Relating land-use and plant biodiversity in Scanian semi-natural grazing lands

anglais. Climate change and biodiversity loss are amongst the most pressing environmental issues internationally, with far-reaching impacts that place natural and semi-natural habitats at ever greater risk of degradation. My project explores the effects of land-use on plant biodiversity in semi-natural grazing lands using datasets covering Sweden’s southernmost region, Scania. Comparative analyses were performed using biodiversity measures and a number of landscape variables. The measures of biodiversity in terms of plant species richness were calculated using a grazing land inventory subset with records from the latest years. These measures were correlated with landscape variables, relating to hydrology and vegetation phenology. Such variables include a Wetness Index and Plant Phenology Index, computed using recent remotely sensed data, namely Sentinel-2 time series data. Land cover data was processed to cluster the study sites into distinct land cover groups which facilitated further correlation analysis between plant biodiversity and the landscape variables. These variables were also analysed at two spatial scales, i.e. at the extent of both the grazing lands and their 1 km buffer zones.

anglais. Grazing land has been described as a biodiversity-rich habitat, and modifying agricultural practices which sustain both fodder production for grazing livestock and the grazing activity itself could disturb the ecosystems in such semi-natural landscapes. Livestock grazing is extensive in Scania and other Swedish regions, and significant changes to this practice could negatively impact biodiversity. The relevance of land-use within semi-natural habitats that have an agricultural purpose is important to study, given the risk that such grazing lands are lost to cropland or other land-use whenever grazing practices are abandoned. The potential for biodiversity loss in these rural settings is known to be significant with increasing land-use change. Finding out if any landscape variables can be associated with the plant biodiversity found in these sites is useful for revealing which, if any, land-use types and/or agricultural practices apart from the grazing itself can be associated to grassland biodiversity. My project explores the effects of land-use on plant biodiversity in semi-natural grazing lands using datasets covering Sweden’s southernmost region, Scania. Comparative analyses were performed using biodiversity measures and a number of landscape variables. The measures of biodiversity in terms of plant species richness were calculated using a grazing land inventory subset with records from the latest years. These measures were correlated with landscape variables, relating to hydrology and vegetation phenology. Such variables include a Wetness Index and Plant Phenology Index, computed using recent remotely sensed data, namely Sentinel-2 satellite data. Land cover data was processed to cluster the study sites into distinct land cover groups which facilitated further correlation analysis between plant biodiversity and the landscape variables. These variables were also analysed at two spatial scales, i.e. at the extent of both the grazing lands and their 1 km buffer zones. Following the use of standard biodiversity metrics and multiple landscape descriptors, the importance of present land-use for plant biodiversity patterns in Scanian grazing lands could not be established. Whilst avoiding unnecessary complexity in my attempt to relate land-use with biodiversity, the results and implications are in agreement with recent literature suggesting that rather than the present land-use, the grassland plant biodiversity is due to historical land-use to a greater extent. Historical land-use remains a valid explanation of present-day plant richness in grazing lands as most land-use effects cannot be associated with biodiversity.