in

the range of forest stocks (10.0 kg m�2) showed that

Fig. 6 Boxplots of SOC stocks for cropland (C), grassland (G)

and forest (F) within the soil class Cambisols (Lines within the

boxes give the median, boxes the 25th and 75th percentile, whis-

kers the lowest and highest values excluding outliers, circles

(1.5–3.0 interquartile range) and triangles (more than 3.0

interquartile ranges) represent outliers and extremes).

© 2012 Blackwell Publishing Ltd, Global Change Biology, doi: 10.1111/j.1365-2486.2012.02699.x

SOC STOCKS IN SOUTHEAST GERMANY 11

SOC-rich Gleysols are responsible for high grassland

stocks determined for the entire data set. The detrimen-

tal effect of cultivation on SOC storage in cropland soils

was still visible, but with only 20% and 18% lower stocks

compared with grassland and forest explicitly lower

than was reported in studies for other central European

countries (Table 4).

Recommendations for future SOC inventories

The reported land use–specific and soil class–specificSOC stocks are regarded as representative for central

Europe, due to the high sampling density and the fact

that all relevant soil parameters (SOC concentration,

BD, SC) were determined by horizon down to the

parent material. The incorporation of subsoil SOC

stocks revealed that land use may not be the main

controlling factor for SOC storage and highlighted the

importance of pedogenetic properties, particularly in

grassland soils. We recommend that pedogenetic soil

information (soil classes, for example) should be

included in SOC stock estimations as well as in C

sequestration studies. Our results further indicate that

SOC depletion in cropland soils due to cultivation is

probably often overestimated because tillage-induced

deepening of the topsoil was ignored by studies with

fixed depths. The application of modelled parameters

in SOC inventories is generally questioned, because

SOC stocks, calculated with pedotransfer functions,

were systematically biased, particularly for forest

soils. Therefore, we propose that in future SOC inven-

tories, soils should be sampled down to the parent

material and completely analyzed by horizon instead

of depth increments in order to increase the accuracy

of SOC stock estimations and to elucidate pedogenetic

effects on SOC storage. Possible overestimations of a

cultivation-induced SOC depletion prompt further

questioning of the accuracy of C emission scenarios

from soils caused by land use changes and probably

also impact the decision-making in policy related to

land use. A quantification of functional SOC pools

with different turnover times according to different

land uses, soil types and climatic conditions would

make it possible to estimate the future development

of SOC stocks under a changing climate. In this

respect, the data set presented here provides a unique

basis for studies on future SOC changes.

Acknowledgements

We thank Alfred Schubert from the Bavarian State Institute forForestry for providing forest soil data. Peter Schad is acknowl-edged for the allocation of soil classes to the WRB system. Weare grateful to the Bavarian State Ministry of the Environment

and Public Health for funding the project ‘Der Humuskorperbayerischer Boden im Klimawandel – Auswirkungen und Po-tentiale’.

References

Adams WA (1973) The effect of organic matter on the bulk and true densities of some

uncultivated podzolic soils. Journal of Soil Science, 24, 10–17.

AD-HOC AG Boden (2005) Bodenkundliche Kartieranleitung. Bundesanstalt fur Geowis-

senschaften und Rohstoffe (Ed.). E. Schweizerbart′sche Verlagsbuchhandlung,

Stuttgart.

Anderson-Teixeira KJ, Davis SC, Masters MD, Delucia EH (2009) Changes in soil

organic carbon under biofuel crops. Global Change Biology Bioenergy, 1, 75–96.

Balesdent J, Chenu C, Balabane M (2000) Relationship of soil organic matter dynamics

to physical protection and tillage. Soil & Tillage Research, 53, 215–230.

Baritz R, Strich S (2000) Forests and the national greenhouse gas inventory of Ger-

many. Biotechnology, Agronomy, Society and Environment, 4, 267–271.

Batjes NH (1996) Total carbon and nitrogen in the soils of the world. European Journal

of Soil Science, 47, 151–163.

Bolliger J, Hagedorn F, Leifeld J, Bohl J, Zimmermann S, Soliva R, Kienast F (2008)

Effects of land-use change on carbon stocks in Switzerland. Ecosystems, 11, 895–907.

De Vos B, Van Meirvenne M, Quataert P, Deckers J, Muys B (2005) Predictive quality

of pedotransfer functions for estimating bulk density of forest soils. Soil Science

Society of America Journal, 69, 500–510.

Diggle PJ, Riberio PJ Jr (2007) Model-based Geostatistics. Springer, New York.

Eglin T, Ciais P, Piao SL et al. (2010) Historical and future perspectives of global soil

carbon response to climate and land-use changes. Tellus Series B-Chemical and Phys-

ical Meteorology, 62, 700–718.

Gee GW, Bauder JW(1986) Particle-size analysis. In: Methods of Soil Analysis, Part 1,

Physical and Mineralogical Methods (ed. Page AL), pp. 383–411. Soil Science Society

of America, Wisconsin.

