Species-specific erosion of genetic diversity in grassland butterflies depends on landscape land cover

Changes in land cover, particularly agricultural intensification, are a primary cause of pollinator decline. Decline and isolation can reduce genetic diversity, reducing adaptive potential and unmasking genetic load. Maintenance of functional connectivity in changing landscapes is necessary to ensure persistence, but how land cover impacts genetic diversity and functional connectivity in butterflies remains poorly understood. We assess to which extent land cover functionally isolates grassland butterfly populations and what consequences isolation has for genetic diversity using whole genome and land cover data. We assess species-specific patterns across three butterfly species that vary in habitat specialization and mobility - Polyommatus icarus, Plebejus argus and Cyaniris semiargus - using samples from 6 to 11 populations across a ~25,000 km2 area in southern Sweden. We find that while generalist Po. icarus is nearly panmictic and heathland specialist Pl. argus maintains modest functional connectivity, grassland specialist C. semiargus exists in largely isolated populations. Genetic diversity in all species is positively related to grassland extent in the surrounding landscape. Arable and forest extent are positively and negatively related to genetic diversity in Po. icarus, while the inverse is true in C. semiargus. Lower functional connectivity in C. semiargus is coupled with higher rates of inbreeding, suggesting more urgent need of conservation action aimed at restoring functional connectivity in this species, illustrating how genetic data can reveal threats that may go unnoticed with abundance data alone. We conclude that specialist butterfly genetic diversity relies upon preservation of well-connected semi-natural grasslands to maintain functional connectivity at longer distances.