The importance of long-term experiments in agriculture: their management to ensure continued crop production and soil fertility; the Rothamsted experience.

Long-term field experiments that test a range of treatments and are intended to assess the sustainability of crop production, and thus food security, must be managed actively to identify any treatment that is failing to maintain or increase yields. Once identified, carefully considered changes can be made to the treatment or management, and if they are successful yields will change. If suitable changes cannot be made to an experiment to ensure its continued relevance to sustainable crop production, then it should be stopped. Long-term experiments have many other uses. They provide a field resource and samples for research on plant and soil processes and properties, especially those properties where change occurs slowly and affects soil fertility. Archived samples of all inputs and outputs are an invaluable source of material for future research, and data from current and archived samples can be used to develop models to describe soil and plant processes. Such changes and uses in the Rothamsted experiments are described, and demonstrate that with the appropriate crop, soil and management, acceptable yields can be maintained for many years, with either organic manure or inorganic fertilizers.

Highlights: Long-term experiments demonstrate sustainability and increases in crop yield when managed to optimize soil fertility.Shifting individual response curves into coincidence increases understanding of the factors involved.Changes in inorganic and organic pollutants in archived crop and soil samples are related to inputs over time.Models describing soil processes are developed from current and archived soil data.