Amphibian distributions and extinction risk in China under climate and land use change
Species extinctions are inevitable, irreversible and a time-delayed process that is accelerating due to the warming climate and the loss of habitats worldwide. The most threatened vertebrate group in terrestrial ecosystems is amphibians. In this thesis, I assess potential range shifts and extinction risks of amphibians in China, and test whether the current distribution of protected areas is effective in protecting amphibian habitats both today and under conditions imposed by climate and land
... climate and land use change in the future. This was done by use of circular statistics, metapopulation models, climate velocity algorithms and ensemble species distribution modeling. Overall, I found: (1) large conservation (protected area) gaps were found throughout China and especially the southern parts of Tibet and the Hengduan Mountains, an amphibian diversity hotspot vulnerable to climate change and human activities; (2) correlations between directional range shift of species and climate velocity were evident with range shifts of amphibians in China being mostly tri-directional in pattern, preferring northern, eastern or northeastern directions for different dispersal scenarios and climatic data used; (3) relaxation time of extinction debt for amphibians in China was related to the strength of the Allee effect, forest cover change and the trade-off between colonization and emigration rates. Metapopulation models, with and without Allee effects, estimated average time to half extinction for endemic amphibians of China to be 44.9 and 71.8 years, respectively. Collectively, this thesis research identifies iii regional conservation needs of amphibians, fuels the development and application of novel statistical methods in the estimate of species extinction, and paves the ways for future studies on extinction debt modeling. iv Acknowledgements I am very fortunate to work with Prof. Fangliang He and under his supervision, to conduct research on biodiversity conservation and ecological modeling. Without his excellent mentorship, great support and many encouragements, this thesis would not have been possible. I sincerely thank my supervisory committee members, Profs. Mark Lewis and Scott E. Nielsen, for their helpful advice and constructive comments on my research over the past years. I appreciate Profs. Mark Poesch, Matthew Wheatley and David M. Green for giving helpful comments on my projects as well. My thesis greatly benefited with discussions and assistance from current and past members in Profs. He and Nielsen's labs and my friends, including, but not limited to Drs.