Simulation of Phosphorus, Zinc and Cadmium Mass Flow in Dairy Farming Systems

Abstract: A simulation model, FARMFLOW, was developed for the calculation of phosphorus (P), zinc (Zn) and cadmium (Cd) stocks, flows and balances in different parts of a dairy farming system. The simulated continuous change in element stocks within the farming system enables evaluation of the long-term effects of management practices. The FARMFLOW model reproduces feedback mechanisms of the internal cycling of nutrients and trace elements between crops and animals on the dairy farm in feed and manure. In addition, FARMFLOW reproduces the balancing feedbacks in the system, consisting of export of elements in milk, crops and animals as well as losses from the soil system. Field data from a case study at the Swedish experimental farm Öjebyn were used to parameterise FARMFLOW. At Öjebyn, organic and conventional management practices were run in parallel over 12 years. Simulations were made over 6 crop rotations, i.e. 36 years. The simulated effects of the management practices on the stocks, flows and balances of P, Zn and Cd are presented for the organic farming system (OFS) and for the conventional farming system (CFS) at Öjebyn. The simulated field-specific element accumulation rates were mainly governed by the difference in manure application rate between the fields, arising from differences in field size, and by differences in losses between the fields, arising from differences in the initial elemental content of the fields. Simulations with increasing animal density in the two management systems showed that FARMFLOW is useful in predicting the impact of intensification on, e.g. the manure P application rates. Annual variations in mass flows, e.g. yields, and element concentrations were observed at the Öjebyn farm. In order to evaluate the impact on the element stocks, flows and balances of variations in the system, FARMFLOW was parameterised with input data including these variations. For P, the variations generated moderate variations in the farm gate P balances, and the variations did not cause a shift in the balances around the policy target of zero. For Zn, it was shown that even the highest simulated accumulation rates of the two systems did not cause soil Zn concentrations risking adverse effects on soil fertility in a 36 year perspective. For Cd, the variations can lead to up to 20% higher removal of Cd in harvested crops than average values.