Output details

Effects of species and habitat positional errors on the performance and interpretation of species distribution models

Significance of output:

Species distribution modelling is now routinely used in biodiversity conservation, the management of invasive species, and for forecasting the effects of climate change on species ranges. It is commonly assumed that the species and environmental data driving such models are error-free, although this is almost impossible to achieve. By artificially inducing error in both species and habitat data, this study shows the impact of imprecise locational data on the models produced. Importantly, it makes the novel observation that models are protected against locational error by spatial autocorrelation, inspiring the subsequent paper by Babak et al. (2011), J. Biogeog. 38, 1497-1509.