Strong genetic differentiation of Pedicularis palustris supports local seed sourcing in wetland restoration

• Deep genetic structure found in Pedicularis palustris across Europe. • Genetic similarity drops sharply between populations separated by about 30 km. • RAD-seq reveals genetically distinct populations within regions. • Local seed of P. palustris should always be prioritized for fen restoration. • Avoid seed transfer between the Carpathian and Hercynian regions. Pedicularis palustris , a hemiparasitic plant, can act as an ecosystem engineer by suppressing dominant vegetation and facilitating fen restoration. As fen deterioration continues due to human activities and climate change, conservation interest in this species is increasing. In temperate Europe, where P. palustris is endangered and regionally fragmented, concerns are rising over future commercial cultivation and uncontrolled seed transfer. In this study, we used Restriction Site Associated DNA sequencing (RAD-seq) to assess the genetic structure and differentiation of 18 P. palustris populations across four regions in temperate Europe, and to provide guidance for seed sourcing in restoration practice. In the spatial autocorrelation analysis, populations within approximately 30 km showed the highest genetic similarity, followed by a steep decline up to about 50 km, suggesting limited gene flow beyond this range. Weak but significant positive autocorrelation persisted up to about 240 km, indicating regional-scale lineage cohesion. Bayesian clustering and Neighbor-Net analyses supported the existence of distinct regional gene pools, while also revealing the genetic uniqueness of certain populations. At the broadest scale, genetic affinities did not follow geographic distance, as the Alpine populations clustered with those from the Bohemian Massif, whereas the Western Carpathian populations clustered with those from the Baltic region. Our results demonstrate strong genetic differentiation both within and among regions, and suggest that meaningful provenance of P. palustris extends only over several tens of kilometres. For restoration, we recommend prioritising local seed sources whenever possible. Seed transfers from populations within approximately 30 km can be considered, although further experimental work is needed to confirm this threshold. Where local seed sources are unavailable, seed transfer should remain confined within the same phytogeographical region, with strict avoidance of translocation between the Carpathian and Hercynian units. Uncontrolled translocation or the use of commercial seeds of unknown origin should be avoided.