Microbial sources of nitrous oxide in models

To ensure that any proposed management practices for mitigation of N2O (see below) have no detrimental effects on future climate, it is essential that we can also predict emissions from each process under different climate scenarios, from a range of ecosystems and under different management options. Most predictive models only consider nitrification and anaerobic denitrification as sources of N2O, and at the time of writing, nitrogen budgets and models ignore the contribution of nitrate ammonification. Models differ in the way they simulate nitrification and denitrification. For example, DNDC (Li et al, 1992) and ECOSYS (Grant, 1995) simulate microbial growth and associated N2O production (Li, 2007), whereas DAYCENT (Del Grosso et al, 2000; Parton et al, 2001) and CERES (Gabrielle et al, 2006) use empirical formulae to simulate the N2O production. No attempt has been made to include nitrifier denitrification or nitrate ammonification as N2O sources, in part because of a lack of data on the controls of these processes for model parameterization or validation.

Scale is a primary consideration when source partitioning and quantifying N2O emissions, because the prediction of emissions and formulation of policy decisions are made on regional, national or even global scales - fundamentally different scales from that at which most experiments are conducted to understand underlying processes (Standing et al, 2007). We discuss below how, with better understanding of the controls and sources of N2O production in terrestrial systems, it should be possible to develop appropriate mitigation strategies. However, for such strategies to be appropriate and effective, outstanding issues of scale may need to be resolved. This is a major challenge, as technical constraints and reliance on stable isotope approaches have meant that efforts to quantify the contributions of different microbial processes to measured N2O emissions have been focused on the microscale to plot scale, but most predictive models rely on the aggregated response offered at field or landscape scale. Linear relationships are often applied when upscaling from the microscale, thereby losing the level of detail offered at the microscale, and assuming that the hierarchy of controlling parameters on N2O emission is the same across all scales. It is possible that uncertainty in these models can be lowered by consideration of the regulation of microbial sources of N2O and the conditions under which different processes predominate, and inherent in that is the need for integration from the microscale. Modelling approaches should also be able to guide us in both temporally and spatially representative sampling for source partitioning.

0 0

Post a comment