The Nelson Mandela Bay Municipality (NMBM) is a semi-arid area along the southern coastline of South Africa (SA). Until recently, there was no systematic approach to research on wetland systems in the NMBM. The systematic identification of wetlands was made more difficult by the relatively large number of small, ephemeral systems that can be difficult to delineate. This has meant that fundamental knowledge on wetland distribution, structure and function has been limited and, consequently, management and conservation strategies have been based on knowledge on systems from other regions of the country. Environmental processes occur at different spatial and temporal scales. These processes have an effect on the abiotic factors and biotic structure of wetlands, resulting in inherently complex systems. The location of the NMBM provides a good study area to research some of these environmental and biological attributes at different spatial scales, due to the variability in the underlying geology, geomorphology, vegetation types and the spatial and temporal variability in rainfall, within a relatively small area of 1951 km2. Thus, the aim of this study was to determine the factors influencing wetland distribution, structure and ecosystem functioning within the NMBM. The first Research Objective of work presented here was to identify wetlands using visual interpretation of aerial photographs. A total of 1712 wetlands were identified within the NMBM using aerial photographs, covering an area of 17.88 km2 (Chapter 5). The majority of these wetlands were depressions, seeps and wetland flats. Valley bottom wetlands (channelled and unchannelled) and floodplain wetlands were also identified. A range of wetland sizes was recorded, with 86% of the wetlands being less than 1 ha in size and the largest natural wetland being a floodplain wetland of 57 ha, located south of the Swartkops River. The identified wetlands were used to create a wetland occurrence model using logistic regression (LR) techniques (Chapter 5), in accordance with Objective 2 of the study. An accuracy of 66% was obtained, which was considered acceptable for a semi-arid climate with a relatively high degree of spatial and temporal rainfall variability. The model also highlighted several key environmental variables that are associated with wetland occurrence and distribution at various spatial scales. Some of the important variables included precipitation, evapotranspiration, temperature, flow accumulation and groundwater occurrence. Wetland distribution patterns were described in Chapter 6. Spatial statistics were used to identify whether wetlands are clustered and, therefore, form mosaics within the surrounding landscape (Objective 3). Systems were found to be highly clustered, with 43% of wetlands located within 200 m of another system. Clustering and wetland presence was especially prominent in the southern portion of the Municipality, which is also associated with a higher mean annual precipitation. Smaller wetlands were also significantly more clustered than larger systems (Average Nearest Neighbour statistic, p-value < 0.0001). Average distances also significantly varied according to HGM type, with depressions being the most geographically isolated wetland type compared to the other HGM types. Overall, distances between wetlands indicated good proximal connectivity. Potentially vulnerable areas associated with wetland systems were identified successfully using landscape variables, in accordance with Objective 4. These variables were: land cover, slope gradient, flow accumulation, APAN evaporation, mean annual precipitation (MAP) and annual heat units. The existing Critical Biodiversity Network was also used in connection with these variables to further identify potentially vulnerable areas. The abiotic and biotic characteristics were decribed for three hydrogeomorphic (HGM) types at a total of 46 wetland sites (Chapter 7), as per Objective 5. Depressions, seeps and wetland flats were sampled across the different geological, vegetation and rainfall zones within the NMBM. The wetland sites were delineated up to Level 6 of the Classification System used in SA, and the various abiotic and biotic characteristics of these systems were defined. A total of 307 plant, 144 aquatic macroinvertebrate and 10 tadpole species were identified. Of these species, over 90 species were Eastern Cape and SA endemic species, as well as three threatened species on the IUCN Red List. Multivariate analyses (including Bray-Curtis similarity resemblance analyses, distance-based redundancy analyses, SIMPER analyses and BIOENV analysis in Primer), together with environmental data, were used to define community structure at an HGM level, in accordance with Objective 5. The importance of the spatial scale of the environmental data used to define plant and macroinvertebrate community structure was described in Chapter 7, to address Objective 6. The results showed that both broad-scale and site-level characteristics were important in distinguishing community structure within the HGM types that superseded general location, the sample timing or the stage of inundation. These results also indicated that a combination of both landscape and site-level data are important in defining the community structure in the various HGM types. Some of the important environmental variables that explained some of species assemblages were similar to those in the wetland occurrence model (Chapter 5), with some additional hydrological and soil physico-chemical parameters (e.g. soil electrical conductivity, soil pH, and surface and subsurface water nutrients). These significant variables indicate the complex, multi-scalar role of environmental attributes on wetland distribution, structure and function.
Identifer | oai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:nmmu/vital:26960 |
Date | January 2016 |
Creators | Melly, Brigitte Leigh, Gama, Phumelele T |
Publisher | Nelson Mandela Metropolitan University, Faculty of Science |
Source Sets | South African National ETD Portal |
Language | English |
Detected Language | English |
Type | Thesis, Doctoral, PhD |
Format | xxi, 314 leaves, pdf |
Rights | Nelson Mandela Metropolitan University |
Page generated in 0.0023 seconds