Study the influence of environmental drivers on N2O emissions under semi-controlled conditions using a rainfall simulator
Abstract
Understanding the dynamics of N2O emissions, the third anthropogenic greenhouse gas, is a major environmental issue. N2O atmospheric concentration, mainly due to soil emissions, rises globally resulting from human activities, especially agricultural practices related to N fertilisation. Microbial denitrification is considered as the major contributor of N2O production, responsible for 40% to 85% of global emissions. Denitrification rate in field is influenced by several regulating factors interacting with time at the scale from the microsite to the landscape (availability of inorganic nitrogen, soil structure and aeration in relation to volumetric water content, organic carbon, pH and temperature). In this work we aimed at analysing the relationships between N2O emissions dynamics and hydric functioning of a cultivated fertilised soil, to better understand N2O production hotspots and hot moments to help reducing uncertainty in N2O emissions estimation. The effect of some environmental parameters, like topography, are hard to assess in field because of the number of regulating factors involved in N2O production and emissions. Oppositely, laboratory studies can easily isolate one factor but they are generally conducted in conditions very different from field, on small samples. Many replicates are necessary to assess studied effects, but it is impossible to reproduce spatial interaction and the effect of distal factors such as topography.
In this study, a 10 m² sloped soil model (0.3 m depth) was installed inside an experimental hall under a rainfall simulator located at INRAE Val-de-Loire Research Centre, which results in an intermediate scale between soil cores and field. This equipment was built in accordance with the rainfall simulator design of the National Soil Erosion Research Laboratory (USDA-ARS, West Lafayette, IN), with 2 rows of 5 swipping nozzles located at 5.1 m over the soil model. Rainfall characteristics can be modified through changes in nozzle type, water pressure and sweep frequency. For this experiment, 10 Veejet flat spray nozzles were used (on each row: 2 x 8070, 2 x 8060 and 1 x 8050) with a pressure of 0.85 bar for a rainfall intensity of 17 mm.h-1. The soil model was cropped (spring barley) and fertilized to reproduce field conditions (3.5° slope). N2O emissions were measured for 70 days at 16 locations, using a fast-box chamber and a quantum cascade laser spectrometer. Soil characteristics (nitrogen, pH, bulk density, water content) were measured up, mid and downstream. We aimed at assessing the temporal and spatial variability of N2O emissions at the model scale, in semi-controlled conditions (fertilisation and rainfall events), considering the interaction of two distal factors (topography and plants).
The N2O emissions measured were in accordance with field measurements. On all points, a first N2O peak was observed after a 24 mm rainfall following a nitrogen application (two conditions favouring denitrification), and a second but lower peak was observed after a 8 mm rainfall. We observed a slope impact, affecting the dynamics of the water processes (lateral flows, accumulation of water at the bottom of the slope) and inducing stronger emissions at the downstream position. Plants also had an impact on N2O emissions, which were significantly higher where the crops had grown better, which increased soil porosity and gas diffusion, limited soil of compaction caused by rainfalls and enhanced the organic matter inputs.
Origin | Files produced by the author(s) |
---|---|
Licence |
Public Domain
|