Study design In the Reijerscamp Natuurmonumenten established a large field experiment in 2006. There were four experimental areas (Area A-D). In each area large top-soil excavations were made down to the minaral soil (top-soil removal (rem) & untreated controls (cont)). In each of the excavated areas treatments were performed where 1) sods were added (sod), 2) where mixed soil was spead out (soil), and 3) a layer of hay was added (hay), and there was a control where nothing further had been done (cont) all three additions were collected from well-developped seminatural sites (sods: Doorwerthse Heide, Wolfheze, Soil: Dennenkamp, Planken Wambuis, Hay: Buurse Zand, Twente) Outside the excavations the soil, hay & control treatmen ts were also performed just as inside the top-soil excavations. The sod addition treatment was not performed outside the top soil excavations (perhaps for financial reasons?) . Hence in total there were 4 experimental areas, with 7 treatments each, creating 28 sampling locations. Within each sampling location a transect of 5 plots was laid out, with the centre plot in the middle of the sampling location. The orientation of the transect was decided randomly as far a the size of the treatment in each sampling location permitted (see tab 'transect orientation' in the primary data files). Each plot was 1x1m and the spacing between plot edges was 5 m (hence each transect from the far edges of plot 1 & 5 spans 25 m). All data in the primary data files were sampled in this way.