Modeling N2O Emissions From Organic and Conventional Cropping Systems in Maryland, USA with DNDC.

The impacts of agricultural management on soil N₂O emissions remain poorly characterized for many practices and many regions.  Mathematical models can help fill these data gaps but most have not been widely validated in diverse agricultural systems.  We compared soil N₂O emissions predicted using a process based model, Denitrification-Decomposition (DNDC), with N₂O emissions data collected in 2005 and 2006 from the long-term USDA-ARS Farming Systems Project in Beltsville, Maryland.  We collected soil N₂O emissions data from the corn phase of two mineral fertilizer based corn-rye-soybean-wheat/soybean rotations under no-till (NT) or chisel till (CT) management; and one corn-rye-soybean-wheat/vetch rotation that receives poultry litter two of every three years (Org3).  We hypothesized that DNDC would better predict soil N₂O emissions from NT and CT than Org3 since there are few long-term datasets from systems relying on organic nutrient sources available for calibrating process-based models.  Measured N₂O emissions in 2005 were similar in the three cropping systems (mean, 2.68 g N₂O-N ha^-1 yr^-1) while DNDC predictions were similar in CT and Org3 (mean, 4.0 g N₂O-N ha^-1 yr^-1) and slightly greater in NT (4.6 g N₂O-N ha^-1 yr^-1).  In 2006, measured emissions were about twice as great in CT and NT (mean, 3.7 g N₂O-N ha^-1 yr^-1) than in Org3 (1.8 g N₂O-N ha^-1 yr^-1).  DNDC predictions, however, indicated greater N₂O emissions in Org3 and CT (5.2 g N₂O-N ha^-1 yr^-1) than in NT (3.7 g N₂O-N ha^-1 yr^-1).  Results indicate that DNDC predicted annual N₂O emissions reasonably well.  However, rankings of these three cropping systems differed based on whether N₂O emissions were measured or predicted.  We will discuss potential reasons for and implications of these discrepancies.