«International 17 Workshop th Nitrogen The was jointly organised by Teagasc and AFBI Printed by Print Depot Suggested citation Authors, 2012. Title ...»
1. Background & Objectives Processing of animal slurry is an opportunity to improve nitrogen (N) use efficiency of manure applied to crops. Separation of slurry into liquid and solid fractions, and further treatment of the liquid fraction using reverse osmosis results in a mineral concentrate which may have similar properties as a mineral N fertilizer (Velthof, 2011). A mineral concentrate is a liquid ammonium fertilizer (ammonium content is on average 90% of total N) with a high pH (7.5). The combination of a high ammonium content and high pH increases the risk of ammonia (NH3) emissions. The risk of NH3 emissions can be decreased by application with a low NH3 emission technique. However, these application techniques may increase nitrous oxide (N2O) emission (Velthof and Mosquera, 2011). In a series of incubation studies, the NH3 and N2O emissions from untreated pig slurry, mineral concentrate, and the solid fraction from separated slurry were quantified. The products were both surface applied and injected.
2. Materials & Methods Two experiments were conducted using incubation bottles with sandy soils without crop, and one experiment was carried using cores from grassland on sand, clay, and peat soils (PVC cylinders of 10 cm diameter and 10 cm depth). The first two experiments included different types of fertilizer (calcium ammonium nitrate (CAN), urea, urean (not shown in this abstract), four pig slurries, four mineral concentrates, and four solid fractions, and two application techniques (surface application and incorporation through the soil and were carried out in three replicates (in total 93 incubation bottles in each experiment). The grassland experiment included the same fertilizer types as the experiment with uncropped soil, except solid fractions. All fertilizers were incorporated in the grassland soil to 5 cm depth experiment (in total 108 cores; 36 per soil type). Fluxes of NH3 and N2O were assessed from the increase in NH3 and N2O concentrations in the headspace of the incubation bottles or flux chambers following closure. Concentrations of NH3 and N2O were measured 8 - 12 times during incubation, using a Innova photo-acoustic gas analyzer.
3. Results & Discussion The emissions of NH3 increased immediately after application of slurry, mineral concentrate, solid fraction and urea, and decreased thereafter. The NH3 emissions were much higher after surface application than after incorporation (Figure 1). Total NH3 emissions from mineral concentrate were on average similar to that from pig slurry, at the same total N application rate. The N2O emissions from mineral concentrate increased just after application (up to a factor 3000), and decreased thereafter. Incorporation of all slurries, treated slurries and mineral N fertilizers increased N2O emission in comparison to surface application. On average, the N2O emissions were about a factor
1.5 higher from mineral concentrate than from untreated slurry (Figure 2) and from CAN. Mineral concentrates contain degradable organic carbon compounds which may have increased denitrification (Paul and Beauchamp, 1989). High NH3 concentrations in the soil after application of mineral concentrates may inhibit nitrification and, thereby increase N2O emissions. These effects are likely to be similar as those found in urine patches (Oenema et al., 1997). Acidification and/or removal of organic carbon from mineral concentrates may be measures to decrease N2O emissions from mineral concentrates.
Figure 1. Total NH3 emission (average ± s.
e.) in experiment I (arable soil). The roman numerals indicate the four different slurries, concentrates and solid fractions).
4. Conclusion A mineral concentrate from treated slurry is a fertilizer with a relatively high risk of NH3 emissions.
However, incorporation of mineral concentrates in the soil strongly decreases NH3 emissions. The N2O emissions from mineral concentrate applied to soil were higher than those from untreated pig slurry and calcium ammonium nitrate fertilizer.
References Oenema, O., Velthof, G.L., Yamulki, S. and Jarvis, S.C. 1997. Nitrous oxide emissions from grazed grassland. Soil Use and Management 13, 288-295.
Paul, J.W. and Beauchamp, E.G. 1989. Effect of carbon constituents in manure on denitrification in soil. Canadian Journal of Soil Science 69, 49-61.
Velthof, G.L. and Mosquera, J. 2011. The impact of manure application technique on nitrous oxide emission from agricultural soils. Agriculture, Ecosystems and Environment 140, 298-308.
