Models fitted to manure emissions contrast with inventory estimates

We developed process based models of N2O and CH4 emissions from pen surface manure in a beef feedlot. We found that much of the variability in emission of nitrous oxide and methane were accounted for by interactions of readily measured manure parameters (e.g. moisture content and temperature). Validation was attempted using data from a second feedlot. We found that the peak N2O emission in manure occurred for a moisture fraction of 0.20 as opposed to 0.59 in soil. This may result from the anaerobic conditions occurring at lower moisture contents in pen manure than occurs in soils. We also developed and validated a physical process-based model of the effect of air temperature, solar radiation and time of day on manure temperature. A physical process-based model was also formulated to characterise the effect of rainfall, evapo-transpiration and drainage on manure moisture. CH4 emissions at the northern feedlot were predicted to be 0.40 kg CH4 head-1 year-1, substantially less than the standard protocol calculated values for manure emission: 4.14 and 1.9 kg CH4 head-1 year-1. N2O emissions at the northern feedlot were predicted to be 0.31 kg N2O head-1 year-1. Standard inventory calculation approaches produce larger estimates of emission: 2.6 and 3.0 kg N2O head-1 year-1. These differences could be partly related to regional conditions and climates, however, the inventory calculation protocols are reliant on a relationship between emission magnitude and total excreted N that was not supported by our study.