Factors controlling vegetation change in long-term experiments designed to restore heathland in Breckland,UK

Abstract. This paper describes the vegetation change in an 18-year experiment designed to test a range of control of Pteridium aquilinum (Pteridium) and heathland restoration treatments; cutting, asulam application and Calluna vulgar is seeding, in a range of combinations at Cavenham Heath in Breckland, UK between 1978 and 1996. Vegetation change was recorded in two Phases; in Phase I (1978-1986) species biomass was sampled and in Phase II (1986-1996) cover was measured. Initially, Calluna establishment was good in some treatments, especially where Calluna seed was added and Pteridium was controlled. Other plots developed either a grass-heath flora dominated by Agrostis capillaris, Deschampsia flexuosa, Dicranum scoparium, Festuca ovina and Rumex acetosella or were dominated by clonal species such as Calamagrostis epigejos or Carex arenaria. An unconstrained ordination showed significant vegetation change through time and that several treatments influenced the vegetation, especially those involving asulam application. When variation partitioning with constrained ordination was used a different explanation emerged. In Phase I the most important factors were the management treatments applied, elapsed time and spatial factors, with little overlap. In Phase II, elapsed time became irrelevant because the variation time explained overlapped that which could be explained by other variables. The most important of these were management treatments, spatial effects, weather, the amount of bare ground caused by disturbance and Pteridium litter cover (an index of Pteridium recovery). The implications of these results in interpreting vegetation change are discussed.