Plant Ecology and Evolution FastTrack: 1–11, 2012 First-year results of a multi-treatment steppe restoration experiment in La Crau (Provence, France) Renaud Jaunatre * , Elise Buisson & Thierry Dutoit Université d’Avignon et des Pays de Vaucluse, Institut Méditerranéen d’Ecologie et de Paléoécologie (UMR CNRS/IRD), IUT, Site agroparc BP 61207, FR-84911 Avignon cedex 09, France *Author for correspondence: renaud.jaunatre@yahoo.fr INTRODUCTION Ecosystems which have undergone long-term severe environ- mental constraints, either biotic (e.g. grazing) or abiotic (e.g. dryness), generally exhibit a high species richness, as well as often unique and highly structured communities (Tilman & Pacala 1993, Alard & Poudevigne 2002, Hopper 2009). Intensive agriculture usually induces strong disturbances which are long lasting and/or at large scale; constrained ec- osystems which have been cultivated often pass biotic and abiotic thresholds (Whisenant 1999). This is especially true when soil has been enriched, leading to increased competi- tion (Marrs 2002), and when target species propagules are not available anymore: either because the seed bank has been de- pleted (Hutchings & Booth 1996) or because their dispersion abilities are too weak (Bakker et al. 1996). Such disturbed ec- osystems cannot be restored without speciic restoration tech- niques (Cramer et al. 2008), which have to focus on lowering non-target species abundance and on improving target spe- cies dispersion (Walker et al. 2004, Baer et al. 2008, Kiehl et al. 2010). Lowering non-target species is achieved either (i) by preventing their emergence: i.e. suppression of seed bank or (ii) by lowering their growth and density: i.e. restoration of low soil nutrient content or suitable disturbance regimes (i.e. grazing, etc.). Among the various techniques available, All rights reserved. © 2012 National Botanic Garden of Belgium and Royal Botanical Society of Belgium ISSN 2032-3921 REGULAR PAPER Background and aims Intense agriculture phases on old plant communities, such as Mediterranean steppes, can lead to low resilience. Two main obstacles to the spontaneous recolonization of these plant communities are often the low dispersal of target species and the high dispersal and establishment potential of unwanted species. The aim of the study is to ind the most eficient restoration treatments to restore these plant communities. Methods – After the rehabilitation of an herbaceous sheep-grazed community on a formerly intensively cultivated orchard in the last French Mediterranean steppe (La Crau, Provence, France), four experimental restoration treatments were applied to restore the steppe plant community: (i) topsoil removal to lower ruderal species seed banks and soil trophic levels, (ii) nurse species seeding to rapidly occupy niches, and then to provide safe sites for target species once sheep grazing is reintroduced, (iii) hay transfer to provide local species seeds, and (iv) soil transfer to provide local species propagules with associated microorganisms. One year later, plant species richness, composition and diversity are compared. Results – Although the communities developing on areas seeded with nurse species and where topsoil was removed differed most widely from the reference ecosystem, i.e. steppe, these restoration treatments succeeded in achieving their goal by signiicantly lowering the abundance of unwanted dominant species. While hay transfer did not have a signiicantly higher species richness than that of the rehabilitated area, it showed promising results, as some germinations of target species were observed. One year only after the treatment was applied, soil transfer provided a community richness and composition very close to that of the reference ecosystem, but not with the same vegetation structure. Conclusion – In order to restore plant community composition, the more the treatment strengthens community dispersal, the more eficient it is. The gain in eficiency is closely linked with the cost of the treatment. Keywords – Plant community composition, former agricultural land, grassland, hay transfer, nurse species seeding, soil transfer, species richness, topsoil removal.