On the Use of Multi-Scale Long-Term Data for Understanding Global Ecosystem Functioning

Friday, 19 December 2014: 11:50 AM
Markus Reichstein, Nuno Carvalhais, Martin Jung and Miguel D Mahecha, Max Planck Institute for Biogeochemistry, Jena, Germany
Recent continuous observations of this exchange of CO2, H2O and sensible heat within the global observation network FLUXNET have enabled us to quantify ecosystem function in response to large eco-climatological spatial gradients and temporal variability of climate. We see that ecosystem function co-varies strongly with climate, but that climate alone does not suffice to explain the variation in total. Instead vegetation biophysical and structural parameters co-determine the exchange of carbon, water and energy between the ecosystem and the atmosphere. By combining ecosystem level observation and information of spatial meteorological and vegetation remote sensing covariates we can infer global patterns of ecosystem atmosphere fluxes and derive key ecosystem functional properties globally. While this approach is powerful and meteorological and vegetation structural predictors explain more than 70% of the spatial variation of monthly fluxes at FLUXNET sites, it ignores the effect of ecophysiological vegetation properties, which is expected from plant physiological leaf or whole-plant studies. Hence, future research has to more strongly link the organismic trait information with ecosystem functional properties. For this we propose a framework that involves 1) the correlation of community aggregated traits with flux-derived ecosystem properties across a range of ecosystems, 2) the up-scaling of vegetation traits using spatially distributed geo-ecological co-variates and the comparison with global ecosystem functional properties and their co-variation with climate, 3) the use of vegetation traits instead of vegetation classes for empirical up-scaling of ecosystem-atmosphere fluxes from ecosystem to globe. We hypothesize that these studies will emphasize the superiority of a trait based approach over classical structural PFT concepts for modeling global ecosystem functioning, but will also highlight the scale-emergent properties at ecosystem level, which cannot be explained by plant traits alone.