The tree species preferences of koalas (Phascolarctos cinereus) inhabiting forest and woodland communities on Quaternary deposits in the Port Stephens area, New South Wales
Stephen PhillipsA, John CallaghanB and Valerie ThompsonC
ASchool of Resource Science and Management, Southern Cross University, PO Box 157, Lismore, NSW2480, Australia.
BAustralian Koala Foundation, GPO Box 9899, Brisbane, Qld 4001, Australia.
CSan Diego Zoological Society, PO Box 551, San Diego, CA92112-0551, USA.
An assessment of the tree species preferences of koalas inhabiting forest and woodland communities growing on Quaternary deposits in the Port Stephens area, New South Wales, was undertaken between November 1994 and March 1996. Using a plot-based methodology, 3847 trees were sampled, comprising 15 Eucalyptus species and 17 species of non-eucalypt. Evidence of tree use by koalas, specifically the presence of koala faecal pellets, was recorded from beneath 10 Eucalyptus species and 9 species of non-eucalypt. Tree species preferences were determined by analyses of log-likelihood ratios derived from data based on the presence/absence of koala faecal pellets, rather than on gross counts. This approach confirmed significant variation in the levels of utilisation amongst and between different tree species, and that two in particular swamp mahogany (E. robusta) and drooping red gum (E. parramattensis) were most preferred. Increases in the levels of use of other tree species were also positively associated with the presence of E. robusta and/or E. parramattensis. Levels of utilisation of E. robusta and E. parramattensis did not alter significantly in response to changes in their respective densities, suggesting that the relative abundance of both was important in terms of understanding the carrying capacity of vegetation communities utilised by koalas. The results have established the success with which an enumerative approach to the interpretation of faecal pellet data can be utilised to clarify the tree species preferences of koalas. Application of the approach for habitat assessment and mapping purposes is also discussed.