Plant diversity dynamics in an eutrophic alluvial reed bed after experimental small-scale disturbances

In an eutrophic alluvial reed bed in western France, artificial disturbance s (mowing) and natural disturbances (cattle, roebucks, boars, voles...) are very frequent. In order to mime these disturbances, experimental small-sca le disturbances of various intensities were generated in April 1996 after t he end of the winter flooding. The two first intensities of disturbances (I 1 and I2) were created by herbicide contact (Round-up (R)) associated (I2) or not (I1) with litter removal. Intensity 3 (I3) corresponded to the mecha nic removal of the top of the soil. In intensity 4 (I4), seed bank was remo ved by exporting the top 10 cm of the soil. The size of disturbances (1 m(2 )) was chosen to mime natural disturbances. Subsequent changes in cover, sp ecies composition and some diversity indexes (species richness, species div ersity, spatial heterogeneity, evenness, dominance) were observed over one following vegetation cycle from May 1996 to August 1996 and compared to und isturbed vegetation. After four months of recolonisation, disturbances remo ved significantly species composition in disturbed areas. Plant recolonisat ion and plant diversity dynamics depended on the intensity of disturbances. The results showed the role of disturbances in plant diversity dynamics an d in particular in the maintenance of biological diversity at the scale of the marsh. In the context of a gradient of disturbances intensities, the re sults confirmed the framework of the intermediate disturbance hypothesis (C onnell 1978, Grime 1979, Huston 1979) which shows that species richness or diversity is maximum at an intermediate intensity of disturbances.