Climate change is predicted to become a major cause of species loss in the coming
century. Shifts in distribution as a response to changing conditions have already been
observed for many terrestrial organisms. A species’ capacity to respond to climate
change will depend greatly on its ability to track suitable conditions; those unable to
track optimum conditions will be under increased threat of local extinction. There is,
therefore, a need to include dispersal parameters in models that forecast the impact of
climate change on species distributions, but this is limited by a paucity of dispersal
data for many species. In this thesis I develop predictive models of dispersal ability to
improve estimates of both distance and rate of dispersal in mammals. Chapter 2
presents a database of empirically derived dispersal distances for mammals and an
analysis of the probability distribution of those distances, aimed at describing the ‘tail’
of the kernel, important in understanding long distance dispersal. Chapter 3 assesses
the explanatory power of species’ life history and ecology, within a phylogenetic
framework, to predict dispersal distances. Chapter 4 examines the roles of dispersal
and colonisation ability in mediating the extent to which a species can fill its potential
environmental niche and quantifies the effects of model accuracy and projection
extent on this approach. Chapter 5 utilises a new technique for identifying patterns of
geographic and phylogenetic constraint to examine the dual roles of evolutionary
history and environment in determining a species’ ability to fill its potential
environmental niche. This thesis helps to clarify controls on range limits and to
incorporate such controls into species distribution models. By providing more
accurate predictions of the impacts of climate change on species range size and
location, this work helps us to better understand the threat to species diversity from
global change.