Carbon sequestration potential of soils in southeast Germany derived from stable soil organic carbon saturation

Download Carbon sequestration potential of soils in southeast Germany derived from stable soil organic carbon saturation

Post on 27-Mar-2017




0 download

Embed Size (px)


  • Carbon sequestration potential of soils in southeastGermany derived from stable soil organic carbonsaturationMART IN WIESME IER * , R ICO HUBNER , P ETER SP ORLE IN , UWE GEU ,



    *Lehrstuhl fur Bodenkunde, Department fur Okologie und Okosystemmanagement, Wissenschaftszentrum Weihenstephan fur

    Ernahrung, Landnutzung und Umwelt, Technische Universitat Munchen, Freising-Weihenstephan 85350, Germany, Lehrstuhl

    fur Wirtschaftslehre des Landbaues, Wissenschaftszentrum Weihenstephan fur Ernahrung, Landnutzung und Umwelt, Technische

    Universitat Munchen, Freising-Weihenstephan 85350, Germany, Bavarian Environment Agency, Hof 95030, Germany,

    Institute for Advanced Study, Technische Universitat Munchen, Garching 85748, Germany


    Sequestration of atmospheric carbon (C) in soils through improved management of forest and agricultural land is

    considered to have high potential for global CO2 mitigation. However, the potential of soils to sequester soil organic

    carbon (SOC) in a stable form, which is limited by the stabilization of SOC against microbial mineralization, is largely

    unknown. In this study, we estimated the C sequestration potential of soils in southeast Germany by calculating the

    potential SOC saturation of silt and clay particles according to Hassink [Plant and Soil 191 (1997) 77] on the basis of

    516 soil profiles. The determination of the current SOC content of silt and clay fractions for major soil units and land

    uses allowed an estimation of the C saturation deficit corresponding to the long-term C sequestration potential. The

    results showed that cropland soils have a low level of C saturation of around 50% and could store considerable

    amounts of additional SOC. A relatively high C sequestration potential was also determined for grassland soils. In

    contrast, forest soils had a low C sequestration potential as they were almost C saturated. A high proportion of sites

    with a high degree of apparent oversaturation revealed that in acidic, coarse-textured soils the relation to silt and clay

    is not suitable to estimate the stable C saturation. A strong correlation of the C saturation deficit with temperature

    and precipitation allowed a spatial estimation of the C sequestration potential for Bavaria. In total, about 395 Mt

    CO2-equivalents could theoretically be stored in A horizons of cultivated soils four times the annual emission ofgreenhouse gases in Bavaria. Although achieving the entire estimated C storage capacity is unrealistic, improved

    management of cultivated land could contribute significantly to CO2 mitigation. Moreover, increasing SOC stocks

    have additional benefits with respect to enhanced soil fertility and agricultural productivity.

    Keywords: agricultural management, climate change, CO2 mitigation, soil organic carbon stocks, soil fractionation, stabilization

    of soil organic matter

    Received 18 April 2013 and accepted 30 August 2013


    Sequestration of atmospheric carbon (C) in soils is

    considered to contribute significantly to CO2 mitiga-

    tion, and several management options for increasing

    SOC stocks have been discussed. For forest ecosystems,

    practices such as a change in tree species composition,

    afforestation, thinning, drainage, fertilization, liming,

    site preparation and harvest management are associ-

    ated with an increase in SOC stocks and are conse-

    quently viewed as having a high potential for soil C

    sequestration (Goodale et al., 2002; Liski et al., 2002;

    Karjalainen et al., 2003; Lal, 2005; Jandl et al., 2007; Ciais

    et al., 2008; Lorenz & Lal, 2010; Luyssaert et al., 2010;

    Carroll et al., 2012; Vesterdal et al., 2012; Wiesmeier

    et al., 2013b). An even higher C sequestration potential

    is assumed for agricultural soils because a distinct

    depletion of SOC stocks has been observed in most

    cultivated soils (Paustian et al., 1997; Lal, 2004; Smith,

    2004). Among several agricultural practices that may

    increase C sequestration in cultivated soils, promising

    management options are promotion of organic inputs,

    conservation/zero tillage, converting cropland to grass-

    land, introduction of perennials, improved manage-

    ment of farmed peatland and organic farming (Cole

    et al., 1997; Paustian et al., 2000; Sauerbeck, 2001; Vlees-

    houwers & Verhagen, 2002; Freibauer et al., 2004; Hol-

    land, 2004; Lal, 2004; Johnson et al., 2007; Smith, 2012).Correspondence: Martin Wiesmeier, tel. +49 (0)8161 71 3679,

    fax +49 (0)8161 71 4466, e-mail:

