Northern Australia has been spared the dramatic loss of mammal species experienced in almost all other parts of the continent. However, there is growing evidence that many mammal species have suffered serious declines in their ranges in northern Australia. An inability to identify the agents of decline has restricted our ability to manage the problem. Changed fire regimes is a potential factor contributing to the declines.

The Kapalga experiment provided an opportunity to examine the influence of four fire regimes on small mammal populations. A total of 96,160 trap-nights over 6 years resulted in over 5000 captures of 12 mammal species.

A distinct difference in the total abundance of mammals can be seen between unburnt and the burnt fire treatments (see Figure below). From a similar starting point, the number of animals steadily declines in annually burnt areas whilst abundance in unburnt changes over time but remains relatively high.

This pattern is reflected in the average number of mammal species. It is apparent from the Figure that a natural cycle of population growth and decline occurs over longer 5-10 year periods in unburnt habitat.

Survival and reproduction are key variables that influence growth in populations, and these were calculated for the common species. Annual survival was highest in Unburnt habitat for three out of four species, Northern Brown Bandicoots (Isoodon macrourus), Northern Quolls (Dasyurus hallucatus) and Grassland Melomys (Melomys burtoni), with Northern Brushtail Possums (Trichosurus vulpecula) the only exception. Similarly, reproduction was highest in Unburnt habitat for two species, Northern Brown Bandicoots and Grassland Melomys.

In general, frequent fires reduce the number of small mammals in tropical savannas. Of all mammal species at Kapalga, Northern Brown Bandicoots showed the most negative response to fire. However, no two species showed an identical response. The frequency of fires is obviously an important aspect of fire management for small mammals, and this should be a priority in any further research through modelling existing data and further field studies.