Farming contributes to emissions of greenhouse gases (CO₂, CH₄, N₂O). It is in large part responsible for N₂O emissions, with land under cultivation contributing around 67% of total emissions in France (CITEPA, 2007). While the risks of emissions are particularly high in the case of mineral fertilisation, measurements of N2O emissions are rarer if organic soil conditioners are used. These are used to increase the OM content in soils. It is important to determine the proportion of OM liable to be mineralised after input, since it diminishes the conditioning value of products input and contributes to CO2 emissions by soils.
Farmed soils also contribute around 5.5% of NO emissions in the atmosphere compared with all other sources in France (CITEPA, 2007). However, given the strong reactivity of this gas and the location of sources, the NO produced by agricultural soils is liable to contribute significantly to the formation of tropospheric ozone in rural areas. Although references exist as far as effluents from livestocks are concerned, there is a considerable shortage of data on the input into the soil of residual products of an urban origin.
This is why measurements of gaseous flows of CO2, N2O and NO were made following the input of various ROPs used at the site in Feucherolles using an automated chamber method. The objective was to obtain initial references concerning the environmental impact of the spreading of ROPs on the field in terms of the emission of greenhouse gas (CO2, N2O) and atmospheric pollutants (NO).
Six automated chambers were put in place. In this system, the chambers were opened and closed using electrically controlled jacks. Once closed, the composition of the air inside the chamber was analysed continuously using pipes and various electromagnetic valves that route the air from the chamber to various gas analysers. The flows are deducted from the rate of accumulation of biogas in the closed chamber, its volume (55 l) and its surface area (1/2 m²). The gas analysers used are simple commercial infrared spectrometers (model 46C, Thermo Environment; modele 820, LICOR, respectively) for N2O and CO2 and a chemiluminescence analyser for detecting NO (modele 42CTL, Thermo Environment). The change in concentrations inside the chamber is recorded using a data acquisition unit every 10 seconds. The closure time per chamber is 15 minutes. The full cycle for the six chambers is therefore 1 hour and 30 minutes.

For the 6 chambers, 6 different treatments were carried out:

a nitrogen solution (UAN) for the equivalent of 117 kgN ha-1 (¾ ammonium and ¼ nitrate)
a green waste plus sludge (GWS) compost
a biowaste (BIO) compost
a residual municipal waste (RMW) compost
a manure
a control

For the 3 composts and the manure, the quantities applied were adjusted to input the equivalent of 4 tC ha-1, or the dose input into the field. The waste was incorporated into the first five centimetres of the soil, corresponding to surface tillage carried out in the field. For the control, a treatment that did not receive any matter, we carried out the same soil work at the same depth.
The gaseous flows were observed for a month and a half after the treatments, from 12 September to 26 October 2006.

Release of carbon dioxide (CO2)

There was a strong increase in flows with the RMW treatment. GWS, BIO and manure were more or less equivalent. After a month and a half, the 'accumulation' functions of the diminished CO2 losses in the control showed no horizontal asymptote and therefore indicate still-active mineralisation of exogenous organic matter. The mineralised proportions are compatible with the quantities remaining in the soil and contributing to the increase in OM levels in the soils (33% for RMW, 45% for GWS, 50% for manure and 55% for BIO). Meanwhile, it appears that the mineralisation of the fraction of easily biodegradable organic C occurs very rapidly after input.

Release of nitrous oxide (N2O)

In the UAN treatment this was low compared with the average emission factor of 1.25% traditionally seen for this gas, undoubtedly due to a technical problem with this treatment. Emissions rose just after application of the various organic treatments, except with BIO, for which no significant additional emission was observed. The flows are significant for around 6 days after input for RMW, manure and UAN. Meanwhile, for RMW, emissions continue for 15 days.

Gaseous release of nitric oxide (NO)

As with N2O, these are significantly higher during the 6 to 8 days following input in the case of RMW, manure and UAN. Again, due to a sealing problem in the chamber, the losses should undoubtedly have been higher in the UAN treatment. In the case of GWS, emissions last for a fortnight and are attributable to its high ammonium content and perhaps also a rapid mineralisation of organic nitrogen for this product.
Mineral analyses of ammonium and nitrate in the soil at the 0-15 cm layer, performed using soil samples taken from experimental plots that received the same applications of effluents, appear to indicate that the release of N2O and NO are linked with the dynamic of ammonium and therefore probably nitrification. The more the ROPs initially contain or rapidly produce NH4 through mineralisation, the higher the emissions of N2O and NO are (Laville, 2008; Wrage, 2001).

In conclusion

It appears that the easily biodegradable organic fraction of ROPs is rapidly mineralised after their input into the soil. The summary of the results of C storage in the soils and its derivative effects in terms of soil properties and CO2 emission measurements in the field will be used in the environmental evaluation of the return of ROPs to the soil. Nitrous oxide emissions (NO and N2O) depend on the ammonium content of the ROPs. A portion of the organic nitrogen contained in the ROPs is also liable to contribute to emissions after mineralisation. These initial results require additional laboratory measurements based on soil columns, which would complete the assessment of gaseous losses for CO2. These measurements would also allow the treatments to be repeated and the field observations to be validated.