Search

Recent observations indicate that long-term N additions can slow decomposition, leading to C accumulation in soils, but this process has received limited consideration by models. To address this, we developed a model of soil organic matter (SOM) dynamics to be used with the PnET model and applied it to simulate N addition effects on soil organic carbon (SOC) stocks. We developed the model's SOC turnover times and responses to experimental N additions using measurements from the Harvard Forest, Massachusetts. We compared model outcomes to SOC stocks measured during the 20th year of the Harvard Forest Chronic Nitrogen Amendment Study, which includes control, low (5 g N m yr) and high (15 g N m yr) N addition to hardwood and red pine stands. For unfertilized stands, simulated SOC stocks were within 10 % of measurements. Simulations that used measured changes in decomposition rates in response to N accurately captured SOC stocks in the hardwood low N and pine high N treatment, but greatly underestimated SOC stocks in the hardwood high N and the pine low N treatments. Simulated total SOC response to experimental N addition resulted in accumulation of 5.3-7.9 kg C per kg N following N addition at 5 g N m yr and 4.1-5.3 kg C per kg N following N addition at 15 g N m yr. Model simulations suggested that ambient atmospheric N deposition at the Harvard Forest (currently 0.8 g N m yr) has led to an increase in cumulative O, A, and B horizons C stocks of 211 g C m (3.9 kg C per kg N) and 114 g C m (2.1 kg C per kg N) for hardwood and pine stands, respectively. Simulated SOC accumulation is primarily driven by the modeled decrease in SOM decomposition in the Oa horizon. [ABSTRACT FROM AUTHOR]

We modeled the expected range of seasonal and annual N2O flux from temperate, grain agroecosystems using Monte Carlo sampling of N2O flux field observations. This analysis is complimentary to mechanistic biogeochemical model outcomes and provides an alternative method of estimating N2O flux. Our analysis produced a range of annual N2O gas flux estimates with mean values overlapping with results from an intermodel comparison of mechanistic models. Mean seasonal N2O flux was 1–4% of available N, while median seasonal N2O flux was less than 2% of available N across corn, soybean, wheat, ryegrass, legume, and bare fallow systems. The 25th–75th percentile values for simulated average annualized N2O flux rates ranged from 1 to 12.2 kg N ha−1 in the conventional system, from 1.3 to 8.8 kg N ha−1 in the cover crop rotation, and from 0.8 to 9.3 kg N ha−1 in the legume rotation. Although these modeling techniques lack the seasonal resolution of mechanistic models, model outcomes are based on measured field observations. Given the large variation in seasonal N gas flux predictions resulting from the application of mechanistic simulation models, this data-derived approach is a complimentary benchmark for assessing the impact of agricultural policy on greenhouse gas emissions. [ABSTRACT FROM AUTHOR]