Estimates of agricultural C sequestration require an understanding of how net ecosystem productivity (NEP) and net biome productivity (NBP) are affected by land use. Such estimates will most likely be made using mathematical models that have undergone well-constrained tests against field measurements of CO 2 exchange as affected by management. We tested a hydraulically driven soil-plant-atmosphere C and water transfer scheme in ecosys against CO 2 and energy exchange measured by eddy covariance (EC) over irrigated and rainfed no-till maize-soybean rotations at Mead, NE. Correlations between modeled and measured fluxes ( R2>0.8) indicated that <20% of variation in EC fluxes could not be explained by the model. Annual aggregations of modeled fluxes indicated that NEP of irrigated and rainfed soybean in 2002 was -30 and -9 g C m -2 yr -1 (net C source) while NEP of irrigated and rainfed maize in 2003 was 615 and 397 g C m -2 yr -1 (net C sink). These NEPs were within the range of uncertainty in annual NEP estimated from gap-filled EC fluxes. When grain harvests were subtracted from NEP to calculate NBP, both the modeled and measured maize-soybean rotations became net C sources of 40 to 80 g C m -2 yr -1 during 2002 and 2003. Long-term model runs (100 yr) under repeated 2001-2004 weather sequences indicated that a rainfed no-till maize-soybean rotation at Mead would lose about 30 g C m -2 yr -1. Irrigating this rotation would raise SOC by an average of 6 g C m -2 yr -1 over rainfed values. Modeled and measured results indicated only limited opportunity for long-term soil C storage in irrigated or rainfed maize-soybean rotations under the soil, climate, and management typical of intensive crop production in the U.S. Midwest.
