Evaluating the carbon balance estimate from an automated ground-level flux chamber system in artificial grass mesocosms

Measuring and modeling carbon (C) stock changes in terrestrial ecosystems are pivotal in addressing global C-cycling model uncertainties. Difficulties in detecting small short-term changes in relatively large C stocks require the development of robust sensitive flux measurement techniques. Net ecosystem exchange (NEE) ground-level chambers are increasingly used to assess C dynamics in low vegetation ecosystems but, to date, have lacked formal rigorous field validation against measured C stock changes. We developed and deployed an automated and multiplexed C-flux chamber system in grassland mesocosms in order rigorously to compare ecosystem total C budget obtained using hourly C-flux measurements versus destructive net C balance. The system combines transparent NEE and opaque respiration chambers enabling partitioning of photosynthetic and respiratory fluxes. The C-balance comparison showed good agreement between the two methods, but only after NEE fluxes were corrected for light reductions due to chamber presence. The dark chamber fluxes allowed assessing temperature sensitivity of ecosystem respiration (R(eco)) components (i.e., heterotrophic vs. autotrophic) at different growth stages. We propose that such automated flux chamber systems can provide an accurate C balance, also enabling pivotal partitioning of the different C-flux components (e.g., photosynthesis and respiration) suitable for model evaluation and developments.