Predicting the potential distribution of an endangered cryptic subterranean mammal from few occurrence records

Knowledge of geographical distributions and habitat preferences are central to the conservation and management of threatened species. Ecological niche models can be used to map potentially suitable habitat, making them valuable tools in conservation biology. These models can assist in searching for new populations of species with poorly known distributions and can also be used to guide area selection for systematic biodiversity planning. These models have traditionally required ten or more occurrence records for calibration and this has limited the use of these tools in conservation biology, since many threatened species are known from fewer records. However, recently models have been successfully calibrated using few records. To illustrate this, we developed an ecological niche model to map the potential distribution of the endangered Juliana's golden mole (Neamblysomus julianae) in South Africa so that it could be used in searching for unrecorded populations and for area selection in systematic biodiversity plans. A model was calibrated in Maxent using only four occurrence records. Predictor variables included various bioclimatic, soil and vegetation types. The model identified limited suitable habitat within the map region. The first model facilitated the identification of two previously unrecorded populations. A second model was calibrated using the two additional occurrence records. This increased the proportion of correctly predicted presence records. Jackknife analyses indicated that the models were successful at predicting known presences as suitable. This paper demonstrates the use of ecological niche modelling to conservation of a cryptic endangered species with a poorly known distribution and discusses issues that are relevant for the application of this approach to other species.