Thesis (MscEng)--University of Stellenbosch, 2005. / ENGLISH ABSTRACT: The environmental impacts of manmade changes to rivers are, in modern days, extremely important. The impact needs to be quantified in order to predict future happenings and to assist in determining preventative measures. Dam construction forms an essential part of modern life to provide the necessary water demand for the ever increasing population. One manmade change that has a major impact on rivers is large dam developments in the upper reaches of rivers. These developments normally have the following effects on the lower reaches of the river due to the change in flow regime: • Narrower main channel, • Deeper main channel, • Reduced sediment transport, • Changes in sediment erosion and deposition patterns, • Less frequent flood plain inundation, • Overgrown flood plains, • Changes in ecological and biological parameters. All the above mentioned factors play an important part in the stability of the ecological and biological parameters. Prior to construction of a dam, however, the baseline ecological, physical Ichemical, hydrological, hydraulic and social conditions of the river system, including its associated groundwater, estuarine and floodplain components, would need to be established, to allow clear identification of future changes as a result of implementation of the instream flow requirements (lFR), as well as to guide the post-construction monitoring programme itself. The discharge and sediment transport changes that might occur as a result of the Berg River Dam development was thoroughly investigated in this reseach. The physical processes and mathematical modelling that formed part of this investigation are discussed in this thesis. The mathematical modelling was carried out using MIKE 11, software developed by the Danish Hydraulic Institute for Water and Environment. This program was developed especially for one dimensional modelling of hydrodynamics and sediment transport within a river system. The results obtained from the simulations were used to calculate a hydrodynamic and sediment mass balance for both the pre and post dam scenarios in the Berg River, Western Cape. The influence of artificial flood releases from the Berg River Dam which is currently under construction was also investigated. Most of the data needed for the calibration and verification of the mathematical model was obtained from field work. Suspended and bed load sediment samples were taken in order to aid in the calibration of the sediment transport model. The bed roughness coefficients of the various sections of the river were calibrated against actual recorded water levels measured during flood events that occurred in 2003 and 2004. An intensive study of the incipient motion of cobbles and boulders in the upper reaches of the Berg River was carried out. Unfortunately it has to be stated that the transport of cobbles and boulders cannot be simulated by most computer programs as almost all the transport models available are only defined up to a maximum diameter of 20mm. The lack of consistent sediment load data as well as the low rainfall during 2003 and 2004 presented problems during the calibration of the sediment transport model; it is therefore recommended that sediment load sampling is continued for at least two years in order to verify the calculated sediment yield for the various catchments. From this study it can be concluded that the Berg River Dam will have a significant effect on the downstream river morphology of the first 70 kilometres of the river. Thereafter the effect will decrease to a minimum. The proposed artifical flood releases are effective in reinstating the river to its present state. This study also showed that a fully hydrodynamic model of a large river system such as the Berg River can be calibrated and validated. / AFRIKAANSE OPSOMMING: Die omgewingsimpak van mensgemaakte veranderinge in natuurlike riviere raak meer en meer belangrik. Hierdie impak moet gekwantifiseer word om sodoende toekomstige gebeurtenisse te verstaan en om die nodige voorkomingsmaatreëls in werking te stel. Om aan die waterbehoefte te voldoen vorm damkonstruksie 'n belangrike deel van die hedendaagse samelewing. Damme is moontlik een van die veranderinge aan die ewewig van riviere wat die grootste impak op die stroomaf dele van die rivier het. Die volgende veranderinge ten opsigte van die hoofkanaal word meestalopgemerk na die konstrukie van 'n dam in die hoërliggende dele van riviere: • Nouer hoofkanaal, • Dieper hoofkanaal, • Afname in die volume sediment wat vervoer word, • Veranderinge ten opsigte van die erosiepatrone, • Vloedvlaktes word minder oorstroom, • Toegegroeide vloedvlaktes, • Veranderings aan die ekologiese en biologiese parameters. Die bogenoemde faktore speel 'n belangrike rol ten opsigte van die stabiliteit van die ekologiese en biologiese veranderlikes. Alvorens die konstruksie begin moet die natuurlike ewewigstoestand van die rivier ten opsigte van die ekologiese, fisiese, chemiese, hidrologiese, hidrouliese, en sosiale faktore gemeet word. Dit moet gedoen word om sodoende toekomstige veranderinge wat moontlik mag plaasvind as gevolg van die damkonstruksie en die omgewingsloslatings te kwantifiseer. Hierdie metings word ook gebruik om die na - konstruksie monitering te evalueer. Die veranderinge ten opsigte van vloei en sedimentvervoer wat moontlik mag plaasvind as gevolg van die Berg Rivier Dam konstruksie is noukeuring ondersoek. Die fisiese prosesse as ook die wiskundige modelle wat gedurende hierdie studie aangewend is word in hierdie tesis bespreek. Al die wiskundige modellering is met MIKE 11 wat deur die Deense Instituut vir Water en Omgewing ontwikkel is, uitgevoer. Hierdie sagteware is spesifiek vir die een dimensionele simulasie van hidrodinamika en sediment vervoer van riviere ontwikkel. Die resultate wat verkry is, is gebruik om 'n hidrodinamiese en sedimentvervoer massabalans vir beide die voor - en na - dam toestande op te stel. Die effek van beheerde, kunsmatige vloede is ook ondersoek. Die meeste van die data wat benodig word om die wiskundige model the kalibreer is verkry deur fisiese veldwerk te doen. Sediment monsters is geneem gedurende die winter seisoen van 2003 en 2004. Hierdie data is gebruik om die sediment vervoer model te kalibreer. Die bodemruheid van die verskillende dele van die rivier is gekalibreer teen gemete vloedhoogtes tydens 2003 en 2004. 