Separating N2O production and consumption in intact agricultural soil cores at different moisture contents and depths.

Agricultural soils are a major source of the potent greenhouse gas and ozone depleting substance, N2O. To implement management practices that minimize microbial N2O production and maximize its consumption (i.e., complete denitrification), we must understand the interplay between simultaneously occurring biological and physical processes, especially how this changes with soil depth. Meaningfully disentangling of these processes is challenging and typical N2O flux measurement techniques provide little insight into subsurface mechanisms. In addition, denitrification studies are often conducted on sieved soil in altered O2 environments which relate poorly to in situ field conditions. Here, we developed a novel incubation system with headspaces both above and below the soil cores and field-relevant O2 concentrations to better represent in situ conditions. We incubated intact sandy clay loam textured agricultural topsoil (0-10 cm) and subsoil (50-60 cm) cores for 3-4 days at 50% and 70% water-filled pore space, respectively. 15N-N2O pool dilution and an SF6 tracer were injected below the cores to determine the relative diffusivity and the net N2O emission and gross N2O emission and consumption fluxes. The relationship between calculated fluxes from the below and above soil core headspaces confirmed that the system performed well. Relative diffusivity did not vary with depth, likely due to the preservation of preferential flow pathways in the intact cores. Gross N2O emission and uptake also did not differ with depth but were higher in the drier cores, contrary to expectation. We speculate this was due to aerobic denitrification being the primary N2O consuming process and simultaneously occurring denitrification and nitrification both producing N2O in the drier cores. We provide further evidence of substantial N2O consumption in drier soil but without net negative N2O emissions. The results from this study are important for the future application of the 15N-N2O pool dilution method and N budgeting and modelling, as required for improving management to minimize N2O losses.

Edaphic factors and plants influence denitrification in soils from a long-term arable experiment.

Factors influencing production of greenhouse gases nitrous oxide (N2O) and nitrogen (N2) in arable soils include high nitrate, moisture and plants; we investigate how differences in the soil microbiome due to antecedent soil treatment additionally influence denitrification. Microbial communities, denitrification gene abundance and gas production in soils from tilled arable plots with contrasting fertilizer inputs (no N, mineral N, FYM) and regenerated woodland in the long-term Broadbalk field experiment were investigated. Soil was transferred to pots, kept bare or planted with wheat and after 6 weeks, transferred to sealed chambers with or without K15NO3 fertilizer for 4 days; N2O and N2 were measured daily. Concentrations of N2O were higher when fertilizer was added, lower in the presence of plants, whilst N2 increased over time and with plants. Prior soil treatment but not exposure to N-fertiliser or plants during the experiment influenced denitrification gene (nirK, nirS, nosZI, nosZII) relative abundance. Under our experimental conditions, denitrification generated mostly N2; N2O was around 2% of total gaseous N2 + N2O. Prior long-term soil management influenced the soil microbiome and abundance of denitrification genes. The production of N2O was driven by nitrate availability and N2 generation increased in the presence of plants.

Stable carbon isotope analysis of fluvial sediment fluxes over two contrasting C(4) -C(3) semi-arid vegetation transitions.

RATIONALE: Globally, many drylands are experiencing the encroachment of woody vegetation into grasslands. These changes in ecosystem structure and processes can result in increased sediment and nutrient fluxes due to fluvial erosion. As these changes are often accompanied by a shift from C(4) to C(3) vegetation with characteristic δ(13) C values, stable isotope analysis provides a promising mechanism for tracing these fluxes. METHODS: Input vegetation, surface sediment and fluvially eroded sediment samples were collected across two contrasting C(4) -C(3) dryland vegetation transitions in New Mexico, USA. Isotope ratio mass spectrometric analyses were performed using a Carlo Erba NA2000 analyser interfaced to a SerCon 20-22 isotope ratio mass spectrometer to determine bulk δ(13) C values. RESULTS: Stable isotope analyses of contemporary input vegetation and surface sediments over the monitored transitions showed significant differences (p <0.05) in the bulk δ(13) C values of C(4) Bouteloua sp. (grama) grassland, C(3) Larrea tridentata (creosote) shrubland and C(3) Pinus edulis/Juniperus monosperma (piñon-juniper) woodland sites. Significantly, this distinctive δ(13) C value was maintained in the bulk δ(13) C values of fluvially eroded sediment from each of the sites, with no significant variation between surface sediment and eroded sediment values. CONCLUSIONS: The significant differences in bulk δ(13) C values between sites were dependent on vegetation input. Importantly, these values were robustly expressed in fluvially eroded sediments, suggesting that stable isotope analysis is suitable for tracing sediment fluxes. Due to the prevalent nature of these dryland vegetation transitions in the USA and globally, further development of stable isotope ratio mass spectrometry has provided a valuable tool for enhanced understanding of functional changes in these ecosystems.

Understanding spatial variability of soil properties: a key step in establishing field- to farm-scale agro-ecosystem experiments.

RATIONALE: The spatial variability of soil properties is poorly understood, despite its importance in designing appropriate experimental sampling strategies. As preparation for a farm-scale agro-ecosystem services monitoring project, the 'North Wyke Farm Platform', there was a need to assess the spatial variability of key soil chemical and physical properties. METHODS: The field-scale spatial variability of soil chemical (total N, total C, soil organic matter), soil physical properties (bulk density and particle size distribution) and stable isotope ratios (δ(13) C and δ(15) N values) was studied using geostatistical approaches in an intensively managed grassland. RESULTS: The scales over which stable isotopes vary (ranges: 212-258 m) were larger than those of the total nutrients, soil organic matter and bulk density (ranges: 84-170 m). Two visually and statistically distinct areas of Great Field (north and south) were identified in terms of co-occurring high/low values of several soil properties. CONCLUSIONS: The resulting patterns of spatial variability suggest lower spatial variability of stable isotopes than that of total nutrients, soil organic matter and bulk density. Future sampling regimes should be conducted in a grid with <85 m distance between sampling locations to sufficiently capture the spatial variability of the measured soil properties on the 'North Wyke Farm Platform'. Consultation of the management histories of the sampled field revealed that it had previously comprised two fields with contrasting management histories, suggesting an effect of management legacy (>5 years) on the patterns of spatial variability.

Long-term release of carbon from grassland soil amended with different slurry particle size fractions: a laboratory incubation study.

Application of animal manure to agricultural soils enhances both native soil carbon (C) and overall (native soil C and added C) respiration. CO(2) effluxes were measured in a laboratory incubation study for 1465 days after the application of different slurry fractions (>2000, 425-2000, 250-425, 150-250, 45-150 and <45 µm) to a grassland soil. The slurry-derived C present in the soil was traced using the natural abundance δ(13)C method. We used two kinetic (single and two pool) models to fit the experimental data and to test the model validity with respect to long-term data sets. Mean residence times (MRTs) of the particle size based slurry-C fractions were estimated using these models and a linear (13)C natural abundance based approach. The results showed that slurry-C degradation in soil over time varied between the different particle size based slurry treatments. The two kinetic soil-C models were successful to predict medium- to long-term carbon release from soil amended with animal slurry. The estimated MRTs did vary between the linear (3.8-5.6 years) and non-linear based (0.8-3.8 years) (model) approaches. Slurry-derived C could still be (isotopically) detected in the soil 4 years after slurry application using the natural abundance δ(13)C method. This suggests that it may take a decadal timescale or longer before the entire amount of C introduced through whole slurry amendments to grassland soils is fully dissipated.