Global terrestrial carbon and nitrogen cycling insensitive to estimates of biological N fixation

Monday, 14 December 2015
Poster Hall (Moscone South)
Joerg Steinkamp1, Bettina Weber2, Christian Werner1 and Thomas Hickler1,3, (1)Senckenberg Biodiversity and Climate Research Centre (BiK-F), Frankfurt/Main, Germany, (2)Max Planck Institute for Chemistry, Mainz, Germany, (3)Institute of Physical Geography, Goethe-University Frankfurt am Main, Frankfurt/Main, Germany
Dinitrogen (N2) is the most abundant molecule in the atmosphere and incorporated in other molecules an essential nutrient for life on earth. However, only few natural processes can initiate a reaction of N2. These natural processes are fire, lightning and biological nitrogen fixation (BNF) with BNF being the largest source. In the course of the last century humans have outperformed the natural processes of nitrogen fixation by the production of fertilizer. Industrial and other human emission of reactive nitrogen, as well as fire and lightning lead to a deposition of 63 Tg (N) per year. This is twice the amount of BNF estimated by the default setup of the dynamic global vegetation model LPJ-GUESS (30 Tg), which is a conservative approach. We use different methods and parameterizations for BNF in LPJ-GUESS: 1.) varying total annual amount; 2.) annual evenly distributed and daily calculated fixation rates; 3.) an improved dataset of BNF by cryptogamic covers (free-living N-fixers). With this setup BNF is ranging from 30 Tg to 60 Tg. We assess the impact of BNF on carbon storage and grand primary production (GPP) of the natural vegetation. These results are compared to and evaluated against available independent datasets. We do not see major differences in the productivity and carbon stocks with these BNF estimates, suggesting that natural vegetation is insensitive to BNF on a global scale and the vegetation can compensate for the different nitrogen availabilities. Current deposition of nitrogen compounds and internal cycling through mineralization and uptake is sufficient for natural vegetation productivity. However, due to the coarse model grid and spatial heterogeneity in the real world this conclusion does not exclude the existence of habitats constrained by BNF.