Tropical Savannas CRC > Research > Weed Managers > Fire and Rubbervine

Fire in the management of rubbervine-infested riparian communities of northern Australia

Project Leader: Dr Tony Grice, CSIRO Sustainable Ecosystems, Townsville, Queensland

Project 3.3.2

Summary | Application of research | Research Progress | Interactions between fire and grazing | Effect of fire on fauna | Effect of fire on riparian vegetation | Future directions | Publications | Project Team |

towers of rubbervine shroud trees alongside the Burdekin River

Towers of rubbervine shroud trees alongside the Burdekin River, north Queensland
Photo: Kate O'Donnell


Rubber vine ( Cryptostegia grandiflora ) is an exotic woody weed that poses a significant threat to northern Australian savanna landscapes. It has already invaded a large portion of northern and eastern Queensland and bio-climatic analysis suggests that it is capable of growing across northern sections of the Northern Territory and Western Australia. The species poses a major problem for pastoral industries and presents a considerable threat to conservation values. It is particularly prevalent in riparian zones.

Experimental work since the early 1990s has demonstrated that rubber vine is prone to fire. Intense fires kill most seedlings and juvenile plants and a large proportion of adults, as well as causing seed mortality. These results indicate that fire has general potential as a tool for the management of rubber vine. Its specific potential for riparian environments must be examined further. In particular it is necessary to document the impacts of riparian burning on both rubber vine and non-target species and communities

This project addressed three questions:

  1. What are the short- and medium-term effects of fire on rubber vine in riparian habitats?
  2. What are the effects of fire on representative plant and animal taxa of north Queensland riparian environments?
  3. How does rubber vine affect the structure and composition of the native plant and animal communities of riparian environments?

This project will help delineate how invasive species influence landscape health. It will also provide recommendations regarding the use of fire for the large-scale management of rubber vine in riparian landscape zones. Recommendations will be built on an understanding of the responses of rubber vine to different fire regimes and knowledge of the impacts of prescribed burning on non-target plant and animal species.

Application of research

Weed management in extensive savannas present particular challenges. Rubber vine occupies very large areas of northern Queensland and there are few economic management options for controlling it.

This project has been able to demonstrate the application of one technique - fire - for managing rubber vine in the all-important riparian environment. It is especially significant because of the scale of the demonstrations involved. The approach is particularly relevant to Burdekin Catchment where the vegetation is structurally and compositionally similar to that of the study sites. Several landholders in the vicinity of the study area have begun to use fire specifically to control rubber vine.

The research would not have been possible without the collaboration of members of the Seventy Mile Range Landcare Group and the Mount Cooper Bushfire Brigade. Linkages with these groups facilitated effective communication of the work's results to landholders who are confronted by the practical challenges of weed management in the region.

The project will also feed outputs into the Information Clearinghouse, the Graduate Diploma and Master of Tropical Environmental Management, and modules of the extension project. These will be coordinated under the Human Capability Development Theme.

Research findings will also feed into the North Australia Landscape, Ecosystem Management and Landscape Processes themes. This will allow us to produce integrated research findings into an overview of savanna health and how to manage savanna landscapes. The findings from the project will also play a role in the Burdekin Catchment Management Study.

using fire to control rubbervine

The project is assessing the impact of using fire to control rubbervine Photo: CSIRO

Research progress

This experiment was designed to incorporate burning treatments at a scale that is relevant to the practical management of rubber vine. To achieve this, each of the three replicate experimental sites encompasses 2-3 km of creek line and the adjacent uplands. Each site is located on a separate tributary of the Burdekin River. At each site, five plots were established in 1999 and, following collection of baseline data, prescribed fires were imposed. The experimental design allowed for each of five burning regimes to be imposed on one of the five plots at each site. These treatments were

  1. unburned;
  2. burned once on the dry season;
  3. burned once in the wet season;
  4. burned twice in successive dry seasons;
  5. burned twice in successive wet seasons.

For the sake of this work, 'wet' and 'dry' seasons were defined in terms of the phenological status of the rubber vine; namely, whether rubber vine was in leaf (wet season) or leafless (dry season, taking into account defoliation by biological control agents). The woody and herbaceous composition of each plot was quantified annually.

Initial burning treatments were imposed in the 1999-2000 wet season and the 2000 dry season. Fuel loads prior to these fires were 3000-4000 kg/ha though not evenly distributed, resulting in a variation of fire intensity in different areas. At least some of this variation in fire intensity is attributable to the presence (and uneven distribution of) rubber vine itself. Low-intensity fires especially will often not penetrate dense stands of rubber vine and in this way the weed can reduce land-management options and so impinge upon landscape health.

