All that glitters is not gold: genetic and post-mortem monitoring indicate dispersal risks for a semi-aquatic apex predator in a protected marshland

This is the authors' abstract. We don't add key points for this paper.

Protected areas are expected to provide suitable habitats for wildlife even within human-altered landscapes, but environmental cues can be misleading. We evaluated whether a strictly protected, rehabilitated marshland (Kis-Balaton, KB) acts as an ecological trap for the Eurasian otter ( Lutra lutra ), a sentinel apex predator. Using non-invasive genetic sampling within the marshland (90 individuals) and nationwide post-mortem monitoring (462 individuals), we found a spatially heterogeneous population, with a genetically insulated peripheral group. Kinship and group sex ratios indicated kin-structured residents, among which frequent attack-related injuries were consistent with strong territorial behaviour. Net gene flow was directed towards the marshland's core area, mainly via the Zala River. However, immigrants were mainly detected in peripheral areas near settlements and roads, where recovered carcasses showed a higher prevalence of attack-related injuries (KB: 35.7%, other regions: 13.1%), and body condition was poorer in injured than in uninjured otters. These patterns indicate that immigrants rarely integrate into the resident population. The core population remains demographically viable, so the marshland is not a classical ecological trap but a dead-end for dispersers rather than a true refuge. Failure of immigrants to establish, despite apparent habitat connectivity, suggests intrinsic constraints that may be set by the area's ecological carrying capacity. We recommend restoring habitat connectivity, improving legal protection and suitability of small watercourses, and implementing road mitigation and free-ranging dog control to reduce mortality risks. These actions are crucial for gene flow and long-term otter population viability in wetland ecosystems.