'n Deeglik studie van die begin van beweging van die groter partikels wat in die bolope van die Berg rivier aangetref word is uitgevoer. Ongelukkig moet dit genoem word dat hierdie een van die groot tekortkominge is van die model aangesien dit nie in die sedimentvervoer model ingekorporeer kan word nie omdat die meeste van die sedimentteorieë wat tans beskikbaar is net gedefinieer is tot 'n partikel groote van 20 mm. Die tekort aan deurlopende sediment lading data as ook die lae reënval gedurende 2003 en 2004 het probleme veroorsaak tydens die kalibrasie van die sedimentvervoer model. As gevolg hiervan word daar aanbeveel dat die insameling van sedimentdata vir ten minste nog twee jaar volgehou word ten einde die beskikbaarheid van sediment in die verskeie opvanggebiede deeglik na te gaan. Uit hierdie studie kan afgelei word dat die Berg rivier dam 'n beduidende effek op die morfologie van die eeste 70 kilometer van die Berg Rivier sal hê. Verder stoomaf sal die effek minder sigbaar wees. The beheerde kunsmatige vloedloslatings speel 'n beduidende rol in die normalisering van die sedimentvervoer van die Berg rivier. Hierdie studie toon dat 'n volledige hidrodinamiese model van 'n groot rivierstelsel gekalibreer kan word.

Thesis (MTech (Tourism and Hospitality Management))--Cape Peninsula University of Technology, 2006 / This research examines the ways in which local residents living in close proximity to the event perceived the Berg River Canoe Marathon as a major event in the Western Cape. By identif'ying these different patterns ofperceptions within the community, it is possible to gain a better understanding of both the tangible and intangible impacts of events and how they differently affect the quality of life of local residents as individuals and the community as a whole. The research method was based on a convenient sampling approach. A questionnaire survey of two hundred residents was conducted and five different dimensions of perceiving the impacts of events were examined ranging from extremely negative to very positive. The impacts analysed included social, economic and environmental impacts. The residents in all the five locations perceived community pride and entertainment to be very positive and were in favour ofthe continuation of the event. 1n profiling the residents who held each of the representations identified, it was found that the most negative group was those residents who had little or no involvement in tourism and are not interested in canoeing as a sport. Thus, it is necessary for event organisers and event managers to get the members of community involved in the planning and management of the event to spread the benefits to a larger portion of the community.

Thesis (MTech (Tourism and Hospitality Management))--Cape Peninsula University of Technology, Cape Town, 2006 / The concept of service quality within sport tourism is important and plays a major role in differentiating each sport tourism product from its competitors. The purpose of this study was to evaluate the level of service quality within South African sport tourism and in the tourism industry in general, by focusing on a specific case study, the Berg River Canoe Marathon. The study discussed the importance of service quality within sport tourism, while the methods and instruments applicable to assess the level of service quality within sport tourism, were also addressed. The survey was conducted across selected viewpoints, which included the starting and the finishing points of the marathon. A total of 100 survey questionnaires were distributed to spectators and participants at the event using purposive sampling, while stratified random sampling was used to determine the viewpoints. The SERVQUAL dimensions were incorporated in the questionnaires as the basis to determine the level of service delivery at the event. The key fmdings for the study include a lack of service delivery consistency across the selected areas, the acceptance of service delivery by some respondents owing to the nature of the event and less interaction between event organisers and spectators. In addition, the starting and the finishing points were the areas, which were found to deliver high quality services

>Magister Scientiae - MSc / Previous studies have shown that the use of chloride mass balance (CMB) method is a suitable and practical approach to estimate groundwater recharge. This enables the prediction of groundwater availability to inform practical strategies for managing groundwater resources. However, such studies have largely applied the chloride mass balance method on national and catchment scales with limited focus on quaternary catchment level (QCL). Neglecting the chloride mass balance method at quaternary catchment level limits practical management and utilization of water resources at quaternary catchment level. The goal of the current study was to prove that 1) the chloride mass balance method should be applied at quaternary catchment level to ensure practical assessment of groundwater availability and that 2) chloride mass balance assessment should be accompanied with supplementary methods for