Predicting rates and patterns of alien plant spread
Doctoral Thesis
1998
Permanent link to this Item
Authors
Supervisors
Journal Title
Link to Journal
Journal ISSN
Volume Title
Publisher
Publisher
University of Cape Town
Department
Faculty
License
Series
Abstract
The invasion of alien plants into natural ecosystems is a widespread phenomenon that impacts negatively on ecosystem structure and functioning. The invasion and subsequent spread of an alien plant population is equivalent to the processes of colonisation and migration. This implies that the existing toolbox of techniques developed for plant succession research should be useful for predicting plant invasions. Practitioners of invasion biology . have, however, found biological invasions frustratingly difficult to predict. The aim of this thesis was to use succession models to develop a modelling protocol for predicting rates and patterns of alien plant spread. The rationale was that such a model would both improve our understanding of the determinants of invasions and allow us to make predictions on the rates and patterns of alien plant spread. Such predictions are likely to be extremely valuable for the tactical and strategic management of plant invasions. Many modelling approaches could be .. adopted: the need to transcend the gap from general models of plant spread to management models led me to select a spatially explicit simulation modelling approach. The modelling approach is developed by comparing the behaviour of an individual based spatially explicit simulation (SEIBS) model of plant spread to the behaviour of the classic Skellam reaction diffusion model. This process also served to define the model's sensitivity and data requirements. The model's heuristic value is demonstrated by exploring why it is so difficult .to predict which plant will invade which environment. The model also ·provides a useful tool for exploring the role of long-distance dispersal in determining invasion rates. I show that long-distance dispersal is extremely difficult to define statistically, but is a key determinant of invasion rates. The model is validated using independent data on the spatial demography of two invasive species, Acacia cyclops and Pinus pinaster, and independent historical reconstructions of invasions. This validated model was then used to develop a dynamic landscape-extent model. This scaled-up model explores the optimal strategies for clearing alien plants and the ability of different clearing strategies and funding schedules to mitigate the threat that alien plants pose to native species. I conclude that models that are tightly linked to understanding of ecological processes and to field data can be used to rapidly develop predictive models. The development of these models challenges our fundamental ecological understanding and, therefore, emphasises the interplay between data, theory and prediction.
Description
Keywords
Reference:
Higgins, S. 1998. Predicting rates and patterns of alien plant spread. University of Cape Town.