Modelling the spatial distribution of species-rich farmland to identify priority areas for conservation actions

Spatial targeting of conservation actions is indispensable to protect and maintain farmland biodiversity and associated ecosystem services. To implement region-specific conservation actions, information on the spatial distribution of farmland biodiversity and its major drivers is a prerequisite. Here, we model the spatial distribution of species-rich farmland (defined by a minimum number of indicator plant species) and landscape elements (e.g. hedgerows, ditches, patches of woodland or scrub) at the national level by combining field-based survey data from a biological monitoring programme with agri-environmental variables. Adopting factor analysis and mixed effect models, we show that the percentage cover of species-rich farmland and landscape elements can be predicted by a combination of factors describing the intensity of arable farming, topography, low-input grassland farming, landscape structure and the intensity of livestock farming. Our results demonstrate that the relative importance of these factors differed between species-rich farmland and landscape elements. The share of species-rich farmland was highest in upland and structurally complex grassland-dominated regions where extensive livestock production is practised; whereas landscape elements were most likely to occur within flat regions where edge density tends to be high. olicy context, our generated maps contribute to identifying priority regions for conservation actions. It is proposed that in regions with high coverage of species-rich farmland priority should be given to sustaining low-input grassland farming by keeping farmers in business and preventing farmland abandonment. We conclude that this modelling approach could be adapted for application in other European countries to facilitate spatial targeting of conservation actions.