Analysis of multivariate abundances

Four ARC Projects are funded for 2012-2018 on theoretical and applied aspects of the analysis of multivariate abundance data.

Multivariate abundance data are abundances collected simultaneously for many taxa (species, orders, functional groups...). This type of data is commonly collected in ecology and the environmental sciences, and has been collected and analysed in thousands of publications. There are many possibilities for significant contributions to this field, using a more rigorous model-based approach to analysis.

Current projects:

  • Design based inference for mixed models of multivariate data (Loic Thibaut).
  • Developing the mvabund package for model-based analysis of multivariate abundance data (Loic Thibaut, Alice Wang).
  • Fast methods for fitting latent variable models to multivariate data (with Francis Hui, Sara Taskinen, and John Ormerod).
  • Model-based methods for vegetation classification (Mitchell Lyons with David Keith, Jane Elith, Stuart Phinn).
  • The PIT-trap for residual resampling of non-normal data (with Alice Wang).
  • Modelling species interaction (Gordana Popovic).

Recent projects:

  • Model-based approaches to multivariate analysis, in the mvabund package (Alice Wang, Ulrike Naumann, Stephen Wright)
  • These methods resolve some undesirable power properties seen in algorithmic approaches to analysis (Alice Wang, Stephen Wright)
  • Incorporating species traits in analyses to explain interspecific variation in environmental response, the "fourth corner problem" (Alex Brown, with Bill Shipley and Trevor Hastie)
  • Finite mixture modelling to cluster species based on their environmental response (Francis Hui, with Scott Foster and Piers Dunstan).

Contact us

Phone: +61 2 9385-7031
Fax: +61 2 9385-7123
E-mail: David.Warton(at)

Red Centre room 2052
(David's office)
Red Centre room 6103
(Eco-Stats office)

School of Mathematics and Statistics
The University of New South Wales

Share |