The challenges of incorporating urine and dung patches in process-based modelling of grazed agricultural systems

Animals are well known to be important in lateral transfers of nitrogen within the farm boundary. Those transfers can be categorised into those that are: (a) primarily random and smallscale dung and urine patches within a grazed paddock, (b) larger and systematic transfers resulting from preferred grazing and resting areas and (c) those that are additionally mediated by management actions such as manure management from housing systems. We review how key simulation models treat the random category of nitrogen transfers and develop recommendations for usage and further development. While urine patches are clearly understood to be important in N cycling in grazed pastures, few simulation models consider the effect of the patchy urine deposition, primarily because the inclusion of the deposition process greatly increases the complexity of the simulation model. Some modellers choose to ignore the effect – and if the primary aim is to model production rather than environmental effect then that position may be quite justifiable. Some authors have taken an approach to model the pasture and leaching at the scale of an isolated patch and then scale up to the paddock level post-simulation. These solutions require relatively little model development and are computationally-efficient but they cannot capture the effects of feedbacks into the system caused by changes in N fluxes and cycling. Far more complex solutions can be informative but they do not lend themselves to practical application at the paddock or farm level. Routine implementation of urine patches into a simulation model will require a balance between realism and pragmatism as it is not generally practical to include the full complexity involved in a whole farm model.