Degree Name

Doctor of Philosophy


Faculty of Science


Tree hollows are semi enclosed cavities that naturally form in mature trees. Hollows are a distinctive component of uneven aged, ‘old growth’ forest. Within Australia they are used by at least 15% of vertebrate fauna, with a large proportion of these species being obligate hollow users: i.e. such species are dependent on the occurrence of suitable densities of hollows. Hollow formation in Australia sclerophyll forests is very slow, taking in excess of 100 years for small hollows to form. Thus if hollows are lost, new hollows may take many decades to form. As a result, logging in either timber production forests or conservation reserves where logging has previously occurred may deplete hollows. Fires may also cause both the loss and creation of hollows. Combinations of logging and fire in the flammable sclerophyll forests of Australia therefore have the potential to lead to differing densities of hollows and hollow-dependent fauna.

While there have been investigations of effects of logging on hollows as well as separate studies of the effect of individual fires on hollows, studies that investigate the interactive effects of logging and repeated fires on the production and loss of hollows are lacking. Logging tends to target trees suitable for recruitment of hollows, thus losses of trees with hollows from fire (as a consequence of branch shear or tree collapse) within a logged forest may result in a net loss of hollows. Such a scenario may depend on the frequency of fires and the consequent balance between rates of loss of hollows and hollow creation. Knowledge of the effect of different fire regimes on arboreal marsupials is, however, limited. Previous work is largely anecdotal and suggests that high severity fires may have a significant impact on populations of arboreal marsupials.

The aim of this study was to investigate the effect of different fire regimes and logging on forest stand structure, tree hollows and populations of arboreal marsupials in sclerophyll forests of Northeastern New South Wales. Specifically I investigated:

1. The effect of fire frequency and logging intensity on forest stand structure and density of tree hollows in wet and dry sclerophyll forest;

2. The effects of tree size (diameter at breast height, DBH), forest type and fire frequency on the occurrence of hollows within unlogged wet sclerophyll forest (WSF) and dry sclerophyll forest (DSF);

3. The effect of individual fires on tree basal injury, tree collapse and consumption of logs in sclerophyll forests;

4. The combined effects of fire frequency and logging intensity on populations of Petauroides volans (Greater Glider) and Petaurus australis (Yellow-bellied Glider) in WSF and DSF.

Research questions 1, 2 and 4 were addressed through a large scale ‘natural experiment’ in north-eastern New South Wales. The study area allowed for a ‘disturbance matrix’ to be established, with sites distributed across varying degrees of logging intensity (i.e. volume of merchantable timber removed off the site) and varying wildfire frequency (i.e. number of wildfires since 1980). On each site I measured forest stand attributes including DBH of all trees and the number of hollowbearing trees to determine the effect of past disturbance on the current forest stand structure and density of hollow-bearing trees. At each site, a 500m long transect for spotlighting of arboreal animals was also established. Sampling was done on three occasions.

Elsewhere on the New South Wales North Coast, I undertook fieldwork before and after four spatially separated prescribed fires to determine the resulting tree basal injury, log consumption and tree collapse.

Logging significantly reduced the average stand DBH, density of hollows and hollow bearing trees in both WSF and DSF, while fire frequency significantly increased the likelihood of basal injury. The abundance of hollow-bearing trees and hollows significantly decreased as a consequence of increasing fire frequency and logging intensity; however on unlogged sites the abundance of hollow-bearing trees and hollows increased as a consequence of increasing fire frequency.

The incidence of hollows in eight species of Eucalyptus, along with stags (i.e. dead trees), was positively correlated with DBH in unlogged stands. In five Eucalyptus species, fire frequency was also found to have a positive effect on a tree containing hollows. Tree basal injury was therefore primarily determined by DBH and fire frequency.

Prescribed fires caused no tree collapse within three months after fire but very low levels occurred after a wildfire. The level of log consumption resulting from both types of fire was high but comparable with other studies. As a result there was disparity between the level of tree collapse and log consumption, indicating that frequent fire could have a detrimental effect on the occurrence of logs suitable as wildlife habitat.

The density of tree hollows was positively correlated with the density of P. volans. Fire frequency was found to be negatively correlated with the population density of P. volans in WSF and the probability of occurrence of P. australis in DSF. Logging intensity was also negatively correlated with the population density of P. volans, however no effect of logging was found on P. australis.

The findings of this study suggest that logging in interaction with fire has major effects on the supply of hollows and hollow-bearing trees. Fires have the potential to reduce the size at which trees first develop hollows and increase the density of hollow bearing trees in unlogged forest, however the suitability of these hollows for hollowdependent fauna is currently unknown. I found that increased fire frequency may increase the abundance of hollows; however an increase in fire frequency results in a negative impact on the abundance of P. volans and P. australis and logging also has a negative impact on populations of P. volans. The study indicates that there are major implications of logging and fire regimes for the conservation of a broad range of both vertebrate and invertebrate fauna in the extensive sclerophyll forests of south eastern NSW. The problem of a lack of hollow recruitment in logged forest subjected to relatively frequent fire could be addressed by retaining additional hollow recruitment trees within timber production forest. These results suggest that populations of arboreal marsupials may be negatively affected by more frequent and intense fires as a consequence of climate change. Further research is required on the effect of fire on arboreal marsupials, in particular the effect of varying fire severities on mortality and habitat attributes. This will assist with the identification of appropriate fire regimes for the maintenance of viable populations of these unique fauna.