Approaches to Model Nitrogen Leaching in Agroecosystems: Current Status and Future Challenges.

Historically the amount of nitrogen (N) leached from the agroecosystems has increased over time leading to water quality problems, but our capacity to predict N leaching has remained the same and in general not yet well developed for reliable and accurate simulations across the landscape. The complexities and challenges in measuring subsoil N dynamics have led to the development of different equations and assumptions to model N leaching in simulation models. Here we review different approaches used in simulation models to predict N leaching, strengths and limitations, as well methodological issues on N leaching definition. Both mechanistic (Richards’ equation) and empirical (tipping bucket) approaches used in crop models to simulate water balance and N movement through the profile are included in this analysis, with particular emphasis on one-dimensional models. Mechanistic approaches use convection-dispersion equation for solute transport while empirical approaches uses mobility or efficiency coefficients to calculate the amount of solute leaving the layer, which in general is moving slower than the water. Modeling of N leaching is further complicated by the inconsistent definitions used by agronomists and modelers to refer to N leaching: a) amount of N leached from the root zone, b) amount of N leached from the bottom of the profile, and c) amount of N leached in subsurface drainage systems. The complexity in modeling N leaching increases in fields with shallow and fluctuating water tables that can bring back part of the N leached to the root zone. Simulation examples illustrating trade-offs between N leaching and crop N uptake among different approaches are presented. In conclusion, improving our capacity to predict and explain N leaching requires a strategic approach that involves high-resolution field measurements combined with modelling techniques to deeper understand the critical processes that control N leaching.