Design and planning of long-term experiments

Abstract

Long-term experiments (LTE) are extremely varied in form, and scientists who undertake this type of experiment face many challenges. Aspects of management, design layout, the role of the coordinator and statistician, and computer use are discussed. The important statistical ideas for LTE are reviewed and related to single perennial crops such as wild blueberries, fruit trees and permanent pastures. Key words: Long-term experiments, planning of experiments, experimental design, REML, residual maximum likelihood