AbstractAimWild bees still face striking shortfalls in knowledge of biodiversity in key regions of the world. This includes Europe, where despite a long tradition of data gathering, the continental scale distribution patterns of wild bees have not been systematically analysed to date. This study aims to characterise large‐scale biodiversity patterns to: (i) understand spatial–temporal heterogeneity in large‐scale databases, (ii) locate genuine diversity hotspots and their relationship with biogeographical patterns or habitats of interests and (iii) identify understudied species and areas to further design conservation actions for most at risk species in key regions.LocationEurope.TaxonBees.MethodsWe present a continental and standardised study of bee taxonomic and phylogenetic diversity patterns in Europe, using a large compilation of occurrence records of nearly three million validated occurrence records for 1515 wild bee species.ResultsSouthern and eastern Europe suffer from the largest gaps in data availability while northern and western regions benefit from better historical coverage. Our models show that higher wild bee diversity in Europe is hosted in xeric, warm areas, as highlighted by a clear latitudinal gradient. However, phylogenetic diversity is predicted to be more homogenous across Europe than taxonomic diversity, suggesting that policies and strategies targeted to protect species richness may differ from those targeting greater phylogenetic diversity.Main conclusionsThis study represents a significant advance in the characterisation of wild bee distribution patterns across Europe and is an important stepping stone towards the design of more targeted survey efforts and conservation actions of this key group of pollinators. This, in turn, will provide the data necessary to improve the spatiotemporal coverage in a context of ongoing and future Europe‐wide monitoring schemes, to ultimately develop cost‐effective, coordinated and evidence‐based conservation actions and tailored habitat management actions that can be implemented on a smaller scale.
Read full abstract