The agricultural experiment is usually reductionistic, that is, one variable is investigated while all others are kept constant. However, in the real world, many factors can vary and interact simultaneously. Increasing oilseed rape yield is often dependent on a few variables (e.g. cultivar and nitrogen), and can be investigated using a traditional reductionist approach. However, enhancing biological control of oilseed rape pests is a much more complicated task, involving many interacting factors, such as plant species, cultivar, structure, soil tillage, parasitoids, pesticide use and so on. Hence, it is not enough to understand the behaviour of individual components. The performance of the system must also be studied to be able to develop new strategies addressing environmental or sustainability questions. The study of the whole farming system has been a main focus within economy and sociology for a long time, but has gained importance in biology only during the last decade or two. Tools to study and quantify the performance of systems have been developed within areas of product control and certification of production processes, and have been adopted by experimental biology. The farming system experiment often aims at continuous improvement, usingthe approach of objectives - design - evaluation - changing design in a repeated yearly process. The tools used for evaluating system performance depend on the scale of the study (field, farm, region, nation) and the problem focus (production efficiencyin relation to resources, environmental questions, rural development and so on).
展开▼
