Abstract. Given the limited scope of unaltered, protected areas in most regions, understanding the contributions to imperiled species conservation by landscapes in which habitat elements have been reconfigured is critical. Commercial forestry has been a driver of altered structure and composition of forests and of distribution and character of aquatic systems. In eastern North America, extensive historical wetland drainage reconfigured hydrologic environments from low-gradient wet flats and isolated wetlands to connected networks of linear ditches. Landscapes where both uplands and aquatic environments differ from historic conditions may affect most ecological aspects of semi-aquatic species, including reptiles. Our objective was to determine if habitat selection and use decisions by spotted turtles (Clemmys guttata), a rare semi-aquatic species, supported population maintenance in a reconfigured forest landscape in eastern North Carolina, USA. We captured 280 individuals and radiomarked 31 adults to examine habitat selection at multiple spatial scales with paired logistic regression, movements and home ranges with location data and utilization distributions (UDs), survival with a known-fate model, and abundance with N-mixture modeling. Across local and landscape scales, turtles selected features associated with ditches despite abundant, more natural aquatic depressions across the study area. Habitat metrics describing understory closure and substrate characteristics were important at local scales, and closed canopy forest and habitat heterogeneity was positively associated with activity areas at landscape-scale spatial grains. Both movements and home ranges were centered on ditches, and turtles exploited ditches to access mates, nest sites, or uplands for estivation. In this highly reconfigured landscape, this species appeared to have sufficient behavioral plasticity to acquire key resources contributing to high survival and an abundant population. Isolation from road traffic and collection, both which negatively affect turtles elsewhere, was facilitated by limited public access. Our results suggest that conservation and management of rare species should not rely solely on habitat information gained from studies in more pristine areas because such results may not demonstrate the range in variation in behavior that might allow persistence in novel environments, absent key threats.