Velthof, G.L. 2011. Synthesis of the research within the frame-work of the Pilot Mineral Concentrates. Alterra, Wageningen, Alterra-report 2224 (http://www.alterra.wur.nl/UK/)
Nitrogen Workshop 2012
Estimating nitrate emissions to surface water: comparison of methods using detailed regional data and national data Dupas, R.a, Gascuel-Odoux, C.a, Durand, P.a, Parnaudeau, V.a, Delmas, M.b, Deronzier, G.c, Domange, N.c a INRA-AgroCampus Ouest UMR Sol Agro et hydrosystème Spatialisation, Rennes, France.
b INRA Unité Infosol, Orléans c ONEMA French National Agency for Water and Aquatic Environments, Paris, France
1. Background & Objectives The European Union (EU) Water Framework Directive (WFD) requires River Basin District managers to carry out an analysis of nutrient pressures and impacts, in order to evaluate the risk of water bodies failing to reach “good ecological status” and to identify those catchments where prioritized nonpoint-source control measures should be implemented. A methodology is developed to estimate nitrate nonpoint-source emissions to surface water, using readily available data at national scale. In addition to this application at national scale, the model was tested in the Brittany region (western France), where detailed regional databases are available. Brittany is a case worthy of study, as it allows comparing prediction of the models taking into account the national-wide databases and more detailed regional data.
2. Materials & Methods The model is inspired from US model SPARROW (Smith al., 1997) and European model GREEN (Grizzetti et al., 2008), i.e. a statistical approach consisting of linking nitrogen sources and catchment land and river characteristics. The nitrate load (L) at the outlet of each river basin is
 where DS is diffuse sources (i.e. N surplus in kgN.ha-1.yr-1), PS is point sources from domestic and industrial origin (kgN.ha-1.yr-1), and R and B are the river system and basin reduction factors,
respectively. Both factors were calibrated as:
  here Xi and Xj are independent variables for the basin and river reduction factors, respectively, αi and αj are parameters to be calibrated. The model was calibrated to fit mean annual nitrate load in 54 independent catchments ranging from 20km² to 2000km² in Brittany, for the 2004-2007 period.
Variable selection was first performed on a simplified version of the model neglecting PS, in order to allow the linearization of equation . Variables were selected according to Bayesian Information Criterion (BIC) in order to optimize the predictive performance of the model. Table 1 summarizes the variables which were tested and entered into the model, considering the national data and the detailed regional data. Note that variables which are expected to have a positive effect on transfer are entered in a reciprocal form. Hence, all αi and αj are expected to be negative and B and R should be lower than 1.
Secondly, the non-linear least-squares estimates of the parameters in equation  were determined using a Gauss-Newton algorithm.
3. Results & Discussion Figure 1 shows that better fitting is achieved when using the detailed regional data rather than the national data. Residual Standard Error was 6.87 kg N.ha-1.yr-1 in the first case and 9.55 kg N.ha-1.yrin the second case.
Figure 1. Predicted vs observed nitrate loads, considering the national database (left), and the regional database (right)
4. Conclusion This study highlights that regional studies should be carried out in regions where detailed data is available, as a complement to the national scale evaluation.
References Grizzetti, B., Bouraoui, F. and De Marsily, G., 2008. Assessing nitrogen pressures on European surface water. Global Biogeochemical Cycles 22, X Lamouroux, N., Pella H., Vanderbecq A., Sauquet E. and Lejot J. 2010. Estimkart 2.0 : Une plate-forme de modèles écohydrologiques pour contribuer à la gestion des cours d'eau à l'échelle des bassins français. Version provisoire.
Cemagref – Agence de l'Eau Rhône-Méditerranée-Corse – Onema Lemercier-Foucault, B., Lacoste, M., Loum, M. and Walter, C. 2011 (in press). Extrapolation at regional scale of local soil knowledge using boosted classification trees: A two-step approach. Geoderma Smith, R.A., Schwarz, G.E. and Alexander, R.B., 1997. Regional interpretation of water-quality monitoring data. Water Resources Research 33, 2781-2798.
