There is much debate about the conservation status of the koala (Phascolarctos cinereus) on a statewide and national scale. Abundance estimates vary greatly, particularly in Victoria and Queensland. My research helps to clarify this situation by deriving an abundance estimate of 63 000 ± 18 000 koalas in the mulgalands of southwest Queensland, a semi-arid rangeland which covers an area of around 20 million hectares. This thesis also provides an insight into the factors that determine koala distribution and density in semi-arid environments.
The broadscale survey design and sampling strategy developed to survey a large heterogeneous region is the initial focus of the thesis. The development of such a sampling strategy is framed by two primary considerations. Firstly, the low density koala populations of the region, arid secondly, the broadscale nature of the research.
Conventional survey methods, such as direct counts, are often unsuitable for surveying low-density animal populations, particularly on a broadscale. This thesis details the development of the first faecal pellet sampling methodology to estimate the abundance of an arboreal mammal. This methodology was applied within the framework of a multi-scaled stratified survey design developed to capture factors that determine the distribution and density of koalas at different scales.
The pellet sampling methodology is based on the quick and accurate identification of trees which have koala faecal pellets beneath them. Pellet abundance was then estimated by conducting counts in quadrats distributed proportionally to canopy area. These data, along with mean daily pellet production and pellet age data were converted to estimates of koala abundance using the faecal standing crop method.
Distribution and abundance data were collected from 149 sites located throughout the study area. These data indicate that the eastern portion of the study area has the most uniform distribution and highest densities of koalas. Koalas were also recorded in the central and western regions of the study area.
Koala distribution within the study area appears to be largely determined by water availability, while the physiography and edaphic characteristics of specific landform types appear to determine the level of utilisation within these boundaries. The highest density populations occur in riverine communities. Low density populations are spread throughout dissected residual communities in the central and eastern portion of the study area, and are also scattered throughout floodplain and plains communities, particularly in the east. To determine the variance around the koala abundance estimate a bootstrap/Monte Carlo exercise was conducted. Riverine and dissected residual koala populations comprise >90% of this estimate. The relative importance of residual communities as habitat is critical to the future management of the koalas, which were considered to be largely restricted to riverine communities.
The highly variable environment of the mulgalands and land management practices associated with grazing have created an inherently dynamic environment with many seral plant communities of varying stability. The region has a long history of vegetation clearance and fragmentation, a change in species dominance from perennial grasses to trees and shrubs and a reduced fire frequency. I also formulated region specific management recommendations that account for the detrimental impacts of land management practices on koala conservation.
For too long the koala conservation debate has been guided by anecdotal evidence and 'best guesses' of their abundance, particularly for inland populations. Regardless of whether koala abundance is considered in light of conservative or more generous estimates, this research shows that the mulgalands contain a significant proportion of Australia's koala population. These populations appear to be relatively stable and are under less immediate threat than many coastal populations, particularly in New South Wales and Queensland. Future koala conservation initiatives should include the vast regions of semi-arid Queensland that comprise a substantial proportion of its current range.