Submitted to: Global Change Biology
Publication Type: Peer reviewed journal
Publication Acceptance Date: 10/11/2009
Publication Date: 10/23/2009
Citation: Castellano, M.J., Schmidt, J.P., Kaye, J.P., Walker, C., Graham, C., Lin, H., Dell, C.J. 2009. Hydrological and biogeochemical controls on the timing and magnitute of nitrous oxide flux across an agricultural landscape. Global Change Biology. 16(10):2711-2720. Interpretive Summary: Reasonable forecasts of greenhouse gas emissions from agricultural soils will depend on mechanistic models that adequately predict emissions under various climatic scenarios and different soil types. Our objective was to determine whether water filled pore space or soil matric potential was a better indicator of nitrous oxide emissions during water table fluctuations for soils representing different landscape positions of the Delmarva Peninsula. Intact soil columns received a usual agricultural nitrogen fertilizer application (100 kg per ha), then were flooded from the bottom to surface, mimicking changing field conditions after a rainstorm. Nitrous oxide emissions, volumetric soil water content, and soil matric potential were measured for 96 hours as the water table receded. Maximum nitrous oxide flux occurred within a narrow range of soil matric potential (1.88 to -4.48 kPa), while the corresponding range of water filled pore space (WFPS) was greater (63 to 98 %). While most estimates of greenhouse gas emissions from soils use WFPS as an indicator of nitrous oxide emissions, soil matric potential was a much better indicator of maximum nitrous oxide flux for the soils of this agricultural landscape.
Technical Abstract: Anticipated increases in precipitation intensity due to climate change may affect hydrological controls on soil nitrous oxide fluxes, resulting in a feedback between climate change and soil greenhouse gas emissions. We evaluated soil hydrology controls on nitrous oxide emissions during experimental water table fluctuations in large, intact soil columns that were amended with 100 kg per ha nitrogen. Soil columns were collected from three landscape positions that vary in hydrological and biogeochemical properties (N = 12 columns). To simulate water table fluctuations that are typical for this site, and expected to increase given future climate change scenarios, we flooded the columns from bottom to surface. After the water table reached the soil surface and the soil was saturated, we allowed the columns to drain freely while monitoring volumetric soil water content, matric potential, and nitrous oxide emissions over 96 hours. Across all landscape positions and replicate soil columns, there was a positive linear relationship between total soil N and the log of cumulative nitrous oxide emissions (r2 = 0.47; p = 0.013). Within individual soil columns, nitrous oxide flux was a Gaussian function of water filled pore space (WFPS) during drainage (mean r2 = 0.90). However, instantaneous maximum nitrous oxide flux rates did not occur at a consistent WFPS across landscape positions and replicate soil columns (range: 63-98% WFPS). In contrast, instantaneous maximum nitrous oxide flux rates occurred within a narrow range of soil matric potential that approximated field capacity (range: -1.88 to -4.48 kPa). The relatively consistent relationship between maximum nitrous oxide flux rates and matric potential indicates that water filled pore size is an important factor affecting soil nitrous oxide fluxes. These data demonstrate that matric potential is the strongest predictor of the timing of nitrous oxide fluxes across soils that differ in texture, structure and bulk density.