Pig slurry treatment modifies slurry composition, N2O, and CO2 emissions after soil incorporation

Résumé

The treatment of manures may improve their agricultural value and environmental quality, for instance with regards to greenhouse gases mitigation and enhancement of carbon (C) sequestration. The present study verified whether different pig slurry treatments (i.e. solid/liquid separation and anaerobic digestion) changed slurry composition. The effect of the slurry composition on N2O and CO2 emissions, denitrification and soil mineral nitrogen (N), after soil incorporation, was also examined during a 58-day mesocosm study. The treatments included a non-treated pig slurry (NT), the solid fraction (SF), and the liquid fraction (LF) of a pig slurry and the anaerobically digested liquid fraction (DG). Finally, a non-fertilized (N0) and a treatment with urea (UR) were also present. The N2O emissions measured represented 4.8%, 2.6%, 1.8%, 1.0% and 0.9% of N supplied with slurry/fertilizer for NT, LF, DG, SF and UR, respectively. Cumulative CO2 emissions ranged from 0.40 g CO2-C kg¿1 soil (0.38 Mg CO2-C ha¿1) to 0.80 g CO2-C kg¿1 soil (0.75 Mg CO2-C ha¿1). They were highest for SF (56% of C applied), followed by NT (189% of C applied), LF (337% of C applied) and DG (321% of C applied). Ammonium was detected in the soil for all treatments only at day one, while nitrate concentration increased linearly from day 15 to day 58, at a rate independent of the type of slurry/fertilizer applied. The nitrate recovery at day 58 was 39% of the N applied for NT, 19% for SF, 52% for LF, 67% for DG, and 41% for UR. The solid fraction generally produced higher potential denitrification fluxes (75.3 for SF, 56.7 for NT, 53.6 for LF, 47.7 for DG and 39.7 mg N2O + N2-N kg¿1 soil for UR). The high variability of actual denitrification results obfuscated any treatment effect. We conclude that treatment strongly affects slurry composition (mainly its C, fibre and NH4+ content), and hence N2O and CO2 emission patterns as well as denitrification processes and nitrate availability. In particular, the solid fraction obtained after mechanical separation produced the most pronounced difference, while the liquid fraction and the anaerobically digested liquid fraction did not show significant difference with respect to the original slurry for any of the measured parameters. Combining data from the different fractions we showed that separation of slurry leads to reduced N2O emissions, irrespective of whether the liquid fraction is digested or not. Furthermore, our results suggested that the default emission factor for N2O emissions inventory is too low for both the non-treated pig slurry and its liquid fraction (digested or not), and too high for the separated solid fraction and urea