its application in quaternary catchments of similar physiographic and hydrogeologic conditions. To achieve these goals, the present study assessed the application of chloride mass balance method on a pilot scale used alongside rainwater infiltration breakthrough (RIB) and water table fluctuation (WTF) methods to estimate the groundwater recharge as an indicator of groundwater availability. The pilot area (PA) was in the upper Berg River catchment in Western Cape in South Africa. Chloride concentrations were determined in groundwater samples collected from boreholes and rain water in rain gauges in the pilot area. Rainfall and borehole water levels in the pilot area were used in water table fluctuation and rainwater infiltration breakthrough analyses. As quality assurance, the specific yield data obtained from the pumping test were compared to those determined with the linear regression model. This established the reliability of the analysis i.e. the relationship between groundwater level and rainfall. Mean groundwater recharge values calculated using the chloride mass balance, rainwater infiltration breakthrough and water table fluctuation methods were 27.6 %, 23.67 %, and 22.7 % of the total precipitation received in the catchment, respectively. These results indicate that the use of these three methods have potential to estimate groundwater recharge at quaternary level which is the basic unit of water management in South Africa. These findings agree with previous studies conducted in the same catchment that indicate that mean groundwater recharge ranges between 18.6 % and 28 % of the total precipitation. In the future, these methods could be tested in catchments which have physiographic and hydrogeologic conditions similar to those of the current pilot area. / African Union (AU)

Magister Scientiae - MSc / A study of the microbial diversity in sediments of the Great Berg River estuary is carried out using modern molecular phylogenetic methods. The aim of this study was to determine the effect of (pollution by) the effluents of the fish industry on the composition of the microbial community in the sediments. The diversity in microbial groups of sediment samples that received wastewater from the local fishing industry was investigated by a PCR-DGGE (polymerase chain reaction-denaturing gradient gel electrophoresis) approach and compared to an unaffected site. / South Africa

Since the demand for fresh potable water increases every year, it is important to have future water demand strategies in place. People expect a secure, high quality, water supply and the water supply industry is governed by increasingly stringent water quality guidelines and legislation. The Cape Metropolitan Area (CMA) faces the challenge of an increasing demand for fresh water in excess of the existing supply. The City is responsible for the planning and development of the local water supply resources as well as managing the water demand in the CMA and to supplement the water supply to the City of Cape Town from local sources. The ‘Voëlvlei Augmentation Scheme’ was identified as one of the options to augment the water supply to the CMA. This option would involve pumping winter water from the Berg River via a pipeline to the Voëlvlei water treatment plant (WTP). The Voëlvlei WTP was designed to treat water from the adjacent Voëlvlei Dam. This Voëlvlei WTP raw water has a higher turbidity and a lower colour in comparison to the Berg River water. The plant’s treatment conditions were optimized to remove this high turbidity. The Voëlvlei WTP raw water also contains a relatively high manganese concentration and coagulation therefore occurs at a high pH with ferric sulphate to remove the manganese during the initial stages of the water treatment process. As the quality of the Berg River water is different to that of the Voëlvlei WTP raw water, it might not be possible to treat the Berg River water at the Voëlvlei WTP using the plants current treatment parameters. The Berg River water could possibly be blended with the Voëlvlei WTP raw water before treatment at the WTP. If the Berg River water, or its blends, could not be treated at the Voëlvlei WTP using the plants current treatment parameters, then this water would have to be pre-treated before entering the plant. Various forms of pre-treatment could be used, e.g., conventional water treatment using either aluminium or ferric sulphate as primary coagulants or ion-exchange water treatment using the MIEX® resin or even a combination of both. The main objective of this study was to determine a treatment strategy for the Berg River water at the Voëlvlei WTP. It is therefore important to determine if the Berg River water could be treated at the Voëlvlei WTP using the current treatment regime. Also, if the Berg River water should be blended with the Voëlvlei WTP raw water, this study would determine which blend would be the most suited for treatment at the Voëlvlei WTP. If the Berg River water could not be treated directly at the Voëlvlei WTP, a pre-treatment strategy for this water should be determined. The cost of pretreatment of the Berg River water as compared to the cost of direct treatment at the Voëlvlei WTP should also be evaluated. In order to determine the best treatment strategy for the Berg River water at the Voëlvlei WTP, it was important to sample the Berg River water and the Voelvlei WTP raw water at regular intervals over a period of at least a year to determine its quality and the impact of seasonal changes. Various laboratory physical (e.g., turbidity) and chemical (e.g., total alkalinity) analyses were conducted on the Berg River water and Voëlvlei WTP raw water to determine its quality. The experimental procedure focused mainly on the Jar test which simulates the coagulation, flocculation