TS-CRC Student project - Spatial patterns of distribution, abundance and diversity in the vertebrate fauna assemblages of the Desert Uplands bioregion, northern Queensland

James Cook University: Completed

Inland forest bat

The inland forest bat, Vespadelus baverstocki, one of the smallest mammal species in Australia, if not the world. Weighing between 3-5 grams, and found in the south-western parts of the Desert Uplands. They can consume 1-1.5 times their body weight in insects per night—and therefore play an essential regulatory role in the ecosystem. An animal to be celebrated not persecuted.

Crucifix or holy cross toad

The Crucifix or Holy-cross toad. A sandy soil dweller that can live for years deep underground enveloped in a cocoon of its own making, waiting for very heavy rains to signal time to breed. Adults are much larger and less brightly coloured, and they all extrude a thick toxic latex when handled, essential predator avoidance strategy when shining like a glowing McDonalds' sign.

Velvet gecko

Oedura castelnaui, the northern velvet gecko. One of the variety of three allied velvet geckos in the Desert Uplands. It hides under heavy bark of tree such as gidgee and black gidgee during the day, coming out at night to feed on a variety of insects and spiders.

Julia Creek Dunnart

It's not easy being endangered: a Julia Creek Dunnart, Sminthopsis douglasi, expresses its absolute fury at being fiddled with by a well-meaning zoologist. Restricted to the cracking grey clays typical of the Mitchell Grass Downs, its discovery in outlying Astrebla grasslands in the Desert Uplands was a bonus and extended its known range.

Alex Kutt

Commenced 1996

Background | Fauna Survey | Unexpected finds and new species | Feral cat diet | Application of the research | Future directions |


The Desert Uplands (DEU) is one of Queensland's six tropical savanna bioregions, covering more than 6 million hectares and sharing boundaries with the Mitchell Grass Downs to the west, the Brigalow Belt to the south and east, and the Einasleigh Uplands to the north. The DEU has a semi-arid climate with vegetation consisting of predominantly acacia and eucalypt woodlands, ephemeral lake habitats and grasslands. It straddles the Great Divide between Charters Towers, Hughenden and Blackall and it is this situation between the wet east coast and the dry interior which makes this area of biological and biogeographic interest.

Unlike the coastal zone and Cape York Peninsula, the vertebrate fauna of Queensland's tropical savannas are almost entirely unknown and unsurveyed, quite inexplicable given the mounting need to find an effective balance between the twin land-management goals of economic viability and nature conservation. The project contributes to the TS-CRC research project Vertebrate Biogeography and to the Desert Uplands Management Study.

Fauna Survey

The survey was undertaken in the Desert Uplands between 1997 and 2000, and was predominantly designed to identify fauna assemblages of the regional ecosystems (the lowest level bioregional planning unit sensu Sattler and Williams 1999), to describe the patterns of the variation in distribution, diversity and abundance of these assemblages, and to characterise the region's biogeographic position within the Queensland landscape.

Sampling in the survey used a standardised nested quadrat array as the basic trapping unit, modified from the strategy developed by Dr John Woinarski for bioregional surveys undertaken by Parks & Wildlife Commission of the NT. This approach uses a set combination of Elliott, cage and pitfall traps with timed day and night searches and bird counts. Incidental fauna records were also collected and a thorough search of all other reliable secondary data sources was made and incorporated into the final database.

Three years later after 23,000 Elliott and cage trap nights, 4200 pitfall trap nights, 600 individual 20 litre buckets dug, more than 7km of drift fence installed and the equivalent of almost seven weeks at 24 hours-a-day of active searching, the field survey was completed. The simple but primary outcome was a database of more than 35,000 records (24,000 from field survey), representing 433 species and translating to 11,300 unique species localities.

Unexpected finds and new species

As one would hope for a predominantly unsurveyed bioregion, there were a number of unexpected finds and range extensions including Spinifexbird and Painted finches, rodents and dasyurids such as the Lakeland Downs mouse Leggadina lakedownensis , Desert Mouse Pseudomys desertor and Pebble-mound mouse Pseudomys patius , Common Dunnart Sminthopsis macroura and the Julia Creek Dunnart Sminthopsis douglasi , and reptiles like the Brigalow Scaly-foot Paradelma orientalis and the Centralian Blue-tongue lizard Tiliqua multifasciata.

