Comparison of Soil CO2 Fluxes by Eddy-Covariance and Chamber Methods in Fallow Periods of a Corn-Soybean Rotation

Soil carbon dioxide (CO₂) fluxes are typically measured by eddy-covariance (EC) or chamber (Ch) methods, but a long-term comparison has not been undertaken. This study was conducted to assess the agreement between EC and Ch techniques when measuring CO₂ flux during fallow periods of a corn-soybean rotation. From 2004 to 2007, we monitored CO₂ fluxes by both continuous EC and hourly Ch measurements. One flux station and two automated soil chambers were permanently deployed in both corn and soybean fields. In November and December, cumulative CO₂ effluxes by EC method showed a tendency towards greater values following a corn crop than after soybean (37.1 ± 5.0 vs. 24.8 ± 7.2 g C m^-2), while the period from January to April showed no significant differences across crops and years. This initial difference in the early fall period may be explained by both greater residue production from corn than soybean as well as fall tillage of the corn field. When contrasting EC to Ch measurements for November and December of 2004, we observed 18 % greater cumulative CO₂ effluxes with the Ch method (54.3 vs. 45.9 g C m^-2). However, this disagreement was not evenly distributed through these two months. The comparison of daily CO₂ fluxes by the two techniques revealed episodic patterns of disagreement with 42 % (3.5 g C m^-2) of the total disagreement concentrated within five consecutive days in late November. The best agreement between the two techniques was found with the highest air temperatures and in absence of snow cover. Both methods captured nearly the same temporal variability of daily measurements in the two months with coefficient of variations ranging from 80 to 90 %. Use of different methods to obtain CO₂ fluxes need to be evaluated carefully to understand the spatial and temporal patterns of the differences.