and sedimentation processes at the Voëlvlei WTP. Jar tests were conducted on the Berg River water and the Voëlvlei WTP raw water using ferric sulphate and aluminium sulphate as coagulants to determine the optimum pH and optimum coagulant dosage concentration for each coagulant. The Berg River water was also blended with the Voëlvlei WTP raw water in three different proportions and Jar tests were conducted on these blends using ferric sulphate as the coagulant at a coagulation pH of 5.0 and a Fe3+ dosage of 5.0 mg/L. Jar tests were also conducted on these blends with the Voëlvlei WTP treatment parameters using ferric sulphate as the coagulant at a coagulation pH of 9.2 and a Fe3+ dosage of 3.5 mg/L. The analytical results showed a similar pattern for the characterization of the Berg River water and the Voëlvlei WTP raw water. The iron and aluminium concentrations were consistently low during the summer months with significant increases during the winter months. There were no significant seasonal impact on the UV absorbance and colour. The Jar test results of the Voëlvlei WTP raw water and the Berg River water with ferric sulphate as the coagulant showed an optimum Fe3+dosage of 3.0 to 4.0 mg/L and 4.0 to 6.0 mg/L, respectively, with an optimum coagulation pH range of 6.6 to 9.5 and 5.0 to 10.0, respectively. The Jar test results of the Voëlvlei WTP raw water and the Berg River water with aluminium sulphate as the coagulant showed an optimum Al3+ dosage of 2.5 to 3.0 mg/L and 4.0 to 5.0 mg/L, respectively, with an optimum coagulation pH of 6.0 to 7.0 and 6.0, respectively. The Jar test results obtained for all 3 blends were similar to each other. The UV absorbance of the treated water was consistently below the operational specification, while the turbidities were inconsistent and did not always comply with the SANS 241:2006 Specification (Class I) for drinking water. The iron of the treated water was also consistently above the specified value of <0.200 mg/L. The Jar tests conducted on all 3 blends, with the Voëlvlei WTP treatment parameters, also yielded similar results. The UV absorbance of the treated water was consistently above the maximum operational specification of 0.100, while the turbidities were also consistently above the SANS 241:2006 Specification of <1 NTU. Both ferric sulphate and aluminium sulphate can be used as coagulants to treat the Berg River water, although ferric sulphate would be the preferred choice due to its wide coagulation pH range and also because of differences in their health effects. The Voëlvlei WTP coagulates at a pH of 9.2 to remove turbidity and any manganese that might be present in the raw water. The manganese would not be removed at the low coagulation pH of aluminium sulphate. The specified treatment parameters, including the Voëlvlei WTP treatment parameters, used in treating the raw water blends were not effective and further investigation and research is necessary to determine its optimum treatment parameters. This study concluded that the Berg River water cannot be effectively treated at the Voëlvlei WTP using the plants treatment parameters, even if it is blended with the Voëlvlei WTP raw water. Therefore, the best treatment strategy for the Berg River water at the Voëlvlei WTP would be pre-treatment of the water before entering the Voëlvlei WTP. Although there are various ways of pre-treating the Berg River water, this study has identified the following possible pre-treatment strategies:<ul><li>pre-treatment with ferric sulphate and lime</li><li> pre-treatment with ferric sulphate and lime in conjunction with MIEX® resin</li><li> pre-treatment with MIEX® resin only</li></ul> Further research and investigation would be necessary to determine the best pretreatment strategy in terms of cost and efficiency. The pre-treated Berg River water would have to pass through the Voëlvlei WTP treatment process (i.e. high coagulation pH) to remove any manganese that might be present in the water. It is recommended that more samples should be taken at various points along the Berg River upstream of the Voëlvlei WTP over a longer period of time to compare the quality of water at these points in the river and also to monitor the effect of various run-off sites. Further research and investigation is necessary to determine the optimum treatment parameters for the Berg River water when blended with the Voëlvlei WTP raw water. Other blending ratios, different to those used in this study, should also be investigated. A more in-depth investigation is also necessary to determine the actual capital and operational costs for the pre-treatment of the Berg River water. / Dissertation (MSc)--University of Pretoria, 2011. / Chemical Engineering / unrestricted

Magister Scientiae (Integrated Water Resource Management) / The Klein Berg River catchment is intensely cultivated with orchards, vineyards and wheat, while also ensuring a water supply to the main urban center, Tulbagh, and the two conservation areas (Waterval and Groot Winterhoek). The primary objective of this thesis is to determine channel change over a long and short time period, and to relate these changes to landuse change within the catchment. Assessing stability of a selected reach within the catchment was done on a short term basis with the use of erosion pins and cross profiles, while aerial photographs of over 55 years (acquired during 1942, 1967, 1987 and 1997) which were analysed using Geographic Informations Systems. Rainfall and discharge data, which were available for a period of 49-years were statistically analysed and used to determine trends. Vegetation characteristics were assessed by means of transects within the study reach. The results over the short time period (18 months) indicate noticeable channel