Trajectories to extinction: spatial dynamics of the contraction of geographical ranges

Aim We examined the range contraction of 309 declining species of animals a nd plants to determine if the contraction dynamics better matched predictio ns based on the demographic characteristics of historical populations (demo graphic hypothesis) or based on the contagion-like spread of extinction for ces (contagion hypothesis). Location Species included in the analysis came from all biogeographic regio ns. Methods We obtained range maps for 309 species from literature or through p ersonal correspondence with authorities. Hypotheses were contrasted by exam ining the sequence of changes in the proportion (C) of the remnant range th at fell within the central region of the historical range. Monte Carlo simu lations and polynomial regressions were employed to examine changes in C du ring the process of range contraction. Results The results of the Monte Carlo simulations indicated that more spec ies had observed range contractions consistent with the contagion hypothesi s than expected by chance (z-score = 2.922, P = 0.002). The Monte Carlo ana lysis also indicated that the number of species whose observed range contra ctions were consistent with the demographic hypothesis was no greater than expected by chance (z-score = 0.337, P = 0.367). The results of the polynom ial regression analysis for the two most common taxonomic groups (mammals a nd birds) and for all geographical regions (Australia, Africa, Eurasia, and North America) we examined also supported the contagion hypothesis. Main conclusions Most of the examined range contractions are consistent wit h the contagion hypothesis and that the most likely contagion is human rela ted disturbance. These results have important implications for the conserva tion of endangered species.