Background The tiny island of Stora Karlsö (just 2.5 km²) is located in the Baltic Sea, 6km from its larger neighbour, the island of Gotland. Originally a fossilised coral reef surrounded by limestone strata, Stora Karlsö is dominated by a plateau that ends abruptly at its northern tip, forming dramatic cliffs which crash down into the sea. Over the centuries the limestone pavements have been greatly weathered and are now dissected by deep crevices. Where the limestone is reduced to a thin layer of gravel, a very special type of habitat has formed: the Nordic alvar. Meanwhile, on the plains and in the hollows on the plateau, there are species-rich grasslands formed by centuries of sheep grazing. In short, the island is a paradise for plants: at the end of May and early June the whole place is covered in thousands of flowering orchids and other rare species. This unique conservation value was recognised as far back as the 1880s when the island became a nature reserve. Unfortunately, in keeping with the conservation philosophy of the time, this resulted in all the sheep being removed to allow the vegetation to develop 'freely'. In addition, some exotic plant species were introduced. This misguided gesture resulted in a strong overgrowth which very badly affected the different habitat types that are so characteristic of the island. Juniper took over most of the island and many of the flora species that are dependent on open grasslands drastically reduced their range. Objectives The main objective was to recreate an open landscape on the plains and plateau of Stora Karlsö Island. The main task was therefore to clear invading scrub and trees, predominantly Juniperus communis, from 200 ha (85%) of grassland and alvar habitats, in order to restore the island's vegetation back to the state it was in during the middle of the 20th century. In addition, exotic plants would be removed, notably the introduced Prunus mahaleb.After restoration, the coverage of trees and shrubs was to be only 10% on 150 ha of the project area, 10-50% on 30% of the project area, and over 50% on 20% of the project area. The long-term maintenance of the project results would be secured through mowing and re-opening the land for sheep grazing, within the framework of a new management plan, and using the EU's agri-environmental subsidies wherever possible. The island also attracts 10,000-15,000 visitors every year. Measures had already been taken to reduce the risk of damage from tourism, for instance by limiting the number of visitors per day to 300, but there was still a need to increase public awareness of the uniqueness of the island. The LIFE project would thus co-finance guided tours, an exhibition and dissemination of information on the beneficiary's web-site. Results Stora Karlsö Island has never been permanently inhabited (except for lighthouse staff), but was used for grazing by sheep at least since Neolithic times. Its unique conservation values were recognised as far back as the 1880s when the island became one of the first protected nature areas in Sweden. In keeping with the conservation philosophy of the time, i.e. to leave the area without any impact of human activities, all the sheep were removed to allow the vegetation to develop ‘freely’. In addition, some exotic plant species were introduced. This resulted in a strong overgrowth which very badly affected the different habitat types. Juniper was recorded on Stora Karlsö only in 1903, i.e. 16 years after grazing ceased in 1887, but by the 1960s it started to expand over most of the island. Against this background, the main project actions were to clear 200 ha of grassland and alvar habitats, i.e. 85 % of the land surface of the island, and to arrange for starting up the traditional land use, i.e. grazing by sheep, with the support of agri-environmental agreements. 79 % (158 ha) of the target of 200 ha of calcareous grasslands and alvar habitats to be cleared had been restored by the end of the project. As of 2004, 96 % (235 ha) of the island qualified for agri-environmental agreements for grazing by sheep, the traditional land use on the island. The principal habitats which benefited from the project were Juniperus communis formations on calcareous heaths or grasslands (5130), rupicolous calcareous grasslands (*6110), semi-natural grasslands on calcareous substrates (*6210), Nordic alvar (*6280) and limestone pavements (*8240). Monitoring of the project actions (clearing and grazing) indicated a positive response by the vegetation and vascular plant communities already 2-3 years after restoration, although this is still too early for more definitive conclusions. The final cost per hectare for clearing was significantly more expensive than foreseen at project start. The beneficiary analysed the reasons for the increased costs in some detail, establishing a reference for the working time and manpower required to restore calcareous grasslands and alvar habitats on very thin soil and where a substantial part of the work has to be done manually. This averaged to 0.057 ha/person/day, but with a wide variation depending on the kind of terrain and on experience. It was also found that collecting and burning cleared material took much more time than previously assessed - in fact, 40% of the total work-time. So the project gained useful experience for the future restoration and management of alvar and calcareous grasslands habitats, with reference to the costs in terms of manpower and consumable materials, and finding that previous estimates might have been much too low. The project involved the control of two exotic plant species, Prunus mahaleb and Acer pseudoplatanus. P. mahaleb was spread over large areas of the island, but once removed its regrowth was found to be quite easy to keep under control. There was a management discussion whether to keep some stands of these scrubs, as they were considered to be an element linked to the cultural heritage of the island. A. pseudoplatanus is more restricted on the island, but much more aggressive and difficult to keep under control. The project site is highly frequented by visitors during the summer period, around 13,000 persons per year, many of them are "naturalists". The majority participate in guided tours. The project grasped the opportunity to promote Natura 2000 and the LIFE Nature financial instrument in general, through information boards and a small permanent exhibition at the visitors' centre on the island. The restoration work involved the seasonal hiring of temporarily-employed staff, many of them unemployed, during late summer and autumn (August-November). Over the whole project period around 100 persons were involved, with a maximum of 60 persons in 2004. Total declared work-time amounted to 3758 work-days or equivalent to around 15 full-time years, i.e. an input not to be neglected in a region with constant problems of unemployment.