To identify management treatments suitable for the conservation of extensively managed grasslands, the ‘Fallow experiments in Baden‐Württemberg’ were set up in 1975. In this investigation, species composition of the grazing, mowing, mulching, controlled burning and unmanaged (succession) treatments were analysed after 25 yr of continuous management in Arrhenatherum elatius and Bromus erectus grasslands. Through ordination analyses it was found that species composition is strongly dependent on the management treatment. The first axis, identified by ordination analysis, essentially corresponded to a gradient of decreasing disturbance frequency. Controlled burning resulted in a unique species composition. Grazing, mowing and mulching twice a year were found to be most suitable for the conservation of unimproved, species‐rich grasslands.

Nomenclature: Oberdorfer (1994).