Modelling the current and potential distribution of woody plants, with special emphasis on the importance of spatial scale and environmental factors

The relationship between species distribution and environment has always been a central issue in ecology. The primary aim of this thesis is to explore the role of abiotic and biotic factors in mediating the distribution of woody plants, particularly invasive alien species. Predictions of the future distribution of invasive species are required for management, conservation actions, and legislation. Using predictive geographical modelling, this thesis assesses the role of spatial scale in understanding the determinants of species distribution, modelling invasion spread, and taking conservation decisions. The species distribution databases were collated in a Geographic Information System from various sources (from field mapping to remote-sensing data), and at various spatial scales (from local to sub-continental). Classification and regression trees, a flexible non-parametric statistical technique, were developed for each case study. I show that static modelling approaches, such as classification trees, are appropriate for modelling species distribution at regional to sub-continental scales. More mechanistic approaches, which include biological attributes, are required for accurate modelling of species abundance at local scale. A hierarchy of environmental factors was observed. Abiotic factors such as climate were more important for modelling species distribution at broad scales, whereas biotic factors were the fine-scale drivers of species distribution and abundance. This highlights the scale- dependence of prediction accuracy and of environmental drivers. Predictive habitat modelling was also applied to modelling future land use changes (including invasive alien species) in the context of conservation planning for the Cape Floristic Region. Effective conservation planning requires a detailed assessment of current land use patterns and future land use scenarios. Conservation planning is generally derived at regional scale but implementation takes place at finer scales. I investigated the transition from broad-scale planning to fine-scale implementation. This work emphasises the importance of diverse and fragmented habitats in measuring conservation value. This thesis concludes by linking predictive habitat modelling, species determinants, and application types to the spatial scale analysed.

Grantor dc:publisher.institution

Department of Biological Sciences

Year dc:date.issued

2002

Author dc:creator

• Rouget, Mathieu

Advisors dc:contributor.advisor

• Richardson, Dave
• Cowling, Richard M
• Lavorel, Sandra
• Milton, Suzanne Jane

Language dc:language.iso

eng

Handle dc:identifier.uri

http://hdl.handle.net/11427/9771

OAI identifier oai:identifier

oai:open.uct.ac.za:11427/9771