|其他摘要||As the core of the material circulation and the energy transmission of the terrestrial system, the carbon cycle and the water cycle play critical roles in the research of the global climate change and the response of the terrestrial system. Now, due to the improvement of eddy covariance technique, we can directly obtain net ecosystem carbon flux or total water flux, but it cannot distinguish their components. Using chamber-based method and eddy covariance method, we simultaneously measured the carbon flux and total water flux in a Haloxylon ammodendron community, located in the southern margin of Gurbantunggut Desert in western China, to partition the contributions of the different kinds of functional plants and soil to the carbon flux and water flux of the community.
In order to investigate different functional plant's contribution to the carbon and water flux of community level, the synchronous observation of carbon and water exchange in shrub layer, herb layer and community levels were carried out during the growing season of 2013 and 2014. Considering of seasonal variation of environmental factors, special attention was paid on the seasonal variation of the ratio of Rsoil to Reco by the continuous observation of soil CO2 flux. The main results include the following aspects:
The contribution of herbaceous layer to carbon-water flux mainly focused in April to June, which was closely related to water conditions in research area.. The soil moisture was sufficient due to snow melting water at the beginning of April, which was beneficial for the fast growth of ephemeral plants. Whereas the carbon and water flux of the shrub layer were relatively stable in the whole growing season, which further illustrated the main contribution of the shrub layer to community carbon fixed.
The values of ecosystem respiration deriving from chamber-based method were higher than values of eddy covariance method, which was consistent with the former consequence of ecosystem respiration deriving from the two contrasting methods, the higher proportion ranging between 20%−40%. The phenomenon of flux underestimation during night has attracted more and more attentions, in spite of flux data were optimized previously by eliminating CO2 advection, removing linear trend, WPL correction and getting rid of friction wind speed threshold, the fact that flux underestimation during night was still not changed.
The Reco was unimodal distribution during the whole growing season, the maximum Reco was 0.75 gC m-2d-1 and the minimum was 0.09 gC m-2d-1. The change of Rsoil was not obvious during May to July (0.34-0.51 gC m-2d-1), and Rsoil began decreasing since the end of July.
The ratio of Rsoil and Reco ranged from 0.38 to 0.98, with an average of 0.67 across the whole growing season. On a seasonal basis, Rsoil/Reco presented an obvious “U” pattern: the value of Rsoil/Reco was high at the beginning and ending of growing season, but was low in the middle of growing season. There was a significant linear correlation between the Rsoil/Reco ratios and biomass of H.ammodendron assimilating twigs (R2 = 0.458, P < 0.05), indicating that as the H.ammodendron growth, the contribution of H.ammodendron aboveground respiration (especially with leaf respiration) to Reco was also increased.|