The Butterfly Project's Data-Driven Approach: from Plant-Pollinator Species Distribution Modelling to Optimizing Local Conservation of Pollination Services

Within the Horizon Europe project 'Butterfly', multiple work packages collaborate to predict species distributions using plant and pollinator observation data. With the rise of citizen science and novel pollinator identification technologies this data pool will exponentially expand. A critical challenge remains: how can stakeholders translate this vast amount of data into actionable policy and informed local landscape planning? We address this by developing tailored tools which use the advanced species distribution modelling frameworks developed within the Butterfly project and other EU projects like VALOR. We aim to advise on local mitigations measures fitting within a wider European conservation strategy; hereto we adhere the principle of end-userand environment-specific guidance. For instance, in the Hesbaye fruit region of Flanders, a cherry farmer benefits primarily from spring-emerging pollinators like Osmia cornuta or Andrena haemorrhoa. Conversely, an adjacent Natura 2000 habitat requires a conservation focus on other more specialized and less prevalent pollinator species. To aggregate and analyse this information unbiasedly across all pollination services, we utilize multiple complementary data layers. First, we match crop blooming phenology with modelled pollinator phenology and regional distributions to identify which specific pollinators are present near a crop of interest. Second, we integrate joint plant-pollinator distribution models with the EuroAPPA plant- pollinator interaction database to extract region-specific plant recommendations (currently targeted at the NUTS 3 level). Crucially, the environmental suitability of mitigation measures is driven by joint distribution modelling; this ensures ecological viability, preventing mismatches such as placing habitat- specific wildflowers in intensive agricultural zones or managed flower strips within protected nature reserves. Beyond informing plant selection, we also develop tools that can assist policymakers by mapping regional pollination needs and identifying overlaps between current pollinator distributions and Red List species. This presentation provides an inside look at the development of these Butterfly applications, discussing methodological choices and potential shortcomings to stimulate discussion on future improvements and alternative approaches.