Assessment of LEACHN and a simple compartmental model to simulate nitrogen dynamics in citrus orchards

Autores UPV
Revista Agricultural Water Management


A simple compartmental model using a tipping bucket approach for the water dynamics coupled with a nitrogen¿carbon transformations model has been adapted to simulate the soil nitrogen and water balance in mature orange groves on a daily step. This model has been compared with the more mechanistic LEACHN model (the N module of the LEACHM model), which uses Richards¿ equation to simulate soil water movement in unsaturated conditions, the convection¿dispersion equation for solute transport, and that, in addition to including evapotranspiration, N transformations and N plant uptake as in the compartmental model, it also considers gaseous losses due to denitrification and ammonia volatilization, that are not considered in the compartmental model. This comparison was made using data from a three-year experiment in a citrus orchard with two nitrogen fertilization rates. After calibration using the first year data, a reasonable match between simulated and measured values in both models was observed for soil water storage in the whole profile for the validation period (2nd and 3rd year), but the agreement was not so good for the soil mineral nitrogen content. In spite of the differences in the nature and in the complexity of the two models, the soil water dynamics and drainage were well simulated during the whole period by both models. However, the LEACHN model predicted nitrate leaching better than the compartmental model, probably because it considers the nitrogen cycle in a more detailed way. This work is the first calibration and performance evaluation of the LEACHN model for citrus in the Mediterranean area and the results obtained in this study indicate that this model can be a valid tool to evaluate the effects of irrigation and N management on nitrate leaching. The compartmental model has a lower data requirement and calibration is less complex than the LEACHN model and, therefore, may be more appealing for advisory N management purposes.