Regional-scale Predictions of Agricultural N Losses in an Area with a High Livestock Density

Abstract

The quantification of the N losses in territories characterised by intensive animal stocking is of primary importance. The development of simulation models coupled to a GIS, or of simple environmental indicators, is strategic to suggest the best specific management practices. The aims of this work were: a) to couple a GIS to a simulation model in order to predict N losses; b) to estimate leaching and gaseous N losses from a territory with intensive livestock farming; c) to derive a simplified empirical metamodel from the model output that could be used to rank the relative importance of the variables which influence N losses and to extend the results to homogeneous situations. The work was carried out in a 7773 ha area in the Western Po plain in Italy. This area was chosen because it is characterised by intensive animal husbandry and might soon be included in the nitrate vulnerable zones. The high N load, the shallow water table and the coarse type of sub-soil sediments contribute to the vulnerability to N leaching. A CropSyst simulation model was coupled to a GIS, to account for the soil surface N budget. A linear multiple regression approach was used to describe the influence of a series of independent variables on the N leaching, the N gaseous losses (including volatilisation and denitrification) and on the sum of the two. Despite the fact that the available GIS was very detailed, a great deal of information necessary to run the model was lacking. Further soil measurements concerning soil hydrology, soil nitrate content and water table depth proved very valuable to integrate the data contained in the GIS in order to produce reliable input for the model. The results showed that the soils influence both the quantity and the pathways of the N losses to a great extent. The ratio between the N losses and the N supplied varied between 20 and 38%. The metamodel shows that manure input always played the most important role in determining the N losses. Other important factors in draining soils were mineral fertilisation and potential drainage, which played a minor role in poorly draining soils. The drainage of the soil and the water table depth also played an important role by influencing the soil water content and N losses. The differences among the crops resulted to be very important and it was impossible to produce any indicator to predict N losses for all the crops together. Estimations of total losses are more reliable than estimations of leaching alone. Estimations of total N losses for maize and meadows are more reliable than for winter cereals.