Some general principles of crop rotation experiments

This review paper deals with some of the general statistical principles of crop rotation experiments. The basic terminology of the subject is defined. Attention is paid to the fundamental design principle that each phase of a rotation should each year be present in an experiment involving that rotation. The inclusion of ‘auxiliary’ treatments - usually fertilizer treatments - is discussed. Consideration is given to matters influencing the choice of size and shape of plots. The statistician's role in planning a rotation experiment is portrayed as onerous. Emphasis is put on the importance of care and good practice in the management of a long-term agricultural experiment. Complications arising in the statistical analysis of data from a rotation trial are sketched.