Return to search

The vegetation ecology of Mfabeni peat swamp, St Lucia, KwaZulu-Natal

A vegetation survey, on the plant community level of organization, was conducted on the Mfabeni swamp on the eastern shores of Lake St Lucia. The survey included both the sedge/grass swamp and the swamp forest. Mfabeni swamp is a peatland with peat up to 10m in depth. Except for limited studies in Lesotho and KwaZulu-Natal, this is a first vegetation survey on the plant community level of peatland vegetation in southern Africa. Total floristic composition and habitat information were recorded in 214 sample plots, fully referenced by GPS for GIS mapping. The data were captured in the TURBOVEG data base for vegetation data, classified using the TWINSPAN numerical classification algorithm and phytosociological tables were compiled using the MEGATAB computerized table management program. Fourteen plant communities were identified, characterized by diagnostic and dominant species, described, and the localities noted by GIS. The data were also processed by the PCOrd numerical ordination program and the axes of the resulting ordination diagrams were related to environmental conditions. This gave an indication of the environmental factors controlling the existence and distribution patterns of the different plant communities. An ordination was run on the habitat data collected to further illustrate the relationship between the plant communities and environmental variables. It seems that water depth, seasonality of the water depth and certain water chemical properties, such as pH conductivity and dissolved oxygen content of the water play an important role in determining floristic composition and therefore plant community type in the Mfabeni swamp. / Dissertation (MSc(Botany))--University of Pretoria, 2006. / Plant Science / unrestricted

Identiferoai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:up/oai:repository.up.ac.za:2263/24480
Date09 May 2005
CreatorsVenter, Catharina Elizabeth
ContributorsProf G J Bredenkamp, upetd@up.ac.za
Source SetsSouth African National ETD Portal
Detected LanguageEnglish
TypeDissertation
Rights© 2003 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria. No part of this work may be reproduced or transmitted in any form or by any means, without the prior written permission of the University of Pretoria.

Page generated in 0.0017 seconds