Methane (CH₄) and nitrous oxide (N₂O) are two major anthropogenic greenhouse gases. Soil-atmosphere fluxes of these gases are driven by a complex network of soil microbial processes; however, the ecological mechanisms controlling their fluxes are not well understood to date. Starting in 2007, we studied trace gas fluxes in three pastures in Switzerland (ranging from montane/moderately managed to alpine/extensively managed ecosystems). Rain exclusion shelters simulated severe summer drought; this treatment was combined with a fertilization treatment consisting of NH₄NO₃, cattle urine, cattle faeces and no fertilizer (control treatment). 

Results/Conclusions

The three sites strongly differed in CH₄ oxidation rates, ranging from 50 µmol m^-2 d^-1 at the most intensively managed pasture to 150 µmol m^-2 d^-1 at the most extensively managed pasture. The drought treatments as well as natural patterns in precipitation showed that water content is the key controller of CH₄ oxidation, with CH₄ oxidation regularly varying by a factor of up to five in the course of drying cycles. This effect was mainly determined by soil moisture, which affected the diffusion of atmospheric methane into the soil and limiting CH₄ oxidation. The reduction of CH₄ uptake due to fertilizer application was significant but rather small compared to many other studies. An additional field experiment testing dose-effect relationships showed progressive limitation of CH₄ uptake with increasing amounts of NH₄NO₃ applied, an effect not observed for urine or feces containing equal total amounts of N, for unknown reasons. Laboratory experiments investigating the effect of individual fertilizer components (anions and cations) indicated that CH₄ oxidation was controlled by different processes at the ecosystem and at the individual soil aggregate level; furthermore, the inhibitory effects of added fertilizer components were strongly site-dependent; while CH₄ oxidation at some sites was inhibited by purely osmotic effects, other sites showed inhibition by ammonium but not other ions.

N₂O emissions strongly depended on the combination of the drought treatment and the fertilizer application. The addition of NH₄NO₃ and cattle urine caused much higher emissions than the application of cattle feces. Simulated summer drought significantly reduced N₂O emissions, with no differences between drought treatments remaining once soil moisture had equalized after removal of rain shelters.

Our results demonstrate that severe drought not only increases CH₄ oxidation but also reduces N₂O emissions, resulting in a net reduction of greenhouse gas emissions when expressed as CO₂ equivalents.