SIMULATING PROCESSES OF SOIL PHOSPHORUS IN GEOLOGICALLY YOUNG ACIDIC SOILS OF FINLAND

Abstract

To properly simulate phosphorus (P) loading from agricultural land to waters, changes in soil P status and processesaffecting soil P have to be described adequately. For practical reasons, a model has to rely on data that are commonlyavailable. In this study, the P submodel of the ICECREAM model, based on the CREAMS, GLEAMS, and EPIC models, wasadapted to simulate young acidic soils, which prevail in Finland. These soils have a high P sorption capacity, plenty of inorganicP in a non-reactive apatitic form, and the organic matter content (Or) in the plow layer is commonly about 5%. Themodel was modified mainly on the basis of 12 long-term fertilization experiments and tested on three other soils. The P poolsconsidered were labile P (Pil), active inorganic P (Pia), stable inorganic P (Pis), fresh organic P (Pof), and more stable organicP (Poh). The major model modifications made were: (1) initialization of Pil from the Finnish soil test results (ammoniumacetate extractable P, PAAc); (2) initialization of Pia and Pis from Pil, assuming equilibrium; (3) revision of the P availabilityfactor (Fl), which regulates the flow between Pil and Pia; and (4) calculation of the initial values of Poh from soil organiccarbon content. These modifications aimed to improve simulation of Pil on the basis of data obtained from soil testing. Propersimulation of changes in Pil, reflected by the soil test P (PAAc), is critical since it regulates the level of dissolved P in runoffwater. After the modifications, the model was able to predict changes in PAAc at annual P application rates of 0 and 30 kgha-1 in cereal cropping, but could not fully cope with a large surplus of P resulting from an annual rate of 60 kg P ha-1.