There is much interest in the role that agricultural practices might play in sequestering carbon to help offset rising atmospheric CO2 concentrations. A number of management methods that might increase soil C levels have been suggested, but there are scant available data to properly support recommendations or policy changes. We have used eddy covariance measurements of net ecosystem exchange (NEE) in contrasting management systems to discern the impact of two specific practices, reduced tillage and a spring cover crop in the soybean year, on the biennial C balance of a corn/soybean rotation, the dominant cropping system in much of the midwestern United States. Measurements commenced in fall 2001, immediately following corn harvest and tillage, and continued through a year of soybean and a year of corn. One of the two fields was farmed conventionally (CONV), with fall chisel/disk tillage after each harvest, soybean planting in late May (2002), and corn planting in early May (2003). In the alternative field (ALT), we used reduced tillage (strip till) each fall following harvest, and a spring oats cover crop in the soybean year (2002) that was planted in early April, then killed with a herbicide shortly after soybean planting. Both fields have the same soil type, and were similarly instrumented, with a sonic anemometer and open-path infrared gas analyzer. Reduced tillage resulted in somewhat lower soil respiration rates in both autumns in ALT, relative to CONV. Also, the spring oats cover crop prior to soybean did fix additional C, but it was rapidly respired after the oats were killed, and the surface crop residue slowed the initial development of the subsequent soybean crop. Soybean yields for the two fields were similar, but slightly higher for CONV, a pattern that was more pronounced with corn the following year. Overall, cumulative NEE was larger (more C fixed) in the conventional field, but C removed in yield in the conventional field was larger too, so that the apparent change in soil organic carbon (ΔSOC), estimated as NEE – harvested C, in the two fields was nearly identical. In both treatments the apparent ΔSOC was negative (approximately 90 g C m−2 SOC lost over the biennium, or about 20% of cumulative NEE) but this may be at least partially due to systematic underestimation by eddy covariance rather than an actual loss of SOC. We conclude that neither of these management practices (reduced tillage, spring cover crop) resulted in any C sequestration within the first two years of implementation that is resolvable with current measurement methods.
