Conserving and restoring wildlife in fragmented urban landscapes: A case study from Brisbane, Australia

Garden, Jenni (2007). Conserving and restoring wildlife in fragmented urban landscapes: A case study from Brisbane, Australia PhD Thesis, School of Geography, Planning and Architecture, University of Queensland.

Attached Files (Some files may be inaccessible until you login with your UQ eSpace credentials)
Name Description MIMEType Size Downloads
Read with bookreader  n0_full_garden.pdf Full text application/pdf 15.26MB 880
Author Garden, Jenni
Thesis Title Conserving and restoring wildlife in fragmented urban landscapes: A case study from Brisbane, Australia
School, Centre or Institute School of Geography, Planning and Architecture
Institution University of Queensland
Publication date 2007
Thesis type PhD Thesis
Open Access Status File (Publisher version)
Supervisor Clive McAlpine
Total pages 248
Collection year 2007
Language eng
Subjects 0502 Environmental Science and Management
Formatted abstract
The environmental conditions that make a location suitable for urban development often coincide with those that support high species diversity and endemism. The resulting loss, fragmentation, and degradation of natural habitats have significant ramifications for urban wildlife. Native wildlife populations fragmented by urban development undergo population declines and localised extinctions often long after the development occurs. Native biodiversity is therefore under threat as urban areas continue to expand and replace natural habitats, yet the processes enabling wildlife to persist in urban areas are not well understood. Consequently, urban planning and management decisions often fail to ensure the long-term conservation of urban biodiversity. This project applied a spatially-explicit, multi-scaled landscape approach to determine the relative importance of site, patch, and landscape-level attributes for the occurrence of reptile and small mammal species living in fragmented forest remnants of Brisbane City, Queensland, Australia.

The study tested a set of a priori models to investigate the importance of site-level habitat attributes relative to patch size and shape, and landscape composition and configuration for native reptile and small mammal species. Field based fauna and habitat surveys were conducted at 59 sites, with fauna surveys repeated for spring/summer over two consecutive years. Fauna surveys used a combination of live trapping, direct observation, and trace analysis to increase the detection probability of the range of target species. The field surveys provided information about reptile and small mammal species occurrences, and the local-level habitat structure and composition within each site. Comparative analysis was used to investigate the detection success and costs associated with the different survey methods employed. Cluster analysis and multi-dimensional scaling ordination were used to investigate relationships between species occurrences and local-level habitat characteristics. Generalised linear modelling and hierarchical partitioning were used to determine the importance of the area of forest habitat and its configuration relative to patch size and shape, and local vegetation composition and structure.

A total of 19 reptile and nine mammal species were identified. All survey methods made a contribution to overall detection success by detecting at least one species not identified by any other method. Pit-fall traps and direct observations were the most successful and cost efficient method combination for detecting reptile species. In contrast, mammals were most successfully and efficiently detected using a combination of hair funnels and Elliott traps (for small bodied mammals), or cage traps (for medium sized mammals), with the one exception being the more successful use of pit-fall traps for detecting planigales (Planigale maculata).

At the local-level, I found that species composition for both taxa was influenced most by habitat structure rather than vegetation composition. Reptile species composition was correlated with: the amount of fallen woody material, the presence of termite mounds, soil compaction, and the weediness of sites. Mammal species assemblages were most correlated with the presence of grass trees and soil compaction. When the importance of local-level habitat attributes was examined relative to patch and landscape-level attributes, I found that attributes across each spatial level were important for determining species richness. Overall, patch level attributes, such as size and shape, were less important than landscape context and local attributes of habitat quality, such as habitat structural complexity. Reptile species responded to attributes at both the local and landscape-levels, with the area of forest habitat and its configuration in the surrounding landscape, and soil compaction and weed cover at the local-level, being correlated with species richness. In comparison, mammal species responded to attributes at all three spatial levels. The key factors influencing mammal species were the amount of forest and rural/low density urban habitat at the landscape-level, habitat composition at the local-level, and patch size and shape at the patch-level.

The research outcomes highlight the need to adopt a multi-species, multi-scaled approach to research and urban conservation planning and management. The major outcomes of this project are synthesised into a set of guidelines and a decision-support tree that will enable urban decision-makers to target priority habitat and landscape attributes for the conservation of native reptile and small mammal communities in urban landscapes.
Keyword Wildlife conservation -- Queensland -- Brisbane
Wildlife management -- Queensland -- Brisbane
Additional Notes The University of Queensland acknowledges that the copyright owner of a thesis is its author, not the University. The University has made best endeavours to obtain author permissions to include theses in this collection, however we have been unable to trace and contact all authors. If you are the author of a thesis included in this collection and we have been unable to contact you, please email

Version Filter Type
Citation counts: Google Scholar Search Google Scholar
Created: Fri, 21 Nov 2008, 15:34:04 EST