change in the form of erosion and deposition within the channel. Bank material composition and riparian invasive alien vegetation play an important role in bank stability. Sand was the dominant grain size of the bank material, and fluvial entrainment occurred during periods of high flow. Woody alien trees prevent the growth of protective ground vegetation, and thus the soil is prone to erosion. Undercutting was also observed with the invasive woody trees, resulting in treefall. Debris dams were also common in the channel and depending on their position in the channel, either cause or prevent bank erosion. Landuse change over the 55-year period illustrated its effects on channel stability. Shrublands within the catchment has been replaced with invasive alien vegetation along the riparian zone, while shrublands along the Obiekwa Mountains, were replaced with cultivated lands. The patterns (shape and size) of lateral and point bars within the study area changed significantly within the 55-year period, which indicates a change in the discharge and sediment dynamics within the catchment. The change in sediment dynamics may be due to agricultural activities and urbanization. The increased trend in rainfall, especially during the winter season within the catchment is also an important catchment control. The study has revealed the integrated nature of variables within the catchment. It is thus recommended that a holistic and integrated approach at a catchment scale is required in the assessment of channel change of a river. / South Africa

Thesis (MSc)--Stellenbosch University, 2012. / ENGLISH ABSTRACT: A trial closure of the emergency gate of the Berg River Dam was undertaken by the Trans- Caledon Tunnel Authority (TCTA) on 12 June 2008. The air vent downstream of the emergency gate was designed to introduce air to mitigate the negative pressures that were expected in the conduit during emergency gate operations. The emergency gate has to close when the radial gate at the downstream end of the outlet conduit fails. Contrary to the theoretical design, the measured air vent velocities in the field indicated that, while the emergency gate was closing, very large volumes of air were apparently continuously being released from the air vent, commencing when the gate was about 30% closed (i.e. 70% open). This is in contrast to what the design intended, namely that air should have been drawn into the vent. This thesis is concerned with the testing of a 1:14.066 physical model representing the outlet works and air vent of the Berg River Dam as a means to determine the reasons for the release of large volumes of air from the air vent during the trial closure in 2008. It also seeks solutions to mitigate the excessive airflow from the air vent. It was concluded that the air velocity in the air vent was independent of the rate of closure of the emergency gate, but to increase with increasing water head. The problem at the Berg River Dam was determined to be one of air blowback. Modifications were made to the configuration of the model in order to determine whether the configuration of the outlet works caused air to be released from the air vent. It was determined that the downward sloping roof at the outlet of the conduit, used to accommodate the radial gate chamber, was the cause of the air blowback phenomenon. An additional air vent was fitted directly onto the conduit at the constriction was found to be ineffective in reducing the air blowback. It was concluded that there are no rational structural change that can prevent or inhibit a recurrence of the blowback phenomenon in the Berg River Dam outlet conduit. The recommendation follows that the outlet conduit should not be constricted by any structural or mechanism further downstream in the conduit. / AFRIKAANSE OPSOMMING: ʼn Toetssluiting van die noodsluis van die Bergrivierdam is op 12 Junie 2008 deur die TCTA (Trans-Caledon Tunnel Authority) uitgevoer. Die lugskag stroomaf van die noodsluis is ontwerp om lug in te voer om die verwagte negatiewe drukke tydens die noodsluissluiting te beperk. Die noodsluis moet sluit indien die radiaalsluis aan die einde van die uitlaatpyp sou faal. In teenstelling met die teoretiese ontwerp, het die gemete lugsnelhede in die lugskag in die veld aangedui dat groot volumes lug voortdurend uit die lugskag vrygelaat word wanneer die noodsluis ongeveer 30% toe is (dit wil sê 70% oop). Dit is in teenstelling met die ontwerp, want die lugskag is ontwerp vir die insuig van lug. Hierdie tesis het ten doel om die redes vir die vrylating van groot volumes lug uit die lugskag vas te stel met behulp van ʼn 1:14.066 fisiese skaalmodel van die uitlaatwerke en lugskag van die Bergrivierdam soos getoets tydens die inwydingstoetssluiting in 2008. Die toetse op die model het getoon dat die lugsnelheid in die lugskag onafhankik van die sluistoemaak tyd is, maar verhoog met die toename in die watervlak. Die Bergrivier dam probleem was bepaal as die van lug terugslag. Die model is gewysig ten einde te bepaal of die spesifieke samestelling van die uitlaatwerke die oorsaak van die vrystelling van lug uit die lugskag is. Die analises en verandering aan die uitleg toon aan dat die skuins afwaartse dak van die uitlaattonnel om die radiaalsluiskamer te huisves die rede was vir die vrylating van die lug uit die lugskag. ‘n Addisionele lugskag was gebou in die dak van die uitlaattonnel reg bo die sametrekking, maar was oneffektief om die terug vloei van lug te verminder. Die gevolgtrekking is dat daar geen rasionele strukturele verandering aangebring kan word aan die Bergrivier dam om die vrystelling van lug uit die lugskag te verhoed of te verminder nie. ’n Aanbeveling vir toekomstige ontwerpe is dus dat die uitlaattonnel nie beperkend by die uitlaatend moet wees nie.

