Statistics

Banding a Little Penguin on Phillip Island. [Photo courtesy of Phillip Island Nature Park]	A Little Penguin chick on Phillip Island. [Photo courtesy of Phillip Island Nature Park]

Current Research Projects

Frailty Modelling (Dr Joanne Chapman with Prof. Robin Henderson)

Random errors and unmeasurable association within clusters cannot be assumed Normally distributed and ignored in survival analysis as in other areas. The frailty term is introduced to represent these unknowns. A typical assumption is that the frailty variable follows a gamma distribution. Models which extend existing models and allow for the possibility of negative correlation within clusters (as often seen in animal litters) are being developed.

Statistical Ecology

In order to ensure the survival of animal populations it is essential to be able to estimate mortality rates in the wild. Two ways of doing this are from the recovery of bands from animals found dead, and from mark-recapture experiments. Models involving survival, recapture and recovery probabilities as unknown parameters, are developed and fitted.

The following projects are currently being undertaken in this area.

•  Effect of banding on juvenile Little Penguins Eudyptula minor (Dr Leesa Sidhu & Assoc. Prof. Ted Catchpole)

This work is being conducted in collaboration with researchers from the Phillip Island Nature Park. For more than 50 years, researchers have been marking penguins with flipper bands. Studies of Adélie Penguins and King Penguins have shown that banded penguins have lower survival rates than unbanded birds, and that they use more energy than unbanded birds when swimming. Our earlier work in this area was the first to study the effect of banding on Little Penguins. We showed that banding had a detrimental effect on the survival of adult penguins, with banded birds having an annual survival probability 6% lower than unbanded birds. While the effect of banding on juvenile Little Penguins is currently unknown, it is likely that banding significantly reduces the survival of young birds, particularly in their first year of life, and that our existing estimates of first-year survival underestimate the true survival probability for unbanded birds. A study examining the effects of banding on juvenile Little Penguins is currently underway on Phillip Island. The main aim of this project is to analyse the data from the Phillip Island study, in order to obtain estimates for survival for banded and unbanded juvenile Little Penguins, and for band loss in penguin chicks. The Department of the Environment and Heritage will use the results of this project, together with those from our earlier work, to determine whether banding of Little Penguins will be allowed to continue in Australia.

•  Analysis of recovery/recapture data for Pacific Gulls Larus pacificus (Dr Leesa Sidhu & Assoc. Prof. Ted Catchpole)

This work is being conducted in collaboration with researchers from La Trobe University. The Pacific Gull is the only large gull occurring naturally in Australia. There is evidence that its population size is falling, its range contracting, and that it could become extinct. While earlier studies of Pacific Gulls have focused on their biology, there are no existing survival estimates for these birds. This study will be the first to produce age- and time-varying survival estimates for these birds, by analysing a long-term mark-recapture-recovery dataset. Such a study is crucial to improve our understanding of these birds, and to ensure the survival of this species.

•  Analysis of recovery/recapture data for Short-tailed Shearwaters Puffinus tenuirostris (Dr Leesa Sidhu & Assoc. Prof. Ted Catchpole)

This work is being conducted in collaboration with researchers from La Trobe University. Short-tailed Shearwaters have been studied continuously on Fisher Island, Tasmania since 1946, making it one of the longest continuous studies of any wildlife population in the world. Although life-history data have been collected sporadically over this time period, a detailed mark-recapture-recovery analysis has not yet been conducted. This study will produce age- and time-varying survival estimates for Short-tailed Shearwaters, and examine the effect of individual covariates such as egg size on first year survival.

•  Comparing first year survival for Little Penguins Eudyptula minor in Phillip Island, Australia and Oamaru, New Zealand (Dr Leesa Sidhu & Assoc. Prof. Ted Catchpole)

This work is being conducted in collaboration with researchers from the Phillip Island Nature Park , the Department of Conservation, New Zealand and the Institute of Marine Research, Norway. While the Phillip Island study has been underway for almost 40 years, the New Zealand study consists of six years of data. Here we estimate and compare the survival of penguins in these two locations and determine to what extent the conclusions of covariate dependence of survival can be extended from Phillip Island to another penguin colony.

The effects of prescribed burning (Dr Joanne Chapman with Dr Karen King)

In many landscapes an important fire management objective is to reduce the negative impacts from unplanned fires on people, property and ecological values. In Australia there exists an inherent assumption that high spatial variability in fire ages and hence fuel loads will have negative effects on both the incidence and spread of subsequent fires, and will enhance ecological values. Our research using the process based computer simulation model, FIRESCAPE-SWTAS, predicts a number of relationships between prescribed burn treatment levels and spatial patterning and management objectives in south west Tasmania, Australia. Further research additionally explores the effects of prescribed burning treatment unit size on unplanned fire incidence and area burned both in the general landscape and specifically in fire-intolerant vegetation.

Other current projects:

Modelling frost tolerance of Eucalyptus trees (Assoc. Prof. Ted Catchpole)

Using computer algebra software to investigate parameter redundancy (Assoc. Prof. Ted Catchpole)

Members of the Statistics Research Group

Academic Staff:

Dr Joanne Chapman (j.chapman@adfa.edu.au) BSc PhD Lancaster
Dr Leesa Sidhu (l.sidhu@adfa.edu.au) BSc (Hons), DipEd, MEdSt Qld, PhD UNSW

Visiting Fellows:

Assoc. Prof. Ted Catchpole (e.catchpole@adfa.edu.au) BSc Lond., MSc Kent, PhD Dund.
Dr Wendy Anderson (w.anderson@adfa.edu.au) BSc Lond., MSc Dund., PhD UNSW

Research Students:

Jennifer Hollis - Research Topic - Coarse woody fuel availability and consumption in Australian forest fires

Recent Graduate Students:

Dr Leesa Sidhu – Research Topic - Analysis of recovery-recapture data for little penguins

Collaborators:

Assoc. Prof. H. Sidhu (PEMS, UNSW@ADFA)
Dr L. Chambers (Bureau of Meteorology Research Centre, Melbourne)
Dr D. Cole (University of Kent, England)
Prof. T. Coulson (Imperial College, London)
Dr M. Cruz (Ensis, Yarralumla, ACT)
Dr P. Dann (Phillip Island Nature Park)
Dr R. Day (University of Melbourne)
Dr P. Ellis (Ensis, Yarralumla, ACT)
Mr J. Gould (Ensis, Yarralumla, ACT)
Prof. R. Henderson (University of Newcastle , Newcastle -upon Tyne, UK)
Mr D. Houston (Department of Conservation, New Zealand)
Dr E. Johannesen (Institute of Marine Research, Norway)
Dr K. King, Postdoctoral Fellow (The Australian National University and Bushfire CRC)
Dr S. Matthews (Ensis, Yarralumla, ACT)
Dr C. Meathrel (La Trobe University Wodonga)
Dr L. McCaw (Department of Environment and Conservation, WA)
Prof. B. J. T. Morgan (University of Kent England)
Mr B. Robertson (La Trobe University, Wodonga)