Browsing by Author "Otto, Mia"

Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    Spatial and temporal changes in Fynbos riparian vegetation on selected upland rivers in the Western Cape
    (Stellenbosch : Stellenbosch University, 2014-04) Otto, Mia; Jacobs, Shayne Martin; Brown, Cate; Reinecke, Karl; Stellenbosch University. Faculty of AgriSciences. Dept. of Conservation Ecology and Entomology.
    ENGLISH ABSTRACT: Restoration practices commonly make use of a reference condition in order to restore a site to a better ecological state than it is currently in. The selection process and relevance of the reference condition has not yet been tested in upland Western Cape rivers especially with relation to spatial and temporal scales. This thesis sought to evaluate whether space (longitudinal) and time (temporal) influence riparian plant community composition (laterally), how it influences the community composition and whether these differences impacts the selection of a reference condition used in restoration practices. In order to investigate the role longitudinal position, sites were selected across three longitudinal zones: mountain stream, transitional and upper foothills. Historic sites used in a previous study on riparian vegetation of upland rivers were resampled and datasets used for temporal comparisons between undisturbed sites, sites recovering after clearing of invasive alien plants and sites affected by fire. Riparian vegetation communities showed differences between longitudinal zones, basins and rivers. The species responsible for marginal zone identity (plants in close proximity to the active channel), determined using relative cover abundance varied, with Isolepis prolifera responsible for the group identity in the mountain stream and transitional sites but in the foothills Calopsis paniculata, Drosera capensis and Metrosideros angustifolia saplings were responsible for lateral zone identity. The lower dynamic (transitional between wet and drybank) had no similarities between different longitudinal zones across rivers. In the lower zone Pteridium aqualinum was mostly responsible for the identity. The upper bank had no single species responsible for group identity. The species described to be typical for the reference condition on these particular rivers by other studies were mostly present in the comparable lateral zone but it was however not always responsible for the identity of the specific lateral zone. By comparing selected environmental variables such as horizontal distance from active channel, elevation and substrate calibre with different longitudinal zones’ riparian vegetation species distribution, different combinations were produced. The mountain streams showed the strongest relationship with horizontal distance and elevation in combination to one another and the upper foothills horizontal distance from the active channel was linked most strongly to vegetation positioning. These results confirm the importance of space when attempting to assess, study or restore riparian communities. Temporally, sites had stronger similarity to data collected during the same sampling period than with historic data. Also, the overall relative species abundance did not show significant change to be present at a site scale. The changes in community composition were found to be due to a lateral zone scale variation in species abundance. As expected the undisturbed rivers showed less variation in species responsible for temporal changes than the recovering and fire-exposed rivers. Species responsible for changes in relative abundance at a lateral zone scale were Metrosideros angustifolia, Morella serrata, Brabejum stellatifolium, Isolepis prolifera, Elegia capensis, Prionium serratum and Calopsis paniculata. Due to the species diversity not changing much temporally but the relative abundance of specific species showing much variation over time it can be concluded that the changes are not diversity based but instead driven by changes in relative abundances of species typical for a lateral zone. The spatial and temporal variation in riparian vegetation community composition was found to be significant enough to suggest that the use of a fixed reference condition for all Western Cape rivers would not be feasible due to clear differences between basins. Secondly when selecting a reference site the spatial location of this site should be within the same longitudinal zone since bank shape does influence riparian plant species distribution. Finally the temporal comparison between sites showed high diversity in species abundances but small differences in diversity overall. This would suggest that a general community description specific to 1) where the site is situated and 2) based on the present riparian vegetation community composition within a specific basin may be more realistic and achievable for restoration and environmental management purposes as opposed to using site descriptions from the past and reference sites too far upstream or downstream from the restoration site.