    2013 John Wiley & Sons Ltd 653

    Global Change Biology (2014) 20, 653665, doi: 10.1111/gcb.12384

  • However, C sequestration by improved management

    of forest and agricultural soils reaches a new equilib-

    rium at a higher SOC level after a certain period of

    time. Several studies have shown that there is an upper

    limit of SOC storage, confirming the hypothesis of soil

    C saturation (Six et al., 2002; Goh, 2004; Stewart et al.,

    2007, 2008; Chung et al., 2008). This is related to the lim-

    ited potential of soils to stabilize soil organic matter

    (SOM) against microbial mineralization (Baldock &

    Skjemstad, 2000). There are three major SOM stabiliza-

    tion mechanisms: selective preservation due to recalci-

    trance of SOM, spatial inaccessibility of SOM due to

    hydrophobicity or occlusion in soil aggregates, and

    interaction with mineral surfaces (Sollins et al., 1996;

    von Lutzow et al., 2006). The last is regarded as quanti-

    tatively the most important in a wide range of soils, as

    indicated by a strong correlation of SOC stocks with

    clay contents (e.g. Oades, 1988; Arrouays et al., 2006).

    Hassink (1997) assumed that the capacity of soils to

    preserve SOC is limited by the proportion of silt and

    clay particles (fine fraction

  • main land uses were adequately represented, with 115 loca-

    tions (22% of the data) as cropland (34% of the total area), 110

    locations (21%) as grassland (16%), 249 locations (48%) as

    forest (35%) and 42 locations (8%) under other land uses

    (15%). The main part of the data constituted a grid sampling

    within Bavaria (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 500 m around the grid node to

    achieve a homogeneous sampling area in terms of vegetation,

    relief, soil type and parent material as well as a central posi-

    tion in the particular land use type. Anthropogenic distur-

    bances in the subsoil were excluded in a pre-exploratory

    survey using a soil auger. Topsoil material was collected as a

    composite sample from eight sub-locations around one main

    soil profile to cover the small-scale heterogeneity of the soils.

    At the main soil profile, steel core samples with a diameter of

    10 cm were extracted for topsoil horizons. A small number of

    soil profiles originated from permanent soil monitoring sites

    (Schubert, 2002) and other regional soil surveys.

    Determination of soil properties

    The proportion of SOC stored in the fraction

  • and curvature were determined. As secondary parameters,

    the contributing area (CA) and the topographic wetness index

    (TWI) were calculated using the following equation:

    TWI ln CAtan a


    where CA is the specific upslope contributing area derived by

    a geographical information system and a is the slope. The TWIis a topographical variable that indicates soil moisture condi-

    tions (Beven & Kirkby, 1978; Sorensen et al., 2006). To include

    geology as a potential parameter influencing the C saturation,

    parent material data were assigned from a map with 35 parent

    material classes (BAG500) with a resolution of 2 km from the

    Bavarian Environment Agency. Information about the soil type

    was included using a generalized soil map (BUK1000N) with

    28 superior soil classes (Leitbodenassoziationen) with a resolu-

    tion of 2 km from the Federal Institute for Geosciences and

    Natural Resources. The factor land use was incorporated by

    using 2006 satellite data from the CORINE Land Cover project

    (CLC2006) from the German Remote Sensing Data Center. For

    climatic variables, annual precipitation and mean annual tem-

    perature determined between 1981 and 2010 by the German

    Weather Service with a resolution of 1 km were allocated. All

    environmental parameters were assigned to 25 9 25 m cells.

    Statistical analysis

    Descriptive statistics were applied to describe the soil data sets

    including mean, minimum and maximum values, median,

    interquartile range, extremes and outliers, skewness and kur-

    tosis. In order to scale current C concentrations of the fine frac-

    tion 100 cm; 2 = 95100 cm;3 = 9095 cm; 4 =

  • fraction, which was calculated according to the equa-

    tion of Hassink (1997), was also similar between land

    uses and ranged between 19.7 and 20.8 mg g1.The current C concentration of the fine fraction was

    measured for soils under major land uses in Bavaria

    (Fig. 1). For cropland soils, the proportion of OC

    stored in the fraction

  • with the proportion of the fine fraction, particularly

    in cropland soils (Fig. 4). In contrast, forest soils and

    soils under other land uses showed no significant

    relationship with the fine fraction content. To gain

    insight into the factors that control the C sequestration

    potential, correlations between several environmental

    Fig. 2 Correlation between the proportion of particles

  • parameters and the C saturation deficit were examined

    (Table 3). Strong positive correlations (P < 0.01) werefound with mean annual temperature and pH and

    strong negative correlations (P < 0.01) with annual pre-cipitation, elevation, slope and soil class. However, a

    multiple linear regression analysis that included two

    factors derived from a PCA (Table 4) revealed that the

    C saturation deficit was strongly controlled by one fac-

    tor that showed high loadings of temperature, precipi-

    tation and elevation.

