Ammonia volatilization following surface application of raw and treated liquid swine manure.
Chantigny, M.H., MacDonald, J.D., Beaupré, C., Rochette, P., Angers, D.A., Massé, D.I., and Parent, L.-É. (2009). "Ammonia volatilization following surface application of raw and treated liquid swine manure.", Nutrient Cycling in Agroecosystems, 85(3), pp. 275-286. doi : 10.1007/s10705-009-9266-7
The treatment of liquid swine manure (LSM) is primarily intended to produce energy and/or decrease P concentration in the liquid fraction, but may have a simultaneous impact on its potential for N volatilization. We compared NH3 volatilization in the field following surface application (May 2004; May 2005; September 2005) of untreated LSM and the liquid fraction of LSM (hereafter called treated LSM) that had undergone either natural decantation, filtration, anaerobic digestion, or anaerobic digestion + flocculation. Though most treatments increased pH and the proportion of total ammoniacal N (TAN) in LSM, the proportion of applied TAN lost as NH3 was 22% lower with treated than untreated LSMs. Most likely, the increased infiltration rate of treated LSMs, associated with reduction in dry matter, compensated for increases in TAN and pH. Emissions of NH3 on the day of application were correlated with LSM pH (R2 = 0.51) and were the highest with the digested and the digested + flocculated LSMs. However, these LSMs generally emitted less NH3 than the other LSMs on the following days. As a consequence, when the volatilization period was short (e.g., dry soil conditions), the proportion of applied N lost from the digested and the digested + flocculated LSMs was similar or higher than the filtered and decanted LSMs. In contrast, when the volatilization period was long, the digested and digested + flocculated LSMs emitted less NH3 than the other treated LSMs. It can be concluded that LSM treatments tended to decrease NH3 volatilization, compared to the untreated LSM; across application dates, the digested LSM most consistently reduced NH3-N losses whereas the filtered LSM was the least efficient.