Thesis (MSc Food Sc (Food Science))--University of Stellenbosch, 2010. / ENGLISH ABSTRACT: Over the last decade, studies on the quality of the water in many of South Africa.s rivers revealed an increase in pollution levels (DWAF, 1996; DWAF, 2000). In agricultural areas of the Western Cape many farmers draw water for irrigation from nearby rivers without knowing what the microbial pollution level is. In order to investigate the current quality of river water in some of the Western Cape.s major rivers and to explore the possibility of pathogen carryover from water onto produce, this research project was initiated. In an exploratory study over a five month period, the microbiological and water chemistry of three selected sites from the upper Berg and two from the Plankenbrug Rivers were assessed. Studied organisms included Salmonella, Staphylococcus, Listeria, endosporeformers, coliforms, Escherichia coli and intestinal Enterococci, while the chemical analysis consisted of pH, alkalinity, conductivity and chemical oxygen demand (COD). Faecal coliform counts ranging from 540 to 1 700 000 cfu.100ml-1 and 490 to 160 000 cfu.100ml-1 were found for the Berg and Plankenbrug Rivers, respectively. The water temperature ranged from 12º to 21ºC with COD values always below 100 mg.L-1. Potential human pathogens such as Salmonella, Staphylococcus, Listeria, endosporeformers, E. coli and intestinal Enterococci were frequently isolated from all five sites that were sampled. These results are of great concern to farmers, fresh produce retailers and consumers alike as the river water is regularly drawn for irrigation of produce that is later consumed raw or after a minimal processing step. From the exploratory study it was concluded that the water from all the sites were not suitable for use in irrigation practices as they regularly exceeded the guidelines for faecal coliforms and E.coli as set out by South African authorities. Irrigation with faecally polluted river water is one way that fruit and vegetables can become contaminated with foodborne pathogens. The risk of disease transmission from potential pathogens present in the irrigation water is influenced by the microbial load present and the numbers carried over to the produce. In this study the carry-over and survival of Escherichia coli on green beans, sugar-snap peas and cocktail tomatoes was assessed under controlled laboratory conditions. The produce was exposed to E.coli under different combinations of exposure times (5, 15 or 30 min), drying times (30 or 120 min) and different inoculum concentration ranges (107, 105, 104, 10³ and 10ª). In all cases a reduction of at least one log value in original inoculum number was found with the 10ª inoculums showing no survivors. Follow-up studies with an exposure time of 60 min and increased drying times of 6 or 12 h. Neither of these parameter changes affected the variation in numbers for the same inoculum or the E.coli survivors. Similar E.coli loads to those detected on the green beans were detected on the sugar-snap peas. The number of survivors on the cocktail tomatoes was much lower than found for the beans and peas. The number of survivors on the cocktail tomatoes was much lower than found for the beans and peas. This was attributed the 'smooth'surface of the tomato skin probably making attachment of the E.coli bacteria difficult. With the exception of the 10ª inoculum range, all E.coli survivors detected on the three types of produce studied exceeded the guideline numbers set for fresh produce. If similar survival patterns are to be found in the environment then results from this study should serve as a warning that the Plankenburg river water is unsafe for use in the irrigation of fresh produce. Some farmers are already treating river water with chlorine prior to irrigation to eliminate the chance of pathogen transfer onto produce. However, this is not a feasible solution for most farmers due to the high cost of implementing a system such as this. / AFRIKAANSE OPSOMMING: Oor die afgelope tien jaar het studies oor die gehalte van die water in baie van Suid-Afrika se riviere 'n toename in besoedelingsvlakke getoon. Baie boere in landbou gebiede van die Wes-Kaap wat watertrek vir besproeiing uit die nabygelee riviere is onbewus van die mikrobiese besoedelingsvlak van die rivier. Hierdie navorsingsprojek is geinisieer ten einde die huidige gehalte van die water in sommige van die Wes-Kaap se riviere te monitor, asook die moontlikheid van patogeen oordrag van die water na vars produkte te ondersoek. Met behulp van 'n verkennende studie oor . vyf maande tydperk, is die mikrobiologiese en water chemie kenmerke van drie geselekteerde moniteringspunte uit die boonste Bergrivier, en twee uit die Plankenburgrivier bemonster. Die bestudeerde organismes het ingesluit: Salmonella, Staphylococcus, Listeria, endosporeformers, fekale kolivorme, Escherichia coli en intestinale Enterococci. Die chemiese parameters het bestaan uit pH, alkaliniteit, geleiding, temperatuur en chemiese suurstof behoeftes (CSB). Fekale kolivorme tellings het