Process-explicit approaches to predicting and managing range changes in Australian marsupials

A species' range‚ÄövÑvÆthe geographical area over which it is found‚ÄövÑvÆis determined by many interacting factors. Outside of small-island settings, the broadest determinant is typically climate, which interacts with biotic factors (predation, competition, pathogens, etc.), human influences including land use, and other abiotic factors (geology, soil chemistry, etc.). Global warming will, therefore, cause many species to either change their range or change how the factors shaping their range interact. Paradoxically, it is projected that to prevent the worst outcomes associated with the resultant geographical redistribution of the world's species due to climate change, we might have to actively and pre-emptively redistribute species ourselves. These movements of species for conservation purposes have been termed conservation translocations. To understand natural range changes and plan conservation translocations, we need accurate predictions of suitable conditions and habitat‚ÄövÑvÆa task that has become a key applied goal of biogeography and ecology. Better predictions will enable more effective and efficient management of endangered species, pests, and pathogens. Currently, the most popular method for implementing these predictions is the Species Distribution Model (SDM). This broad group of techniques can be split into two strategies: correlative (pattern oriented) and mechanistic (process-explicit). Each have their associated advantages and disadvantages relating to their underlying assumptions and ease of use. In this thesis, I examined the geographic ranges of three Australian marsupials in the past, present, and future; identify translocation sites for the contemporary cases; and sought to distinguish the characteristics of successful translocations globally. Specifically, I aimed to : (i) undertake a systematic review to identify the features of successful and failed conservation translocations, (ii) assess the role of past climate change on the mainland of Australia in the extirpation of the Tasmanian devil (Sarcophilus harrisii), (iii) identify current translocation sites for the vulnerable brush-tailed rock-wallaby (Petrogale pencillata), and (iv) quantify the effect of future climate change on Australia's only hibernating marsupial, the mountain pygmy possum (Burramys parvus), and evaluate potential sites for its translocation. I found that translocation success is positively affected by investment‚ÄövÑvÆgreater number of individuals over longer periods‚ÄövÑvÆthough historical context is important‚ÄövÑvÆOceania's difficulty in controlling invasive species is the likely reason it had less positive outcomes than Europe or North America. I showed that there is no clear evidence, using process-explicit SDMs, for invoking climate change as the sole cause of the loss of the Tasmanian devil from mainland Australia in the Holocene, and consequently other factors, such as human intensification and dingoes, must be involved. I demonstrate that parts of Tasmania seem to be an ideal and immediate climate-habitat refuge for the highly endangered southern population of brush-tailed rock-wallabies and potentially also the critically endangered mountain pygmy possum in the future. I show the importance of food availability for the mountain pygmy possum if it is to cope with a rapidly changing alpine climate. Although our ability to predict species' distributional changes has improved in recent decades, the fact that ranges are an emergent property of a variety of processes means that many challenges (scientific and conservation management) are still to come. However, process-explicit SDMs and ecologically relevant predictors in correlative SDMs provide viable platforms to meet these challenges, as demonstrated in this thesis.