Gingrich S, Erb KH, Krausmann F, Gaube V, Haberl H (2007) Long-term dynamics of

terrestrial carbon stocks in Austria: a comprehensive assessment of the time period

from 1830 to 2000. Regional Environmental Change, 7, 37–47.

Goidts E, van Wesemael B (2007) Regional assessment of soil organic carbon changes

under agriculture in Southern Belgium (1955–2005). Geoderma, 141, 341–354.

Goidts E, van Wesemael B, Van Oost K (2009) Driving forces of soil organic carbon

evolution at the landscape and the regional scale using data from a stratified soil

monitoring. Global Change Biology, 15, 2981–3000.

Gregorich EG, Greer KJ, Anderson DW, Liang BC (1998) Carbon distribution and

losses: erosion and deposition effects. Soil & Tillage Research, 47, 291–302.

Gruneberg E, Schoning I, Kalko EKV, Weisser WW (2010) Regional organic carbon

stock variability: a comparison between depth increments and soil horizons. Geo-

derma, 155, 426–433.

Guo LB, Gifford RM (2002) Soil carbon stocks and land use change: a meta analysis.

Global Change Biology, 8, 345–360.

Hamza MA, Anderson WK (2005) Soil compaction in cropping systems – a review of

the nature, causes and possible solutions. Soil & Tillage Research, 82, 121–145.

Harrison RB, Footen PW, Strahm BD (2011) Deep soil horizons: contribution and

Importance to soil carbon pools and in assessing whole-ecosystem response to

management and global change. Forest Science, 57, 67–76.

Hartge KH, Horn R (1989) Die physikalische Untersuchung von Boden. Enke Verlag,

Stuttgart.

Houghton RA, Goodale C (2004) Effects of land-use change on the carbon balance of

terrestrial ecosystems. In: Ecosystems and land use change (ed. DeFries R, Asner G,

Houghton RA), pp. 85–98. American Geophysical Union, Washington, DC.

IPCC (2000) Land Use, Land-Use Change, and Forestry. Intergovernmental Panel on Cli-

mate Change (IPCC), Geneva.

IPCC (2003) Good Practice Guidance for Land Use, Land-Use Change and Forestry. Inter-

governmental Panel on Climate Change (IPCC), Geneva.

IUSS Working Group WRB 2006. World reference base for soil resources 2006. World

Soil Resources Reports No. 103, FAO, Rome.

Jackson RB, Canadell J, Ehleringer JR, Mooney HA, Sala OE, Schulze ED (1996)

A global analysis of root distributions for terrestrial biomes. Oecologia, 108,

389–411.

Janssens IA, Freibauer A, Schlamadinger B et al. (2005) The carbon budget of ter-

restrial ecosystems at country-scale – a European case study. Biogeosciences, 2,

15–26.

Jobbagy EG, Jackson RB (2000) The vertical distribution of soil organic carbon and its

relation to climate and vegetation. Ecological Applications, 10, 423–436.

© 2012 Blackwell Publishing Ltd, Global Change Biology, doi: 10.1111/j.1365-2486.2012.02699.x

12 M. WIESMEIER et al.

Joneck M, Hangen E, Martin W et al. (2006) Wissenschaftliche Grundlagen fur den

Vollzug der Bodenschutzgesetze in Bayern (GRABEN) – ein Projekt stellt sich vor.

Bodenschutz, 2, 32–38.

Jones RJA, Hiederer R, Rusco E, Montanarella L (2005) Estimating organic carbon in

the soils of Europe for policy support. European Journal of Soil Science, 56, 655–671.

Lark RM, Cullis BR (2004) Model-based analysis using REML for inference from sys-

tematically sampled data on soil. European Journal of Soil Science, 55, 799–813.

Leifeld J, Bassin S, Fuhrer J (2005) Carbon stocks in Swiss agricultural soils predicted

by land-use, soil characteristics, and altitude. Agriculture Ecosystems & Environ-

ment, 105, 255–266.

Lettens S, van Orshoven J, van Wesemael B, Muys B, Perrin D (2005a) Soil organic

carbon changes in landscape units of Belgium between 1960 and 2000 with refer-

ence to 1990. Global Change Biology, 11, 2128–2140.

Lettens S, Van Orshovena J, van Wesemael B, De Vos B, Muys B (2005b) Stocks and

fluxes of soil organic carbon for landscape units in Belgium derived from hetero-

geneous data sets for 1990 and 2000. Geoderma, 127, 11–23.

Liski J, Perruchoud D, Karjalainen T (2002) Increasing carbon stocks in the forest soils

of Western Europe. Forest Ecology and Management, 169, 159–175.

von Lutzow M, Kogel-Knabner I (2009) Temperature sensitivity of soil organic

matter decomposition-what do we know? Biology and Fertility of Soils, 46,

1–15.

von Lutzow M, Kogel-Knabner I, Ludwig B et al. (2008) Stabilization mechanisms of

organic matter in four temperate soils: development and application of a concep-

tual model. Journal of Plant Nutrition and Soil Science, 171, 111–124.

Martin MP, Wattenbach M, Smith P, Meersmans J, Jolivet C, Boulonne L, Arrouays D

(2011) Spatial distribution of soil organic carbon stocks in France. Biogeosciences, 8,

1053–1065.

