Soil and Litter Carbonyl Sulfide Fluxes in an Oak Woodland in California.

Carbonyl sulfide (COS) is considered a promising tracer for partitioning gross photosynthesis and respiration from net carbon fluxes. This requires knowledge of soil COS fluxes to obtain the net COS mass balance. At the global scale, uptake by soil is considered the second largest COS sink after vegetation. The soil COS uptake is due to the enzyme carbonic anhydrase that is active in both leaves and soil micro-organisms. However, soil COS flux estimates are severely limited by the lack of data, particularly direct field observations. At an oak woodland in southern California, we measured soil COS and CO₂ fluxes for 40 days in spring 2013 using automated chambers and quantum cascade laser (QCL) spectroscopy. As the soil at the site was covered with a thick litter layer, we characterized litter COS fluxes separately from surface (soil+litter) fluxes. Soil and litter were primarily COS sinks, with a strong dependence on moisture content. We observed an optimum temperature of 14°C for surface COS uptake, whereas litter-only COS uptake was not sensitive to temperature. The relative uptake ratio of COS to CO₂ (that characterizes the relationship between COS fluxes and respiration) increased with soil temperature, i.e. was more negative at lower temperature. Concurrent bursts of respiration and COS uptake after rain events indicated that the soil microbial/enzyme activity was strongly water limited. We constructed a soil diffusion-reaction model, and used data-driven optimization of COS uptake and production activity parameters to simulate surface COS fluxes that agreed well with field observations. We suggest that the new process-based model framework can be used for modeling surface COS fluxes in semi-arid and other ecosystems.