Peter Wilson

Research Fellow

Telephone: +61 (02) 9850 8404
Facsimile: +61 (02) 9850 8245

Email:peter.wilson[at]mq.edu.au

Building E8A 276

Department of Biological Sciences

Faculty of Science

Macquarie University

North Ryde, NSW, Australia, 2109

Background

My academic and work experience background is long and checkered. I began studying mechanical engineering at the University of New South Wales in 1972. Many jobs later, and after bits and pieces of leave of absence and part-time study, I ended up graduating with a Bachelor of Science (Double Major in Zooolgy) in 1983. I began a PhD at UNSW in 1984 but withdrew in 1986. Work experience includes mattress assembler, storeman and packer, engineering workshop gopher, clerk in a major telecommunications company, research assistant on an applied physics project studying bat echolocation, work as programmer/analyst in two state government departments, and middle and senior management roles at the New South Wales National Parks and Wildlife Service. Finally, I went crazy and started an MSc (Hons) at Macquarie University under the supervision of Prof. Emeritus Andy Beattie (MSc (Hons) awarded 2003) and then began a PhD with Andy (submitted 2008).

Isolated kurrajong in paddock

Kurrajongs in cleared landscape

Predicted distribution for Scoteanax rueppellii

Research Interests

I have always been interested in the physics and mechanics of organisms, and in the "big picture" patterns of distribution and abundance. In particular, I am fascinated by the thermodynamics of organisms and communities of organisms, by ecological morphology, and by aspects of biogeography and species distribution.

My abiding biological passion is the ecology and evolution of bats. I have been enthralled by these organsims for almost 40 years, and I am particularly interested in echolocation, insectivorous bat ecology, biogeography and comparative functional morphology within this group of mammals. My long-term passion for this group of organisms combines the diverse, but not entirely unrelated, threads of my research interests.  

Current Research

My current research focus is refining the application of species distribution modelling (SDM) to better understand patterns and processes of plant invasions. In particular, the research group I am part of (lead by Dr Michelle Leishman and Prof. Lesley Hughes) is looking at the implications of climate change on the distribution of invasive plant species. I am looking at the potential for ensemble modelling methods (e.g. Bayesian model averaging) to combine the results of multiple models to give more precise indications of species distribution under various climate change senarios. I am also investigating the impact of verious sources of error on SDM. Recent work (e.g. Graham et al. 2008. Journal of Applied Ecology 45:239-247) has shown the impact of spatial location error on the performance of a suite of important SDM tools. However, there are several sources of error (and potential interactions between them) that need to be understood so that we may have greater confidence in the results of SDM applied to invasive plants and climate change.

Another active research area is the development of methodologies for predictor subset selection for two of the most powerful SDM tools - boosted regression trees and maxiumum entropy prediction (MaxEnt). Like linear models in traditional statstical modelling, highly correlated predictors can lead to instabilities and low quality models for prediction using these two methods. Because they are relatively new methods in terms of their application to SDM, there are a number of issues to be considered when selecting the best subset of predictors.

Previous Research

MSc project: Kurrajong trees (Brachychiton populneus) are found throughout the woodlands and open forests of the drier parts of south-eastern Australia. They exist as scattered individuals but have also been extensively planted as stock shade and drought fodder for sheep and cattle. During an fauna survey in mid-western New South Wales, I became interested in the impact that radical change in the vegetation surrounding a kurrajong might have on leaf and seed eating insects and in turn the impact they might have the trees. This became the subject of my MSc project. I found that isolated kurrajongs in paddocks suffered higher levels of leaf damage and seed predation than those in relatively undisturbed woodland remnants. However, the dieback syndrome commonly seen in isolated eucalypt species was not observed in kurrajongs even though levels of leaf herbivore damage were similar in the kurrajongs I observed compared to published ecualypt data. I concluded that kurrajongs are unlikely to disappear from paddocks like eucalypts and that the dieback paradigm is not universally applicable to all landscape elements in the woodlands and open forests of south-eastern Australia.

PhD project: I dealt with aspects of a key macroecological pattern - the abundance-occupancy relationship. I concluded that much of the research on the nature of this relationship was mis-guided because it sees the relationship as a uniquely biological phenomenon, and ignores the influence of bias and distortions caused by combining incompatible measures of distribution and abundance. I argued that the relationship is in fact a simple physical relationship present in all spatially distributed entities - animate and inanimate (i.e. a universal natural phenomenon). In addition, I argued that simple applications of measure theory from mathematics (particularly spatial point pattern analysis) lead inevitably to the conclusion that we cannot infer causes for the relationship in biological systems from the distribution of points on an abundance-occupancy scatterplot. Newer approaches to causal analysis (e.g. structural equation modelling, Bayesian belief networks) may be able to help make the link between pattern and process in macroecology. However, I suggested that these newer methods should be applied to the abundance-occupancy relationship with caution because characterising the links between field observations of distribution and abundance, and summary measures such as mean abundance and occupancy for assemblages of species presented in scatterplots, as mappings between metric spaces reinforces the fact that vital information about causality is lost in the mapping process. The first of several planned papers arising from this research is in press (see Publications below).

Australian bat species distribution modelling: I have already published two papers on Australian bat biogeography (see Publications below) and have a few more in various stages of development. This is an on-going "background" project that I try to keep ticking along in my spare time.

Selected Publications

Wilson, PD, Downey, PO, Leishman, MR, Gallagher, R, Hughes, L & O’Donnell, J (2009) Weeds in a warmer world: predicting the impact of climate change on Australia’s alien plant species using MaxEnt. Plant Protection Quarterly 24: 84-87.

Wilson, P.D. (2008). The pervasive influence of sampling and methodological artefacts on a macroecological pattern: the abundance-occupancy relationship. Global Ecology & Biogeography. 17: 457-464