A urine patch framework to simulate nitrogen leaching on New Zealand dairy farms

On New Zealand dairy farms, it is the nitrogen excreted directly onto pasture, particularly urine, that drives nitrogen (N) leaching from the farm. A new framework (UPF: Urine Patch Framework) is presented that post-processes the results of a whole farm model and runs a mechanistic soil model to simulate the urine patches. Two alternative methods to simulate the spatial distribution of urine patches were implemented and compared (Grid: spatially explicit, and Probabilistic: based on the probability of different temporal urination patterns). This paper describes the implementation of these two methods in connection with a Whole Farm Model; and compares the N leaching predictions with observed data. Two examples are provided, one analyzing the impact of urine patch overlap and another, the relative risk of N leaching at different times of urinary N deposition. The model showed good correlation and predictive ability between simulated annual N leaching results and observed data [R² = 94 %, mean relative prediction error (MRPE) = 10 % for Grid and R² = 72 %, MRPE = 20 % for Probabilistic]. The two methods produced similar results across an 8-year period for monthly and annual N leaching (R² = 96 %, MRPE = 10 % and R² = 86 %, MRPE = 8 %; respectively). Only 8 % of the paddock area was covered with multiple urinations during 1 year, but as much as 39 % of the total urine volume was deposited on overlapped patches. Systematically removing all urinary N for 1 month in either May or June reduced N leaching by approximately 20 %. Avoiding urinary N deposition during autumn or early winter could be highly effective in mitigating N leached during the following winter.