Capturing uncatalogued distribution records to improve conservation assessments of data-deficient species: a case study using the glossy grass skink

Effective conservation planning is often predicated on detailed and current information about a species' geographical distribution. However, traditional sources of occurrence data (e.g., online biodiversity databases) may be insufficient for estimating the range of rare, poorly understood species that are readily misidentified. Here, we demonstrate a more holistic approach to this problem, using the poorly known glossy grass skink (Pseudemoia rawlinsoni) as a case study. We first compared the relative contribution made (to our geographical knowledge of the species) by online database records, with that of photo-substantiated records obtained via personal communication (PC). We used ecological niche modelling (ENM) to predict the species' distribution, then performed field surveys at both historical and predicted suitable sites to further clarify its occurrence. 20% of all known records came from the PC method, which resulted in 35 new sites and increased the species' area of occupancy (AOO) by 176 km². Most records obtained via PC came from the past decade, demonstrating that this method is more effective at elucidating the current distribution. ENM revealed that P. rawlinsoni has a disjunct range, and is mostly a low-elevation coastal species, with the exception of suitable habitat in parts of the high-elevation Australian Alps bioregion. The species' AOO has likely declined over recent decades owing to anthropogenic disturbance, given that 38% of the species' predicted range is now cleared agricultural land, and our field surveys failed to detect the species at 52% of historical record sites. Together, these findings provide a robust foundation of geographical knowledge on which to develop strategic conservation actions for the species.