Carbon Isotope Discrimination of Terrestrial Ecosystems — How Well Do Observed and Modeled Results Match?
Nina Buchmann and 1. Introduction 253
fed O. Kaplan 2. Experimental and Analytical Methods 255
Max Planck Institut for 3_ Description of the Model 255
jena Germany ' • Signature or hcosystem Respiration 256
5. Modeled Ecosystem Carbon Discrimination 258
6. Comparison of Observed and Modeled Ae Estimates 259
7. Ecophysiological Information from ae 261
8. Conclusions 262
Terrestrial ecosystems play an important role in the global carbon cycle. Recently, S'-'C ratios of CO, in the atmosphere have been used in general circulation models to constrain the global carbon budget and imply location and magnitude of carbon sources and sinks. These models rely on scaling modeled 5' 'C ratios of soil and plant components to the ecosystem level, but no validation with measured ecosystem level estimates has been accomplished. However, the isotopic signature of the biosphere is highly variable through space and time. We have compiled a global dataset of measurements on ecosystem carbon isotope discrimination (ac) and used this dataset to validate a global terrestrial biosphere model that simulates ac (BIOME3.5). Measured ac values (based on ecosystem measurements) averaged 18%o globally, while the global modeled estimate (with BIOME3.5) averaged 15.6%o. These differences between the measurements and the model may be due to site selection and lack of representative coverage of certain ecosystem types as well as to model parameterization. Field measurements in deserts, C3 and C4 grasslands, and savanna systems are very limited or do not exist yet. The latitudinal bands between 40° and 20° S, 20° and 30° N or >70°N are not covered. The model, which does not incorporate information about land use, simulates a mean ae intermediate between those used in other modeling studies. The effects of land use may confound the global ac signal. The model also shows that the ratio of ecosystem assimilation to canopy conductance is closely related to the ecosystem's ac except in tropical savannas where roughly equal amounts of C3 and C4 vegetation coexist. Thus, ae is a useful tool for investigating the global carbon cycle as it provides information not only on isotopic fractionation during terrestrial C02 exchange with the atmosphere but also ecophysiological information on the water status of the vegetation. Future analyses of the global carbon budget need to account for the magnitude and the heterogeneity of the terrestrial isotopic signature as a 3%o underestimate of modeled ae can cause up to a 20% reduction in the estimated strength of the terrestrial carbon sink.
Was this article helpful?