Soil Greenhouse Gas Flux Measurements with Automated and Manual Static Chambers, Forced Diffusion Chamber, and Concentration Profiles

Abstract:

Accurate measurements of soil greenhouse gas fluxes are critical for determining the role of ecosystem dynamics, both natural and managed, in climate change. We compared concentration profile methods with static, forced diffusion, and automated flux chambers using a combination of infrared gas analyzers (IRGA), gas chromatography, and cavity ring-down laser absorption spectroscopy (Picarro G2508) during field campaigns in managed ecosystems in California. At a drained peatland pasture site, we observed large differences between methods (fluxes ranged between 2-15 µmol CO₂ m^-2 s^-1). However, low temporal/high spatial replication measurements (manual LI6400 chamber measurements; n=6 collars) encompassed the full range of CO₂ fluxes observed across all other methods. This suggests that the majority of variability in CO₂ emissions was due to high spatial variation in soil respiration and not due to methodological differences across measurement systems. At a dry upland pasture site, water, nitrate solution and manure were applied during the experimental period to expand the range of greenhouse fluxes. Preliminary results showed good agreement of gas fluxes between static and automatic chamber sampling. We observed large CO₂ and N₂O fluxes after manure application with both methods. The two chamber types were highly significantly correlated for N₂O (slope=0.74, r²=0.94). Mean CH₄ fluxes measured by static chambers was -0.36 µg cm^-2 h^-1, similar to the -0.57 µg cm^-2 h^-1 measured by the automatic chamber and Picarro analyzer during the study period. Overall, our results suggest that both automated and static chamber methods are in good agreement, but automated chambers are advantageous for capturing diel dynamics and pulse responses to experimental treatments. Our results also highlight the importance of spatial replication, which can be difficult to achieve using expensive automated chambers. We suggest future research efforts to seek a combination of high spatially replicated manual static chambers and continuous sampling methods.