And the highlight (satisfying a zoologist's unspoken of desire for immortality) with the discovery of two new species, both reptiles: Ctenotus terrarossa sp. nov, (the etymology identifying the deep red sandy habitat in which it was found and is seemingly restricted to), currently being described in conjunction with the Queensland Museum; and Lerista sp. nov., still awaiting a formal classification.

Feral cat diet

As an adjunct to the trapping survey, the diet of feral cats were examined via their stomach contents. The cats were collected systematically across the bioregion and the directly adjacent areas of the Mitchell Grass Downs and Northern Brigalow Belt.

Samples were obtained by local professional kangaroo and pig shooters, and those participating were furnished with a collection kit containing preservatives and sample containers, labels for recording locality and morphometric data for the cat. A small bounty was paid to the shooters to facilitate and maintain their interest in the project.

Gut content samples obtained were then washed, sorted and identified to the lowest taxonomic level possible. The Queensland Museum verified gut material that could not be identified confidently and intact, good-quality specimens were incorporated into their collections. A total of 194 cat guts were collected over a two-year period, comprising 1300 prey items. While this number is large, it is likely that it is far less prey than these cats actually consume.

Of all prey items identified, 16 per cent were birds, 33 per cent reptiles, 5 per cent amphibians, 25 per cent mammals, and 21 per cent insects. Conversely the proportion of taxonomic groups represented on a per cat basis equates to 33 per cent containing bird prey items, 63 per cent reptiles, 9 per cent, amphibians, 47 per cent mammals and 42 per cent insects.

Because feral cats are not an economic pest, they are generally ignored as a target for research - except where re-introduction or management of rare species is in conflict. The lack of systematic Australia-wide examination of regional and local patterns and impacts is worrying, as the influence of feral cats on native fauna, particularly in recently modified habitats, is not well understood. These data will hopefully contribute to the debate regarding the effects of cat predation.

Application of the research

While there is obvious value in information where none previously existed, this survey will also provide a spectrum of outputs including:

  • examination of the recent and later biogeographic patterns of the Desert Uplands fauna and distribution (it seems the bioregion is both barrier and refuge for the Torresian fauna of the coast and the Eyrean fauna of the inland);
  • the environmental determinants controlling the local and regional patterns of fauna assemblages, including the influence of spatial scale (vertebrate fauna can be characterised into discrete assemblages based on abundance at local sites, species interactions and environmental characteristics, though on larger scales the species composition and structure changes according to area of habitat and geographic location in the regional landscape);
  • how adequately landscape units act as a surrogate for identifying vertebrate fauna assemblages (apart from the species' abundance and distribution, vertebrate fauna assemblages can be characterised on multiple overlapping levels, be it vegetation, habitat or landscape or management units. However the adequacy of these partitions vary enormously in their ability to act as a surrogate for capturing and describing the fauna diversity across the bioregion);
  • on a smaller scale the influence of fire and grazing on fauna within a particular vegetation association (in spinifex-dominated woodlands, fire is an important influence on fauna assemblage, due to changes in vegetation composition and structure, though grazing complicates pattern of structural change to the disadvantage of some species); and finally
  • using the current system of protected areas created using regional ecosystems as the planning unit, test how well the fauna diversity is represented, what additional or alternative reserve designs would be needed to adequately capture the majority of species, and what species or groups act most suitable surrogate for the entire species pool.

Future directions

This inventory is merely a snapshot of the vertebrate fauna in the bioregion and barely provides a baseline from which to extend our understanding of the dynamics of wildlife in a rapidly modifying landscape. One key future goal is passing the information back to the Desert Uplands community, via an atlas of distribution or a field guide to describe the variety and ecology of the species in the area. The data will also be used in Draft Desert Uplands Conservation Strategy (EPA)

Additional targeted follow up work is also essential, such as directly examining in detail the interplay between to the variety of land management regimes (fire frequency, stocking rates, tree clearing) and how best to balance needs for farm viability with native fauna protection.


Dr Alex Kutt
CSIRO Sustainable Ecosystems
Tel: 07 4753 8547

Fax: 07 4753 8600

Davies Laboratory, Private Mail Bag PO