In spite of the differing character of the prescribed fires, their overall impact on rubber vine was considerable. For example, a single wet-season fire reduced the density of rubber vine from 2147 to 1165 plants per hectare. These fires did not cause significant changes in the densities of most species of native trees and shrubs. Most shrub species sprouted from the base even when they had been subjected to intense fires. The fires stimulated germination of Acacia spp., notably Acacia holosericea. After burning, the herbaceous layer remained dominated by the exotic stoloniferous grass Bothriochloa pertusa (Indian couch) and native perennial tussock grasses, though there was an increase in the abundance of native legumes. In the dry season after burning, herbaceous biomass was lower on burned plots than on unburnt plots.

Interactions between fire and grazing

There were obvious interactions between the effects of burning and the effects of grazing. These interactions would need to be considered in developing practical fire-management strategies for rubber vine fire. Importantly, the concentration of cattle on the unfenced experimentally burnt plots meant that the annual burning regimes of the design could not be followed. Although the paddocks carried sufficient fuel to support a fire, the low fuel loads meant that the fires would not have been effective against rubber vine. It was seen as preferable to delay the second round of fires to give two years between fires. A regime consisting of fires in years one and three of a 10-year cycle would allow for two fires of adequate intensity and sufficiently close in time to be very effective against rubber vine. For rubber vine management on pastoral properties it will always be essential link a fire management program with a grazing management plan.

Effect of fire on fauna

The impacts of riparian burning for rubber vine management on two important faunal groups, birds and reptiles, are also being examined by James Cook University Masters student Leonie Valentine.

Preliminary faunal surveys indicate that reptiles may be useful indicator species to address impacts of burning. This is because of the micro-scale of their habitat requirements and their key position as major predators of another major element of biodiversity, the arthropods. An additional question to be addressed is what effect rubber vine itself has on biodiversity. Early observations suggest that there was less reptile diversity in dense rubber vine infestations than elsewhere in riparian landscapes.

Effects of fire on riparian vegetation

To the extent that data have been analysed so far, it is apparent that the side-effects of individual fires on non-target components of the riparian communities are tolerable. This observation must be interpreted in the light of the fact that effects were of single fires in a single type of riparian vegetation. The results should not be interpreted to mean that riparian vegetation in general is tolerant of fire or that the specific riparian communities that were studied would be resilient in the face of frequent burning.

Future directions

At present this project will end in December 2001. By December 2001, the full complement of planned wet season experimental fires will have been imposed; our current plan is to complete dry-season burning by October 2002. Their effects should be documented over the subsequent 12 months.

The results of this project raise the importance of the interactions between grazing and fire. This lies in the impact that grazing has on the landholders capacity to impose fires of a type that will have a useful impact on populations of rubber vine.

Another issue that should be considered involves the interactions between weed species. Most savanna systems are under threat from more than one weed species. Management that targets one species may open opportunities for invasion by others. This demands that attention be given to understanding and managing 'weed complexes' that may include species of a variety of growth forms and functional responses.



Grice, A.C., Radford, I.J and Abbott, B.N. 2000, 'Landscape and regional-scale patterns of shrub invasion in tropical savannas,' Biological Invasions, 2:187-205.

Grice, A.C. and Campbell, S.D. 2000, 'Weeds in pasture ecosystems: symptom or disease', Tropical Grasslands, 34:264-270.

Campbell, S.D. and Grice, A.C. 2000, 'Weed biology: a foundation for weed management,' Tropical Grasslands, 34:271-279.

Grice, A.C., Campbell, S.D. and Vitelli, J.S. 2000, 'The principles and practice of weed control for northern Australia', Proceedings of the Northern Grassy Landscapes Conference, Tropical Savannas CRC, Darwin.

Project team

Dr Tony Grice, CSIRO Sustainable Ecosystems
Dr Ian Radford, CSIRO Sustainable Ecosystems
Lindsay Whiteman, CSIRO Sustainable Ecosystems
Michael Nicholas, CSIRO Sustainable Ecosystems
Dr Shane Campbell, Queensland Dept Natural Resources and Mines
John McKenzie, Queensland Dept Natural Resources and Mines
Tony Johnson, Queensland Dept Natural Resources and Mines

Janice Jackson
Exotic grass species in tropical savannas of northern Australia


Fire and weeds: what works, what doesn't

Article on the impact of fire on woody weeds and how it affects management decisions. From Savanna Links, Issue 19, July - Sept 2001 [read more...]

Fire burns out large-scale riparian threat

Rubbervine research in Queensland has shown fire to be an effective control method. From Savanna Links, Issue 22, May - July 2002 [read more...]


Dr Tony Grice
Senior Research Scientist (Ecologist)
Tel: 07 4753 8543

Fax: 07 4753 8600

Davies Laboratory, PMB


Info Sheet: Could fire put the rubbervine invader to rout?
Could fire put the rubbervine invader to rout? First-round CRC project on rubbervine and fire [pdf 295.1 kb]