Simulating nitrate leaching under crops fertilized with pig slurry in lysimeters

Regions in the Po Valley, Northern Italy, are characterized by intensive crop-livestock farming systems. A simulation model has been chosen for an inter-regional project, which should help in defining groundwater vulnerability and pollution risk on a regional scale, in relation to agricultural land use, by allowing the prediction of nitrate leaching under different climate, soil, crop and management scenarios. The model derives from the coupling of a hydrological model, MACRO, simulating water flow and solute transport in structured soils, with a model simulating soil N dynamics, SOILN. The aim of this work was to test the model's ability to simulate nitrate leaching through soil after land spreading of pig slurries. A dataset obtained from lysimeter experiments which had been carried out in the period 1976-1981 was used for this purpose. Four soil types were compared (silty clay, sandy loam, loam and sand) in factorial combination with four rates of pig slurry (0, 142, 284, 426 g of N m(-2), in, accumulated values from 1976 to 1979) for a seven crop sequence. The efficiency of the MACRO model ranged from 0.96, in the sandy-loam soil, to 0.81, in the sand. Percolation was usually under-estimated, the relative error ranging from 0.7 to 14.6, depending on the soil. The low efficiency of the SOILN model in simulating nitrate leaching is attributed to the lack of knowledge of the mechanisms regulating N transformation processes and especially the mineralization of pig slurry N. This lack of knowledge hampers the correct setting of the N transformation parameter values. A remarkable improvement of the model's performance was obtained by changing a few coefficients which control the mineralization-immobilization turnover of the faeces-organic N. The model efficiency, following this recalibration, ranged from -0.62 to 0.84, and the relative error ranged from -56 to 35, depending on soil and treatment. N leaching was under-estimated at the low pig slurry N application rates and over-estimated at the high ones.