gewissel van 540 - 1 700 000 cfu.100ml-1 en 490 tot 160 000 cfu.100ml-1 is gevind vir die Berg- en Plankenburgriviere, onderskeidelik. Die water temperatuur het gewissel van 12 º tot 21ºC met die CSB-waarde altyd onder 100 mg.L-1. Potensiele menslike patogene soos Salmonella, Staphylococcus, Listeria, endosporeformers, E. coli en intestinale Enterococci is dikwels geisoleer uit al vyf die moniteringspunte. Hierdie resultate is van groot belang vir boere, kleinhandelaars en varsprodukte verbruikers omdat die riviere se water gereeld onttrek word vir besproeiing van produkte wat later gebruik word in rou of in minimale geprosesseerde vorm. Vanuit die verkennende en uit die langer termyn studie (12 maande) is die gevolgtrekking gemaak dat die water vanuit die riviere nie geskik is vir gebruik in besproeiingspraktyke waar die riglyne vir fekale kolivorme en E. coli, soos uiteengesit deur die Suid-Afrikaanse owerhede, gereeld oorskry word nie. Besproeiing met fekale besoedelde rivierwater is een manier waarop groente en vrugte besmet kan word met voedsel patogene. Die risiko vir die oordrag van moontlike patogene teenwoordig in die besproeiingswater word beinvloed deur die mikrobiese lading teenwoordig en die getalle wat oorgedra word na die produkte. In hierdie studie is die oordrag en oorlewing van E. coli op groenboontjies, ertjies en tamaties onder beheerde laboratoriumtoestande beoordeel. Die produkte is blootgestel aan E. coli onder verskillende kombinasies van blootstellingstydperke (5, 15 of 30 min), drogingstye (30 of 120 min) en verskillende inokulum konsentrasies (107, 105, 104, 10³ en 10ª). In elke geval is . afname van ten minste een log waarde van die oorspronklike inokulum konsentrasie gevind. Die 10ª inokulum reeks het geen oorlewende selle getoon nie. Opvolg studies is gedoen met 'n blootstellingstyd van 60 min en verlengde drogingstye van 6 of 12 uur. Daar is gevind dat hierdie parameter veranderings die variasie in getalle vir dieselfde inokulum of die oorlewende E. coli nie beinvloed het nie. Soortgelyke E. coli ladings wat op die groenbone aangetref is, is ook op die ertjies aangetref. Die aantal oorlewendes op die tamaties was heelwat laer as die resultate vir die boontjies en ertjies. Dit word toegeskryf aan die "gladde" oppervlak van die tamatie-vel wat waarskynlik die aanhegting van die E. coli selle bemoeilik. Met die uitsondering van die 10ª inokulum reeks, het al die E. coli oorlewende selle gevind op die drie tipes produkte, die riglyn getalle oorskry wat vir vars produkte gestel is. As soortgelyke oorlewingspatrone gevind word in die omgewing, behoort die resultate van hierdie studie as waarskuwing te dien dat die water van die Plankenburgrivier ongeskik is vir die besproeiing van vars produkte. Sommige boere is reeds besig met chloor behandeling van die rivierwater voor besproeiing, om die kans van patogeenoordrag na die produkte uit te skakel. Dit is egter nie 'n haalbare oplossing vir die meeste boere nie as gevolg van die hoe kostes verbonde aan die implementering van' n stelsel soos hierdie.

On t.p.: Masters of Agricultural management. / Thesis (MAgricAdmin)--Stellenbosch University, 2002. / ENGLISH ABSTRACT: The need to take account of risk in agriculture must be part of every decision taken in agriculture. Yet risk is nothing to be too afraid of Risk is a choice rather than a fate. The actions we dare to take, which depend on how free we are to make choices, are what the theory of risk is all about. The task is rather to manage risk effectively, within the capacity of the farmer, business or group in order to withstand adverse outcomes. Some methods of managing risks are feasible for all types of farms. Others are only feasible for certain sizes and types of farms. Therefore, farmers in general need a systematic technique that will enable them to choose an efficient investment strategy from among all feasible strategies. Specifically, given n risky assets (such as the different enterprises in the PaarlIBerg River region), it is essential to seek a diversification strategy which yields a portfolio lying on the efficient frontier. The research question was whether different diversification models (Markowitz diversification model, Single Index Model and the Capital Asset Pricing Model) that are normally applied in capital markets for the construction of optimal diversified portfolios consisting out of different shares, are also applicable on risky portfolios in agriculture comprising different enterprises in the PaarlIBerg River region. The efficient frontier can be seen as the graphical representation of a set of portfolios that maximize expected return for each level of portfolio risk. The Microsoft Excel portfolio optimiser (SOLVER) programme was used to illustrate the investment proportions, expected returns, and standard deviations of the portfolios ofthe efficient frontier. The Single Index Model (SIM) can be used as an alternative to Markowitz diversification model. It drastically reduces the number of parameters needed to be estimated and yields the efficient set relatively easily