Applied Vegetation Science 17 (2014) 236–245 SPECIAL FEATURE: ECOLOGICAL RESTORATION Topsoil removal improves various restoration treatments of a Mediterranean steppe (La Crau, southeast France) Renaud Jaunatre, Elise Buisson & Thierry Dutoit Keywords Community composition; Grassland; Hay transfer; Nurse species seeding; Rehabilitation of former agricultural land; Restoration of grazing; Soil transfer; Species richness; Topsoil removal Nomenclature Base de Donn ees Nomenclaturales de la Flore de France (Tela-Botanica 2011) Received 25 February 2013 Accepted 29 July 2013 Co-ordinating Editor: Rob Marrs Jaunatre, R. (corresponding author, renaud.jaunatre@yahoo.fr), Buisson, E. (elise.buisson@univ-avignon.fr) & Dutoit, T. (thierry.dutoit@imbe.fr): Universit e d’Avignon et des Pays de Vaucluse, Institut M editerran een de Biodiversit e et d’Ecologie (UMR CNRS/IRD 7263/237), IUT, Site Agroparc BP 61207, 84911, Avignon Cedex 9, France Abstract Question: Can topsoil removal improve restoration efficiency of several ecologi- cal engineering techniques (nurse species seeding, hay transfer and soil transfer) for restoring a Mediterranean steppe, species-rich plant community? Location: La Crau area, southeast France (Provence). Methods: After the rehabilitation of a 357-ha herbaceous, sheep-grazed range- land suitable for threatened steppe birds in a formerly abandoned industrial orchard in the last French Mediterranean steppe (La Crau area), four experi- mental treatments were applied, with or without topsoil removal. Topsoil was removed on a 5000-m² area and the four treatments (control, nurse species seeding, hay transfer and soil transfer) were applied within this area where top- soil was removed and within the area where topsoil was not removed. Three years later, soil chemistry, soil seed bank, vegetation cover and height, plant spe- cies richness, composition and diversity were compared. Results: Removing topsoil partly restored soil conditions and significantly reduced the non-target seed bank. Moreover, species richness and similarity to the reference steppe was significantly increased with topsoil removal. Topsoil removal, combined with soil transfer and hay transfer, created a community that most closely resembled the reference during the study period. Conclusion: In order to restore plant community composition, topsoil removal is a relevant method to increase the efficiency of other ecological restoration techniques, especially those that strengthen dispersal processes. Introduction Restoration ecology is at the interface between theoretical ecology and ecological restoration (Falk et al. 2006). This science, which focuses on how to repair damaged and destroyed ecosystems (Society for Ecological Restoration International Science & Working Policy Group 2004), is increasingly requested since restoration was identified as a tool to slow biodiversity loss (Millennium Ecosystem Assessment 2005; Nellemann et al. 2010; Convention on Biological Diversity 2011). The choice of techniques used to restore ecosystems depends both on ecosystem charac- teristics and on the causes of their destruction, and this choice is based on knowledge provided by ecological theory (Falk et al. 2006). Intensive cultivation is one of the modern land uses responsible for a significant part of the 39–50% of transformed land surface (Vitousek et al. 1997). Intensive cultivation causes important and lasting dam- age to established vegetation. One of the first steps of intensive cultivation is usually deep ploughing, which sup- presses at least the above-ground vegetation. This distur- bance (i.e. a constraint that limits plant biomass by causing its destruction; Grime 1977) releases nutrients in soils and therefore also reduces stress (with stress defined as a growth limitation; Grime 1977). Its effects on vegetation are significant even a very long time after cultivation aban- donment (Dupouey et al. 2002; € Oster et al. 2009). The pre-disturbance plant community is indeed destroyed by ploughing and the seed bank usually does not contain Applied Vegetation Science 236 Doi: 10.1111/avsc.12063 © 2013 International Association for Vegetation Science