Phosphorus dynamics in permanent pastures: differential fertilizing and the animal effect

The general objective of this study was to evaluate the deviation between measured and estimated soil phosphorus (P) content, over 3 years, in a permanent bio-diverse pasture system in Southern Portugal with grazing sheep, and subjected to annual, differential P fertilization. The results can be a contribution to the development of practical and effective site-specific management strategies to minimize pasture yield variation. The soil and pasture samples were taken from a 6 ha field and were geo-referenced with RTK GPS. This procedure was carried out, each year, from March to May, according to the vegetative growth of the pasture. In October of each year the field was fertilized in a differential manner, as a function of soil P concentration measured at the end of the pasture growth cycle (May-June). Maps were developed in ARCGIS 9.3 considering: (1) the soil P concentration and pH; (2) the average P plant uptake; (3) the differential P application; and (4) the extractable P. The significant temporal variability of soil P concentration and pasture dry matter yield confirm the complexity of soil P dynamics in pastures involving two biological systems: plant and animal, under Mediterranean conditions and the consequent difficulties in implementing precision agriculture techniques. The results of this work indicate that 3 years of P variable-rate application rate were not sufficient to obtain homogeneous and adequate levels of P in the whole field for crop production. Despite differential P fertilization with the objective of obtaining homogenous values of soil P content in experimental field, the undulating topography of the experimental field associated with the presence of grazing animals adds a notable short-range spatial variation in nutrients that generally arises from heterogeneous excreta depositions. The small and positive final deviation between measured and estimated P levels suggests the irregular and gradual release of P by the fertilizer over years, contributing to an increase in soil P concentration. P input in fertilizer and removal in the crop greatly exceed all other inputs and outputs. However, the pH effect on extractable P, the estimation of amounts of export/import by livestock, atmospheric deposition or erosion/leaching losses are complex and justify more experiments to evaluate the confidence of long-term estimated P dynamic balance before sustained decision making is possible for differential pasture fertilization and site-specific management strategy implementation.