The invasion of natural environments by alien species is a significant threat to the ecological integrity of these systems. Senna obtusifolia is an aggressive invasive weed recently introduced to Australia that is having significant impacts on grassland ecosystems on the Cape York Peninsula. Currently the species is inadequately managed and so range expansion continues. The invasion potential of S. obtusifolia in Australia remains unknown, as does much about its behaviour and management in natural systems. This project undertakes extensive mapping and modelling of the current and future distributions and the invasion dynamics of S. obtusifolia in Australia to facilitate early detection of outbreak populations and the development of appropriate management strategies. The mapping and modelling of S. obtusifolia was conducted at three different scales: continental, landscape and local (population). To address these spatial scales, eco-climatic modelling, remote sensing analysis, field experimentation and creation of a model of seed fate was undertaken. Using the climatic preferences of S. obtusifolia displayed internationally, an eco-climatic model (using CLIMEX software) ascertained that S. obtusifolia has a very large invasive potential in Australia. The predicted geographic distribution comprised the entire eastern and northern Australian coastlines, with spread further inland being largely restricted by a lack of moisture. The regional distribution of S. obtusifolia was not successfully delineated using remote sensing technology. Despite possessing favourable traits for detection by remote sensors, poor data quality and inappropriate image scales prevented the weed from being distinguished from other vegetation by multi-spectral satellite imagery and aerial photography. However, the results indicated that refining the data and the techniques used, single S. obtusifolia populations may be detectable in the future. Investigation of the invasion dynamics of S. obtusifolia at the local scale involved multiple field surveys and manipulative experiments during 2002-2005. Field work indicated that little variation in population characteristics (e.g. stem density, soil seed reserve, seed production) existed within populations, but there was variability across populations and between years: the variation between years was very significant. The vegetation type adjacent to the weed population did not affect population attributes; however less competitive, more open and disturbed environments may better facilitate the invasion. The compartment model of seed fate reflecting S. obtusifolia population dynamics demonstrated that change in annual rainfall was unlikely to explain the variation evident between populations and years. Instead, the rate at which dormancy is broken in seeds and the intensity and regularity of fire provided a better explanation of the weed's population dynamics. Early detection of invaders and the prediction of likely sites of invasion provide the most effective means of preventing future invasions. How best to achieve these goals still remains largely unknown. The process undertaken in this study was a relatively quick and reliable method for assessing the seriousness of S. obtusifolia, predicting future outbreaks and for providing clues to long term management. The appropriate use of fire, maintaining high interspecific competition and shade, as well reducing the rate at which dormancy is broken in seeds are all possible methods of managing S. obtusifolia.
| Identifer | oai:union.ndltd.org:ADTP/265482 |
| Date | January 2007 |
| Creators | Dunlop, Elizabeth A. |
| Publisher | Queensland University of Technology |
| Source Sets | Australiasian Digital Theses Program |
| Detected Language | English |
| Rights | Copyright Elizabeth A. Dunlop |
Page generated in 0.0018 seconds