    Fig. 4 Correlation between proportion of particles

  • Discussion

    The C saturation deficit of agricultural soils

    The estimation of the C saturation deficit in soils of

    Bavaria revealed that it was dependent on the current

    C content of the fine fraction. The potential C saturation

    as a function of the proportion of particles

  • compared with cropland soils. This is in the range

    reported for pastures in the south-eastern United

    States, where a silt- and clay-associated C saturation of

    60% was determined (Conant et al., 2003). A higher C

    saturation in pastures compared with cropland was

    also detected in Australia, though on a lower C satura-

    tion level (Chan, 2001).

    In summary, both cropland and grassland soils in

    Bavaria have a substantial potential to sequester addi-

    tional amounts of C in a stable form. A positive rela-

    tionship of the C saturation deficit with the silt and clay

    content (Fig. 4) revealed that soils that are particularly

    fine textured have a large potential for C sequestration.

    C saturation in forest soils and limitations of Hassinksequation

    In forest soils, a distinctly different level of C saturation

    in the fine fraction was observed as compared with

    agricultural soils. The current C concentration

    increased slightly with the proportion of particles

  • degraded grasslands worldwide (Conant & Paustian,

    2002). For each land use, thresholds of temperature and

    precipitation were derived (point of intersection of the

    regression line at a C saturation deficit of 0) that divide

    the areas of the respective land uses into regions with

    saturated and unsaturated conditions (Table 5). The C

    sequestration potential for each land use was estimated

    by multiplying the area of unsaturated conditions by

    the median value of the C saturation deficit of this area.

    A regionalization of the C sequestration potential using

    geostatistical methods was not conducted as the C satu-

    ration deficit of the investigated locations was calcu-

    lated using only an estimation of the current C

    saturation of the fine fraction, which was based on a

    smaller data set.

    The results revealed that cropland and grassland

    soils of Bavaria could potentially sequester 32 and 6 Mt

    C in the uppermost 10 cm respectively. The high poten-

    tial of cropland soils is related to the high C saturation

    deficit of intensively cultivated soils and a large area

    with unsaturated conditions. Less than 1% of the total

    cropland area of Bavaria was assigned to saturated

    conditions as cultivation is not feasible in cool, humid

    areas. For forest soils, a C sequestration potential of

    only 4 Mt was estimated with a high uncertainty as

    previously explained. Other land uses have a potential

    to sequester 9 Mt C. The low potential of forest soils to

    sequester C can be ascribed to almost saturated condi-

    tions in forest soils and the fact that only half of the

    total forest area was associated with unsaturated condi-

    tions. About 50% of Bavarian forests are located in

    regions where cool, humid conditions result in a

    complete C saturation of silt and clay particles.

    The C sequestration potential estimated for the first

    10 cm of the soil was extrapolated to the median depth

    of A horizons of each land use, assuming that soil

    texture and SOC contents are comparable within the A

    horizon. The C sequestration potential of A horizons

    under cropland, grassland, forest and other uses was

    estimated to be 96, 12, 4 and 22 Mt respectively. In total,

    soils of Bavaria could additionally sequester 134 Mt C,

    18% of total SOC stocks of 764 Mt. This amount corre-

    sponds to 490 Mt CO2-equivalents (CO2-eq.), which is

    more than five times higher than the annual greenhouse

    gas emission (in 2009) in Bavaria of 94 Mt CO2-eq.

    (UGRdL, 2012). The majority, 395 Mt CO2-eq.

    Fig. 5 Correlation between mean annual temperature and the C saturation deficit (Csat-def) for cropland (C), grassland (G), forest (F)

    and other land uses (O). Values above the dashed line refer to a deficit of C saturation, values below the dashed line refer to an oversat-

    uration of C.

    2013 John Wiley & Sons Ltd, Global Change Biology, 20, 653665

    662 M. WIESMEIER et al.

  • (approximately 80%), could be sequestered in agricul-

    tural soils. An increase in C saturation in forest soils

    and soils under other land uses is associated with high

    uncertainty. Assuming that due to an improved man-

    agement of cultivated soils the theoretical stable C stor-

    age potential is reached after a mean period of 30 years

    (West & Six, 2007), a mean annual amount of 13 Mt

    CO2-eq. could be sequestered in Bavarian soils over this

    period of time, which is 14% of the annual emission of

    greenhouse gases in Bavaria (in 2009). On an area basis,

    4.1 t CO2-eq. ha1 yr1 could potentially be seques-

    tered in agricultural soils, which is considerably higher

    than observed and modelled C accumulation rates

    through various management options aimed at increas-

    ing SOC stocks in cultivated soils (Vleeshouwers &

    Verhagen, 2002; West & Post, 2002; Freibauer et al.,

    2004; Smith et al., 2008).


    A comparison of the current C amount with the poten-

    tial C saturation of...


View more >