 |
| Titel |
Dynamics of carbon in deep soils inferred from carbon stable isotopes signatures : a worldwide meta-analysis |
| VerfasserIn |
Jerôme Balesdent, Isabelle Basile-Doelsch, Joël Chadoeuf, Sophie Cornu, Delphine Derrien, Zuzana Fekiacova, Christine Hatté |
| Konferenz |
EGU General Assembly 2014
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 16 (2014) |
| Datensatznummer |
250093888
|
| Publikation (Nr.) |
 EGU/EGU2014-9052.pdf |
|
|
|
|
|
| Zusammenfassung |
| The contribution of soil carbon deeper than 30 cm to the atmospheric carbon balance is still
poorly understood. A very straightforward quantification of the gross exchange of carbon
between the atmosphere and soil organic matter can be obtained at places where the
13C/12C signature of vegetation has been changed for known durations, due to switch
of the photosynthetic metabolism (C3 or C4) or to Free Air Carbon Enrichment
experiments.
We compiled C and 13C profile data of 113 sites of this type, either gahered from the
literature or from our own measurements. Each site comprised two profiles : one where the
13C/12C of the vegetation had been changed, and a reference profile with unchanged
vegetation 13C/12C. An isotope mixing equation was used, which takes into account the
natural isotope enrichments with depth and decay. Three main variables were calculated at
any depth from 0 to 100 cm and in a few sites down to 200 cm : the carbon content, the
proportion of new carbon (aged less than the duration of change t) and the amount of new
carbon. The database concerned 23 countries, various climates (58% intertropical and 42%
between 23° to 56° latitude) and various soil types and textures. Landuses and
vegetation consisted in 26% of forests and woodlands, 35% of grasslands and 38% of
cropped systems. The duration of the natural labelling t ranged from 2 years to
ca. 4000 years. Peatlands, boreal, and desert environments were absent from the
database.
Non-linear regressions with time across the dataset yielded kinetic parameters of the age
distribution on one hand and of the flux of new carbon incorporation (kg C m-2 yr-1) on the
other, each calculated by 10 cm depth increments. On the average, the median ages of carbon
increase from ca. 15 years at 0 cm to more than 1000 years at 100 cm. Turnover is on the
average 2 to 3 times slower for the subsoil (30-100 cm) than for the topsoil (0-30 cm).
Based on the incorporation of new C in the first decades, the carbon input to the
30-100 cm layer was estimated to 0.3 times (with 1 S.D. = 0.2 times) that to the
topsoil 0-30 cm, whereas the corresponding ratio for total carbon stocks is close to
1.
A multivariate analysis confirmed that the turnover rate in the topsoil is dependent on land
use and mean annual temperature, and related to a lesser extent to aridity index and clay
content. The relative proportion of carbon input to the subsoil is higher in croplands than in
forests or grasslands, in probable accordance with the exportation of plant aerial parts as
crops.
We derive from this study quantitative constraints on depth-dependent mechanisms that
drive carbon dynamics, such as decreasing decay rates down the depth and the magnitude of
priming effects, the rate and intensity of carbon transport downwards, or the occurrence of
stable C throughout the profiles. We also propose simplified expressions for the
parameterization of models of carbon exchanges between deep soil organic carbon and the
atmosphere. |
|
|
|
|
|
|