B41C-0075:
Soil-atmosphere carbonyl sulfide (COS) exchange in a tropical rainforest at La Selva, Costa Rica

Thursday, 18 December 2014
Wu Sun1, Kadmiel S Maseyk2, Sabrina Juarez3, Celine Lett4 and Ulrike H Seibt1, (1)University of California Los Angeles, Los Angeles, CA, United States, (2)Open University, Milton Keynes, MK7, United Kingdom, (3)BIOEMCO Biogéochimie et Ecologie des Milieux Continentaux, Thiverval-Grignon, France, (4)NERC British Antarctic Survey, Cambridge, United Kingdom
Abstract:
Carbonyl sulfide (COS) has recently been proposed as a promising tracer for partitioning ecosystem carbon assimilation due to the close analogy between leaf uptake processes of COS and CO2. This emerging framework requires accurate characterization of the source and sink components of COS, including soil fluxes. Here we present the first direct, continuous observations of soil COS fluxes for 4 months at a tropical rainforest, La Selva Biological Station, Costa Rica. Three soil plots with contrasting water content were selected for chamber measurements. Our observations confirmed that soils are principally COS sinks, with daily mean COS fluxes averaged across all chambers ranging from -3 to 0 pmol m-2 s-1. When compared with net ecosystem COS uptake which peaks around -30 pmol m-2 s-1, their contributions should be considered in ecosystem COS balance. We did not find a temperature optimum, but soil COS uptake slightly increased with soil temperature, indicating biotic control on soil COS fluxes. Diurnal cycles of COS fluxes were observed during drying out periods after rain. The diel periodicity of COS fluxes was probably obscured by frequent raining at the site. Diffusional control of soil COS fluxes is shown from increasing soil COS uptake at lower soil water-filled pore space. These confirm that soil COS fluxes are mediated both by soil physical and biological factors. Using a depth-resolved diffusion-reaction model with data-driven enzyme activity parameterization, we simulated the COS fluxes from measured soil environmental variables, consistent with observations. This modeling scheme is useful for separating soil COS fluxes from net ecosystem COS fluxes, which lends support to the emergent COS-based approach of carbon flux partitioning.