Unexplained limits on species distributions: how important is climate in defining species’ range edges?

The study of species’ range margins has a long history of academic interest, but is of particular relevance today due to its applications in modelling species range shifts induced by climate change, and predicting the spread of invasive species. Climate has long been assumed to structure species’ ranges over broad scales, but this consensus has recently been challenged by work on non-climatic factors, such as dispersal, biotic interactions and gene flow. It remains unclear how and where non-climatic factors can structure species’ ranges, and to what extent species’ ranges will consistently match sets of climatic conditions. In this thesis I investigate what can lead to a species underfilling its climatic niche (when a species fails to colonise all climatically suitable areas), or expanding its climatic niche (when a species is able to colonise new types of climate). I find evidence that several non-climatic factors can slow or prevent non-native species colonising all climatically suitable areas in their naturalised regions, including dispersal, fragmentation of climatically suitable areas and the area of introduction. I also find that species will readily spread into new precipitation regimes with which they have not been previously associated. This suggests that species ranges can be constrained by non-climatic factors in the wettest part of their native range, and these constraints are frequently lifted in their naturalised range. I find evidence that species range limits set by temperature, in particular temperature maxima and minima, are more conserved and species will rarely expand into new thermal regimes. I also find evidence that species have different phenotypic responses to temperature across their range. Together these results indicate that a species’ current range frequently does not indicate its overall climatic tolerance, particularly in relation to precipitation, hence predictions that rely on associations between occurrence and environmental variables will frequently be flawed. Future work should consider a systematic way of detecting and including non-climatic factors that constrain the edges of species’ ranges