Ecological patterns and processes influence ecosystem function at scales from nanometres to global scales depending on the organisms involved. Predicting the presence and abundance of species, at scales appropriate to the organisms and the underlying processes, is central to ecology. Models of species’ distributions can provide important insights into pattern-process-scale relationships including the relative importance of various environmental factors and their interactions that influence habitat selection at the individual and population levels. Mapping current and potential distributions informs the conservation of threatened species by providing spatial information on where a species is likely to occur and the identification of habitat elements and their spatial configurations which influence occupancy and persistence. The aim of this thesis was to incorporate the principles of pattern, process and scale in the identification of habitat associations for threatened species within a species’ distribution modelling framework. Accurate modelling of species’ distributions depends on robust sampling designs, reliable data input and appropriate statistical methodologies that align with the ecological model. I applied a range of innovative statistical methods to various sources of data to identify important habitat associations for a threatened species at different scales and tested the discriminative ability of the resultant models. I integrated the results from extensive field sampling and expert elicitation to build connectivity networks using graph theory algorithms to identify important conservation priorities for threatened species. The threatened brush-tailed rock-wallaby (Petrogale penicillata) was chosen as a suitable study species for quantifying habitat relationships at multiple spatial scales using species’ distribution modelling. The distribution of brush-tailed rock-wallabies is restricted to a set of suitable habitat characteristics related to rocky terrain supporting cliffs and boulder piles that occur infrequently across a landscape. At the site scale, they require suitable resting and refuge sites provided by rocky habitats, while at a landscape scale their dispersal is dependent on the connectivity of suitable habitats. The species is listed as threatened throughout eastern Australia and endangered in some states. Information about its current distribution and occupancy status is essential to support habitat conservation and threat management. The first chapter provides a broad view of the literature on modelling of species’ distributions and the thesis aims and structure. In chapter 2, I assess the ecological scale relevant to habitat modelling for the brush-tailed rock-wallaby. In chapter 3 I test whether habitat models from one region can be extrapolated to neighbouring regions. I use a novel approach and elicitation tool in chapter 4 to collect expert knowledge and assess it with a comprehensive set of field data in a Bayesian framework. In chapter 5 I assess whether landscape connectivity is a determinant of site occupancy by using graph theory algorithms to identify important habitat patches and dispersal pathways for rock-wallaby movement in fragmented landscapes. The final chapter synthesises the individual chapters’ findings within the context of species’ distribution modelling. Management implications are discussed for the conservation of the brush-tailed rock-wallaby and its habitat network. Wider implications are also suggested for other rock-wallaby species and species living in similar environments. The results of the thesis showed the habitat of the brush-tailed rock-wallaby was affected by site-scale and landscape-scale factors, supporting the need for a multi-scale approach when investigating species-environment associations. I found that models performed well within a region at both scales. Extrapolating the models to neighbouring regions resulted in good predictive performance at the site scale but substantially poorer predictive performance at the landscape scale. When there is insufficient field data to build robust data models, management bodies would benefit from incorporating expert knowledge. The study demonstrates the potential errors in using experts with knowledge gained from outside the area of interest. Finally, I highlight the importance of accounting for the landscape connectivity between patches from the perspective of the individual animal. Least cost analysis, using graph theory algorithms, provides a cost-efficient and effective framework for identifying landscape connectivity patterns and key paths and patches to help inform suitable land management strategies for conservation of threatened species. There is much pressure from conservation and management agencies to produce models of species’ distributions that could be potentially be used in other regions or with similar species. The thesis combines ecological theory with rigorous statistical methodology to test different modelling techniques for species distribution modelling. It demonstrates how a combination of expert knowledge, extensive field data and landscape connectivity measures successfully predicts ecological relationships at a number of scales. Species’ distribution models can benefit from applying a robust sampling design and suitable modelling techniques to various data sources to generate ecologically-based information to improve our understanding of species-habitat associations and provide a reliable component to incorporate into conservation planning. This thesis therefore provides important advances to spatial ecology and ecological modelling of species distributions and management of threatened species.

In this thesis a number of statistical methods have been developed and applied to habitat suitability modelling for rare and threatened species. Data available on these species are typically limited. Therefore, developing these models from these data can be problematic and may produce prediction biases. To address these problems there are three aims of this thesis. The _rst aim is to develop and implement frequentist and Bayesian statistical modelling approaches for these types of data. The second aim is develop and implement expert elicitation methods. The third aim is to apply these novel approaches to Australian rare and threatened species case studies with the intention of habitat suitability modelling. The _rst aim is ful_lled by investigating two innovative approaches for habitat suitability modelling and sensitivity analysis of the second approach to priors. The _rst approach is a new multilevel framework developed to model the species distribution at multiple scales and identify excess zeros (absences outside the species range). Applying a statistical modelling approach to the identi_cation of excess zeros has not previously been conducted. The second approach is an extension and application of Bayesian classi_cation trees to modelling the habitat suitability of a threatened species. This is the _rst `real' application of this approach in ecology. Lastly, sensitivity analysis of the priors in Bayesian classi_cation trees are examined for a real case study. Previously, sensitivity analysis of this approach to priors has not been examined. To address the second aim, expert elicitation methods are developed, extended and compared in this thesis. In particular, one elicitation approach is extended from previous research, there is a comparison of three elicitation methods, and one new elicitation approach is proposed. These approaches are illustrated for habitat suitability modelling of a rare species and the opinions of one or two experts are elicited. The _rst approach utilises a simple questionnaire, in which expert opinion is elicited on whether increasing values of a covariate either increases, decreases or does not substantively impact on a response. This approach is extended to express this information as a mixture of three normally distributed prior distributions, which are then combined with available presence/absence data in a logistic regression. This is one of the _rst elicitation approaches within the habitat suitability modelling literature that is appropriate for experts with limited statistical knowledge and can be used to elicit information from single or multiple experts. Three relatively new approaches to eliciting expert knowledge in a form suitable for Bayesian logistic regression are compared, one of which is the questionnaire approach. Included in this comparison of three elicitation methods are a summary of the advantages and disadvantages of these three methods, the results from elicitations and comparison of the prior and posterior distributions. An expert elicitation approach is developed for classi_cation trees, in which the size and structure of the tree is elicited. There have been numerous elicitation approaches proposed for logistic regression, however no approaches have been suggested for classi_cation trees. The last aim of this thesis is addressed in all chapters, since the statistical approaches proposed and extended in this thesis have been applied to real case studies. Two case studies have been examined in this thesis. The _rst is the rare native Australian thistle (Stemmacantha australis), in which the dataset contains a large number of absences distributed over the majority of Queensland, and a small number of presence sites that are only within South-East Queensland. This case study motivated the multilevel modelling framework. The second case study is the threatened Australian brush-tailed rock-wallaby (Petrogale penicillata). The application and sensitivity analysis of Bayesian classi_cation trees, and all expert elicitation approaches investigated in this thesis are applied to this case study. This work has several implications for conservation and management of rare and threatened species. Novel statistical approaches addressing the _rst aim provide extensions to currently existing methods, or propose a new approach, for identi _cation of current and potential habitat. We demonstrate that better model predictions can be achieved using each method, compared to standard techniques. Elicitation approaches addressing the second aim ensure expert knowledge in various forms can be harnessed for habitat modelling, a particular bene_t for rare and threatened species which typically have limited data. Throughout, innovations in statistical methodology are both motivated and illustrated via habitat modelling for two rare and threatened species: the native thistle Stemmacantha australis and the brush-tailed rock wallaby Petrogale penicillata.