The Nyl River floodplain is one of the jewels in the arid Limpopo Province. The conservation and protection thereof is thus vitally important. The Nyl River Floodplain is an ephemeral floodplain and the largest of its type in South Africa. The wetland is a Ramsar site and provides habitat for a number of endangered species of birds and animals. The aims of this project were to (1) assess the water and sediment quality in the Nyl River system, (2) to determine baseline levels of pollution, (3) to develop a rapid wetland assessment protocol for biomonitoring and (4) to provide a framework for wetland management. Eighteen sites in the Groot Nyl and Klein Nyl rivers, as well as in some of the larger tributaries were selected. Water and sediment were sampled and analyzed to determine metal and nutrient levels. Bacterial analysis also took place at five of the sites along the course of the system. The results obtained from the water analysis indicate that bacterial levels in the system are cause for concern. Although metal levels in the water and the sediment are higher than Target Water Quality and Sediment Guideline Ranges, the metal levels remained relatively constant throughout the system. The metal levels indicated that they pose no potential threat to the system. The comparison between the present and historical ecological state indicated that nutrient levels are increasing in the system. The levels of toxic ammonia did not increase and thus the water quality in the system can thus be classified as fair. The sequential extraction of the sediment indicated that the majority of the metals in the sediment are not readily bioavailable. They were released by the fourth and fifth fractions and will only become available in the presence of strong reducing or oxidizing agents. Organic contaminant levels were also analysed in the sediment. The results indicated traces of PCB’s (Poly-chlorinated Biphenyls) and pyrethroids (Cypermethrin), but concentrations were too low to quantify. iii The third aim of the project was to develop a Wetland Assessment Protocol. A draft version of the protocol was developed using a modified version of the South African Scoring System version 5 (SASS5). Due to the lack of biotopes in the system, the vegetation biotope was chose as sampling habitat. Aquatic invertebrates were collected and a total score was given to each sampling site. The total site rating was determined using a combination of the SASS5 scoring system, a newly developed habitat assessment system and a human impact assessment system, The Wetland assessment protocol identified changes in water quality, but more refinement is required on a system with a greater pollution gradient. The fourth aim of the project was to set up a draft framework for wetland management. The framework is based on the National Estuary Programme of the USEPA. It has been interpreted and adapted for use in wetlands, in a similar way to which USEPA ecological risk assessment guidelines have been adapted for the South African scenario. This research project was thus able to (1) provide baseline values for the Nyl River System, (2) to produce a first draft of a Wetland Assessment Protocol and (3) provide a framework for wetland management. It is envisaged that the information in this thesis will provide useful information in the protection and management of the Nyl River. / Prof. J.H.J. Van Vuren

The Nyl River and its floodplain are situated on the eastern foothills of the Waterberg mountain range in the Limpopo Province of South Africa. Tributaries flowing out of the Waterberg range display unusual downstream changes, as they approach and converge with the Nylsvlei (or Nyl floodplain). Tributary channels decrease in size downstream until, eventually, they disappear altogether forming unchannellized floodouts. On one such floodout, on the farm Driefontein, an actively aggrading piedmont has formed adjacent to the famous Wonderkrater peat mound, known for its pollen record dating back ~45,000 years. Sediments from the aggrading piedmont interlace with Wonderkrater’s peat layers, suggesting that as the piedmont aggrades so too does the peat mound. This setting presents a unique opportunity to study active aggradational processes, and their products, on hillslope deposits and floodout environments. This study aims to describe the geomorphology and nature of depositional processes along the length of the piedmont adjacent to the Wonderkrater peat mound. Cross-sections, drainage channels and vegetation indices based on topographic maps, orthophotographs and hyperspectral images, were created using ArcGIS in order to describe and determine the surface morphology and hydrology of the Driefontein piedmont in detail. Surface soil samples were collected in order to determine particle size distribution, which were in turn compared to vegetation indices and changes in slope elevation. Further grain samples were collected from depth for age dating using Optically Stimulated Luminescence (OSL), as well as to determine grain size distribution in relation to surface sediments and other fluvial environments. Hyperspectral indices were found to correlate to surface grain size distribution, demonstrating that the presence of vegetation acts as a retaining mechanism for particles along hillslopes where incline should be too steep to support fine-grained sedimentary material. Surface sediments were found to demonstrate the characteristics of an alluvial floodout system, affected greatly by the presence of vegetation and slope inclination. Sub-surface samples were characteristic of a colluvial setting, suggesting that pediment retreat and basin fill, coupled with evidential climatic changes, were dominant controls on the pediment’s morphological and aggradational mechanisms. OSL age results estimated the sediments to be between 37.33 and 58.66 ka old. As a result of its unique sedimentary characteristics, a new type of ‘slow creep fan’ class was established in order to describe the characteristics of the Driefontein piedmont.

A Dissertation submitted to the Faculty of Science, University of the Witwatersrand, Johannesburg, in fulfilment of the requirements for the Degree of Master of Science in Geology. Johannesburg 2013 / The Nyl River floodplain, located in the Limpopo Province, is one of the few active sedimentary basins that exist within the South African interior, providing a unique opportunity to study the effect of climate change on fluvial systems. Progradation of tributary fans into the Nyl/Mogalakwena River has raised the surface by 30 m and forced the course of the river westwards towards the Waterberg. Periods of progradation deposited thick sequences of coarse-grained sediments with sand- to gravel-sized mean grain sizes and coarsely-skewed populations in the distal reaches of the tributary fans. These periods were interspersed with periods of relative non-deposition, when active sedimentation on the fan ceased and shallow lakes (or vleis) developed in the trunk river, resulting in deposition of fine-grained, organic-rich, floodplain sediment layers with silt-sized mean grain sizes and finely-skewed distributions in the extreme outer reaches of the tributary fan. The alternating progradational sequences and non-deposition events produced interlayered floodplain and fan deposits in the furthest reaches of the tributary fans along the banks of the Nyl/Mogalakwena River. Incised river cuts within the Rooisloot tributary fan were dated using OSL and 14C techniques. For OSL samples, the SAR protocol was used to measure the equivalent dose and the burial dose was determined using the CAM and MAM. Emission counting methods, including TSAC, GM-beta counting and HRGS were used to determine the dose-rate. The OSL ages ranged from 99 years to 3884 years, constraining the sampled deposits within the late Holocene. Although the 14C ages agreed with this range, carbon contamination of the samples resulted in inverted and overestimated ages. Based on stratigraphic relationships the non-deposition events have been dated at approximately 750–800 years ago, 600 years ago, 475 years BP and 100–150 years ago and two major periods of aggradation at ~ 800–1000 years ago and ~ 500–700 years ago. The rate of aggradation (0.29 cm/year) calculated implies that the entire 30 m deposit could have been deposited in 9 000 years. However, an independent study by McCarthy et al. (2011) proved that tributary sedimentation began prior to 220 ka. Therefore, in order to deposit 30 m of sediment over 220 ka, either the mid – late Quaternary sedimentation rate was lower than the recent past (Late Holocene) or the system periodically undergoes extensive erosion in order to flush the accumulated sediment from the tributary fan system.