MODELLING NITROGEN DYNAMICS IN CROP AND SOIL: FROM SITE-SPECIFIC TO REGIONAL APPLICATION IN NORTHERN ITALY.

The aim of my Ph.D. work was to investigate the main factors related to N leaching from arable land under different pedoclimatic conditions and cropping systems in Lombardia plain. In situ monitoring and modelling analysis were defined to evaluate the potential N losses via leaching from arable land and the effect of agricultural management. At monitoring sites, representative of Lombardia arable land, data of soil, crop, water, and N-related variables were collected for a period from 2 to 5 years. Soil characterization, crop yield, leaf area index, harvest index, crop nitrogen uptake, soil water content and soil solution nitrogen concentration at different depths were measured over the monitoring period. All the collected data were used to calibrate and validate the ARMOSA model. Such dynamic model was developed by our research team to predict N leaching risk from arable land in northern Italy. The calibration and validation procedures allowed to parameterize (i) pedological parameters related to soil water balance and nitrogen dynamics, such as mineralization, denitrification, volatilization, wet and dry atmospheric deposition, immobilization; (ii) six crops growth and development parameters which lead the gross assimilation of CO2, leaf area index, stem and root elongation, respiration loss, nitrogen dilution curve, crop development based on growing degree days, dry matter partitioning, evapotranspiration and residuals calculation. The outstanding result was that crop, water and N-related variables were accurately simulated, being in full agreement with observed data. Once calibrated and validated, ARMOSA model was applied at regional scale in order to evaluate the potential risk of N leaching. The model run over 20 years in 35 simulation units, obtained by dividing Lombardia plain in homogenous districts in terms of pedological, climatic and cropping systems features. Each district was characterized by two representative soil types, meteorological observed data set, crop rotations according to the regional land use analysis, organic N load, calculated on the basis of livestock density. With regard to results, similar or even higher N use efficiency resulted with increasing organic N supply and proportionally reduced mineral fertilization. In such way, N leaching decreased by half in maize-based forage systems. Moreover, the eventual choice to introduce a catch crop in rotation strongly contributed to minimize N leaching.