Meersmans J, van Wesemael B, Goidts E, van Molle M, De Baets S, De Ridder F

(2011) Spatial analysis of soil organic carbon evolution in Belgian croplands and

grasslands, 1960–2006. Global Change Biology, 17, 466–479.

Mestdagh I, Sleutel S, Lootens P et al. (2009) Soil organic carbon-stock changes in

Flemish grassland soils from 1990 to 2000. Journal of Plant Nutrition and Soil Science,

172, 24–31.

Murty D, Kirschbaum MUF, McMurtrie RE, McGilvray A (2002) Does conversion of

forest to agricultural land change soil carbon and nitrogen? a review of the litera-

ture. Global Change Biology, 8, 105–123.

Neufeldt H (2005) Carbon stocks and sequestration potentials of agricultural soils in

the federal state of Baden-Wurttemberg, SW Germany. Journal of Plant Nutrition

and Soil Science, 168, 202–211.

Nieder R, Richter J (1986) C- und N-Festlegung in Boden Sudostniedersachsens nach

Krumenvertiefung. Journal of Plant Nutrition and Soil Science, 149, 189–201.

Poeplau C, Don A, Vesterdal L, Leifeld J, Van Wesemael B, Schumacher J, Gensior A

(2011) Temporal dynamics of soil organic carbon after land-use change in the tem-

perate zone – carbon response functions as a model approach. Global Change Biol-

ogy, 17, 2415–2427.

Post WM, Kwon KC (2000) Soil carbon sequestration and land-use change: processes

and potential. Global Change Biology, 6, 317–327.

Prechtel A, von Lutzow M, Schneider BU, Bens O, Bannick CG, Kogel-Knabner I,

Huttl RF (2009) Organic carbon in soils of Germany: status quo and the need for

new data to evaluate potentials and trends of soil carbon sequestration. Journal of

Plant Nutrition and Soil Science, 172, 601–614.

R Development Core Team 2011. R: A Language and Environment for Statistical Comput-

ing. R Foundation for Statistical Computing, Vienna. Available at: http://www.

r-project.org/ (accessed 15 February 2012).

Rawls WJ (1983) Estimating soil bulk-density from particle-size analysis and organic-

matter content. Soil Science, 135, 123–125.

Riberio . PJ Jr, Diggle PJ (2001) geoR: a package for geostatistical analysis. R-News, 1,

15–18.

Rumpel C, Kogel-Knabner I (2011) Deep soil organic matter-a key but poorly under-

stood component of terrestrial C cycle. Plant and Soil, 338, 143–158.

Schrumpf M, Schulze ED, Kaiser K, Schumacher J (2011) How accurately can soil

organic carbon stocks and stock changes be quantified by soil inventories? Biogeo-

sciences, 8, 1193–1212.

Schubert A (2002) Bayerische Waldboden-Dauerbeobachtungsflachen – Bodenuntersuchun-

gen. WissenschaftszentrumWeihenstephan fur Ernahrung, Landnutzung und Um-

welt der Technischen Universitat Munchen and Bayerische Landesanstalt fur

Wald und Forstwirtschaft, Freising.

Schubert A (2010) Organisch gebundener Kohlenstoff im Waldboden. LWF Waldfors-

chung aktuell, 78, 11–14.

Six J, Elliott ET, Paustian K, Doran JW (1998) Aggregation and soil organic matter

accumulation in cultivated and native grassland soils. Soil Science Society of America

Journal, 62, 1367–1377.

Sleutel S, De Neve S, Hofman G (2007) Assessing causes of recent organic carbon

losses from cropland soils by means of regional-scaled input balances for the case

of Flanders (Belgium). Nutrient Cycling in Agroecosystems, 78, 265–278.

Smith J, Smith P, Wattenbach M et al. (2005) Projected changes in mineral soil carbon of

European croplands and grasslands, 1990–2080. Global Change Biology, 11, 2141–2152.

Stevens A, van Wesemael B (2008) Soil organic carbon dynamics at the regional scale

as influenced by land use history: a case study in forest soils from southern

Belgium. Soil Use and Management, 24, 69–79.

Stoate C, Boatman ND, Borralho RJ, Carvalho CR, de Snoo GR, Eden P (2001) Ecologi-

cal impacts of arable intensification in Europe. Journal of Environmental Manage-

ment, 63, 337–365.

Stolbovoy V, Montanarella L, Filippi N, Jones A, Gallego J, Grassi G (2007) Soil Sam-

pling Protocol to Certify the Changes of Organic Carbon Stock in Mineral Soil of the

European Union. Office for Official Publications of the European Communities,

Luxembourg.

Tranter G, Minasny B, McBratney AB, Murphy B, McKenzie NJ, Grundy M, Brough

D (2007) Building and testing conceptual and empirical models for predicting soil

bulk density. Soil Use and Management, 23, 437–443.

© 2012 Blackwell Publishing Ltd, Global Change Biology, doi: 10.1111/j.1365-2486.2012.02699.x

SOC STOCKS IN SOUTHEAST GERMANY 13