On scale problems in modelling: an example from soil ecology.

Agricultural research aims to increase our understanding of systems and to reveal manipulable characteristics at the farm scale, so that management practices can be improved. Because agricultural systems are too complex to understand as a whole, subsystems (e.g. soil, plant, microclimate) and subprocesses (e.g. transport processes, photosynthesis, energy balance) are distinguished (Figure 19) and studied (systems analysis) with the ultimate objective of interconnecting the resulting knowledge and of returning to the farm scale (systems synthesis). Subsystems and subprocesses usually operate on much smaller spatial and temporal scales than the total system: they represent a different level of organization than the farm level. For example, crops in the Netherlands usually have a fairly homogeneous appearance on a hectare scale, indicating the integrative power of the root-soil system with respect to space and the homogenizing effect of soil tillage and fertilization. For the crop growth process the characteristic time in which a substantial change in amount of biomass occurs, i.e. the time coefficient, is of the order of a week. Soil oxygen concentration, however, which may have a pronounced influence on crop growth (Drew, 1983), is known to vary substantially at a