without the technical difficulties characterising the fullrank solution. However, if the SIM assumptions are in contradiction to the actual data, the simplification of the calculations is achieved at the cost of getting imprecise results. The simplicity of SIM calculations was attained at a cost of constructing a sub-optimal portfolio, which does not lie on the corresponding efficient frontier. The Capital Asset Pricing Model (CAPM) reveals that there is a great deal of systematic risk in relation to the portfolio enclosed in this study. By using the CAPM it is possible to determine which part of the risk the producer can control (non-systematic risk) and which part the producer has no control over (systematic risk). The proportions of systematic risk that can be diversified away are small, relative to the total risk of the Farm Sector Portfolio. The success of these models depends on the efficiency of the market, as weU as a large, up-to-date and reliable data source. Many younger cultivars could not be included in this study, due to the limited availability of data. In the next few years as data become available, it will be possible to construct efficient frontiers out of a wider range of enterprises. Different enterprises and cultivars will increase the number of alternative uses for natural resources in the PaarlIBerg River region through diversification. This will result in more choices for the farmer, and more flexibility in the decision-making process. Without reliable data, the result will be "garbage in, garbage out." / AFRIKAANSE OPSOMMING: In elke besluit wat geneem word in landbou moet risiko as 'n faktor in ag geneem word. Tog is risiko nie iets om te vrees nie. Dit is eerder keuse as noodlot. Die stappe wat ons waag om te neem, wat afhang van hoe vry ons is om keuses te maak, is waaroor die teorie van risiko gaan. Die doel van die tesis is om risiko effektief te bestuur binne die vermoëns van die boer om sodoende negatiewe resultate die hoof te bied Sommige metodes van risikobestuur is lewensvatbaar vir alle soorte plase. Ander is slegs lewensvatbaar vir sekere groottes en tipes plase. Daarom benodig boere in die algemeen 'n tegniek wat dit vir hulle moontlik maak om 'n effektiewe beleggingstrategie te kies uit die verskillende uitvoerbare strategiee. Gegewe n as riskante aktiwiteite (soos die verskillende gewasse in die PaarllBergrivierstreek) is dit noodsaaklik om 'n diversifiseringstrategie te vind wat 'n portefeulje sal lewer wat raak aan die effektiewe grens. Die navorsingsvraag was of verskillende diversifiseringsmodelle (Markowitz diversifiseringsmodel (MVC), "Single Index Model" (SIM) en die "Capital Asset Pricing Model" (CAPM)) wat gewoonlik toegepas word in kapitaalmarkte vir die samestelling van optimale gediversifiseerde portefeuljes bestaande uit verskillende aandele, ook van toepassing sal wees op riskante portefeuljes in die landbou in die PaarlJBergrivierstreek, wat verskillende gewasse insluit. Die effektiewe grens kan gesien word as die grafiese voorstelling van 'n stel portefeuljes wat die verwagte winste vir elke vlak van portefeuljerisiko vermeerder. Die Microsoft Excel portefeulje optimeringsprogram (SOLVER) word gebruik om die beleggingsverhoudings, verwagte winste en standaardafwykings van die portefeuljes aan die effektiewe grens te illustreer. Die "Single Index Model" (SIM) kan gebruik word as 'n alternatief vir die Markowitz diversi:tikasiemodel. Dit verminder drasties die getal parameters en lewer maklik die effektiewe reeks, sonder die tegniese probleme wat ondervind word met die oplossing by die Markowitz model. Nietemin, indien die SIM die werklike data weerspreek sal die vereenvoudiging van die berekenings bereik word ten koste van onakurate resultate. Die eenvoud van die SIM is verkry ten koste van die samestelling van 'n suboptimale portfeulje, wat nie aan die ooreenstemmende effektiewe grens lê nie. Die "Capital Asset Pricing Model" (CAPM) wys dat daar baie sistematiese risiko gekoppel is aan die portfeulje ingesluit in hierdie studie. Deur gebruik temaak van die CAPM is dit moontlik om vas te stel watter deel van die risiko (nie-sistematies) die produsent kan beheer en watter deel die produsent nie kan beheer nie (sistematiese risiko). Die verhouding van sistematiese risiko wat weggediversifiseer kan word is klein in verhouding tot die algehele risiko van die boerderysektor portefeulje. Die sukses hang afvan die doeltreffendheid van die mark, sowel as 'n groot tot-op-datum en betroubare bron van data. Baie van die jonger aangeplante kultivars kan nie ingesluit word in hierdie studie nie as gevolg van beperkte data In die volgende paar jaar, soos data beskikbaar word, sal dit moontlik wees om effektiewe grense van 'n wye reeks gewasse saam te stel. Verskillende gewasse en kultivars sal die hoeveelheid alternatiewe gebruike van natuurlike hulpbronne in die PaarllBergrivierstreek vermeerder deur diversifikasie. Dit sal lei tot meer keuses vir die boer en meer buigsaamheid in die besluitnemingsproses. Sonder betroubare data kan betroubate resultate nie verkry word me.