 |
| Titel |
Modelling soil CO2 diffusion to investigate the effect of non-steady-state on isotopic composition of respired CO2 |
| VerfasserIn |
U. Gamnitzer, A. B. Moyes, D. R. Bowling, H. Schnyder |
| Konferenz |
EGU General Assembly 2009
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 11 (2009) |
| Datensatznummer |
250023927
|
|
|
|
|
|
| Zusammenfassung |
| In the steady-state, the carbon isotopic composition of the soil CO2 efflux is determined by
the respiratory CO2 production in the soil. But in reality steady-state conditions are rare.
Non-steady-state can be introduced either naturally, for example by changes in
photosynthetic discrimination due to changes in environmental parameters, or artificially, for
example by changing δ13C of CO2 in air during a labelling experiment. This complicates the
experimental determination of δ13C of respired CO2. To quantitatively assess the impact of
such changes, we used a one-dimensional iterative diffusion model to calculate the depth
profile of CO2 concentration in soil air as well as its isotopic composition by considering
12CO2 and 13CO2 as separate diffusing gases. Model predictions were confirmed by
observations in two different soil depths.
The model was adapted to allow simulation of chamber-based measurements
of non-steady-state effects on the apparent isotopic composition of respiration.
These are compared with a dataset from a 13CO2 labelling experiment in a grassland
ecosystem, where the tracer was observed in total ecosystem respired CO2 during
nighttime. δ13C of ecosystem respiration, measured with closed static chambers,
differed significantly from steady-state measurements using open dynamic chambers.
Translated to the fraction of labelled carbon in respired CO2, the closed static chamber
measurements suggested a much larger fraction of tracer in respired CO2 (70-80% after two
weeks of labelling) than the open dynamic chamber measurements (40-50%). The
simulations demonstrated that the closed chamber measurements were biased due
to non-steady-state effects. Consideration of storage of labelling CO2 in the soil
gas pores during the preceeding labelling period caused a bias of the respiratory
signal in the observed direction, but could not explain the magnitude of the bias.
When CO2 dissolved in soil water was also included in the simulation, the order of
magnitude of bias of the closed static chamber measurements was also explained. |
|
|
|
|
|
|