Nitrogen Workshop 2012
Evaluation of a closed tunnel for field-scale measurements of N2O fluxes at the soilatmosphere interface Schäfer, K.a, Böttcher, J.b, Weymann, D.c, von der Heide, C.b, Duijnisveld, W.H.M.d a Karlsruhe Institute of Technology, Institute for Meteorology and Climate Research, Department Atmospheric Environmental Research (IMK-IFU), Garmisch-Partenkirchen, Germany b Leibniz University Hannover, Institute of Soil Science, Hannover, Germany, c Forschungszentrum Jülich GmbH, Agrosphere Institute, Jülich, Germany, d Federal Institute for Geosciences and Natural Resources, Hannover, Germany
1. Background & Objectives Emissions of the powerful greenhouse gas nitrous oxide (N2O) from soils are commonly characterized by huge spatial variability. An upscaling of classical small-scale chamber measurements is thus questionable and may add uncertainty to emission inventories or emission factors. Therefore, field-scale approaches will become increasingly important. Since micrometeorological techniques are limited by stable atmospheric conditions (Pihlatie et al., 2005) and their low spatial resolution (Smith et al., 1994), we used a closed tunnel equipped with an openpath Fourier Transform Infrared (FTIR) spectrometer to (i) evaluate its feasibility for measuring field-scale N2O fluxes from an unfertilized grassland soil and (ii) compare those results with smallscale fluxes obtained from closed chambers.
2. Materials & Methods The measuring tunnel, consisting of a 99×5×0.6 m aluminium liner structure was closed with a commercial plastic cover prior to each measurement. The cover was sealed at the frame and the soil using sand-filled hoses. The FTIR technique enabled precise concentration measurements longitudinal through the whole tunnel atmosphere at five minute intervals. Based on those measurements, we used a non-steady-state approach to calculate the predeployment N2O flux (q0).
This was achieved by taking into account diffusive gas transport between soil and tunnel atmosphere to simulate the N2O concentration at the height of the FTIR measuring beam. We estimated q0 inversely using measured concentration courses at the FTIR beam height, assuming
that the boundary condition at the soil/tunnel interface can be described by:
q0 ql(z 0,t)  (t 1 )ε with ql(z = 0, t): time course of N2O flux at the soil/tunnel interface, : fit parameter (0 0.5).
Predefined representative emission scenarios showed that this approach yields robust results within an uncertainty range of up to ± 30 % for the inversely estimated q0. Concurrent manual chamber measurements were performed in close vicinity to the tunnel, where each of four chambers covered an area of 0.045 m2. N2O fluxes were calculated using the non-steady-state diffusive flux estimator (NDFE) (Livingston et al., 2006).
3. Results & Discussion During twenty-four measuring campaigns we found that the tunnel system is generally feasible for calm and dry weather conditions. Rain and high wind speeds were disadvantageous and add uncertainty to this otherwise precise method. Therefore, we restricted the measurements to the evening hours and to the first hour of tunnel closure. Field-scale N2O fluxes determined by the
Nitrogen Workshop 2012
tunnel method were generally small and in a range between 0 and 28 µg N2O-N m-2 h-1 (Figure 1), which is typical for ‘background’ emissions from an unfertilized grassland site. In contrast, smallscale chamber based fluxes were spatially and temporarily variable (Figure 1). The mean daily N2O fluxes for the chamber measurements ranged between -4 and 228 µg N2O-N m-2 h-1. The cumulative flux for the 11 comparable measuring dates was 513 µg N2O-N m-2 h-1 for the chamber method, but only 61 µg N2O-N m-2 h-1 for the tunnel approach. This difference was mainly caused by peak emissions occurring at three measuring dates which were only exhibited by the chambers.
chambers N2O emission (µg N2O-N m-2 h-1)
-100 1.1.06 1.5.06 1.9.06 1.1.07 1.5.07 1.9.07 1.1.08 1.5.08 date Figure 1. N2O emissions (predeployment flux q0) inversely estimated from tunnel measurements and concurrent chamber-based emissions (four replicates) calculated with NDFE during the whole measurement period.
Since chamber and tunnel measurements represent different scales it seems not appropriate to validate one method with the other one. However, we argue that the chambers were occasionally susceptible to detect hotspots and hot moments of N2O emission, which results in an overestimation of the actual field-scale emission. This confirms the uncertainty associated with an up-scaling of N2O fluxes obtained from the small-scale and underlines the need for field-scale measurements.
4. Conclusion This study introduced a tunnel coupled to an open-path FTIR spectrometer which enables reliable measurements of N2O fluxes particularly during stable atmospheric conditions. We conclude that this flexible field-scale approach has potential to fill an experimental gap between small-scale chamber and ecosystem-level micrometeorological methods.
References Livingston G.P., Hutchinson G.L. and Spartalian K. 2006. Trace gas emission in chambers: a non-steady-state diffusion model. Soil Science Society of America Journal 70, 1459-1469.