PhD Thesis
Carbon fluxes of an extensive meadow and attempts for flux partitioning
Michael Riederer (06/2009-03/2014)
Support: Thomas Foken
In times of climate change and increasing carbon dioxide concentrations, three questions arise for ecosystem sciences: At first, which ecosystems can contribute to mitigate those processes? Secondly, how will ecosystems react on the changing conditions? And finally, is the performance of our commonly applied research methods adequate under those complex and continuously changing environmental conditions? This thesis is integrated in the joint research project FORKAST which investigates those questions. The role of grassland ecosystems’ source or sink, related to the carbon cycle is currently not welldefined. At least, extensively managed grassland in mid European low mountain ranges may be able to contribute to climate change mitigation by carbon sequestration.
In ecosystem sciences, two dominant approaches are used to gain access to the carbon cycle. On the one hand these are the micrometeorological methods as the eddy-covariance technique which provides a top view from the atmosphere and, on the other hand, leading isotopic methods used in agricultural and soil science which allow a more interior view on the ecosystem. In this thesis, the advantages of both are turned to account.
In a first step, the investigated area, an extensively managed grassland in a mid European low mountain range, was defined as a net carbon sink. The carbon uptake accounted for – 91 g C m–2 a–1 in 2010. It has to be mentioned, too, that the long term climate measurements on the site revealed an upward trend of spring droughts. In a forty year time series a decrease of precipitation of 21 mm in April and May had been detected. Hence, the reaction of the carbon cycle was investigated by inducing a 1000-year spring drought event (i.e. 38 days without any precipitation) and comparing the carbon allocation into shoots, roots, soil and respiration fluxes to those detected on plots with normal precipitation. Therefore, a stable isotope pulse labeling experiment had been conducted. This fact indicated an increase of carbon allocation by 6.2% to below ground pools as soil and roots and a reduction of shoot respiration by 8.5% due to spring drought.
Gaining absolute values of carbon allocation, the relative portion, provided by pulse labeling and tracing, was set off the absolute carbon input into the ecosystem, obtained by eddy-covariance measurements of the net ecosystem carbon exchange in combination with partitioning of that into underlying assimilation and respiration flux. With the absolute XIII carbon input of –7.1 g C m-2 d–1 and the relative allocation of the labeling, into fluxes of 2.5, 0.8, 0.5, 2.3 and 1.0 g C m–2 d–1 into shoots, roots, soil, shoot respiration and CO2 efflux could be determined and validated.
Flux partitioning is an important tool in ecosystem sciences. It can be accomplished in different ways. The commonly applied flux partitioning model based on Lloyd–Taylor and Michaelis–Menten functions had been compared to dark and transparent chamber measurements and to a partitioning by an isotopic approach, based on isoflux measurements with the relaxed eddy accumulation technique. The latter comparison revealed a lack of sensitivity of the common flux partitioning model for ecosystem reactions on short term changes in the weather conditions. The isotopic model based on detecting the isotope discrimination worked well on grassland compared to former experiments over a forest. Furthermore, relaxed eddy accumulation based 13CO2 isoflux measurements confirmed only minor influences of atmospheric isofluxes on isotopic labeling experiments by detecting only a negligible portion of 13CO2 of the entire CO2 flux. However, there are certain restrictions for applying relaxed eddy accumulation on managed grassland, found in this study. Scalar similarity, a precondition for proper relaxed eddy accumulation fluxes, cannot be guaranteed directly after the management. It is suggested to wait at least 22 days in summer and 12 days in autumn after the management. The ecosystem needed this span of time to recover the regular source/sink distribution of water vapor, CO2 and temperature.
The chamber method was applied to validate the assimilation flux, provided by the common flux partitioning model. This was done during the day at time of turbulent atmospheric conditions. In a comparison experiment between the chamber and eddycovariance a good agreement was found at that time. In the late afternoon and during night, the chamber could not reproduce present atmospheric conditions, as, for example, increasing stable stratification due to the oasis effect or coherent structures. This resulted in smaller chamber CO2 source fluxes of 26% during night and larger chamber CO2 sink fluxes of 14% during day. The chamber technique is important for small scale measurements (especially in treatment experiments). Thus, it is important to know the reasons for those differences to eddy-covariance.