The aim of this study was to use plants to determine the degree of heavy metal contamination in water and sediments in order to effectively monitor and provide possible recommendation to improve the water quality in the aquatic ecosystem of the Bottelary River.

Thesis (MSc Food Sc (Food Science))--University of Stellenbosch, 2010. / ENGLISH ABSTRACT: Agriculture in the Western Cape is not only one of the most important economic sectors but also provides many job opportunities. Over the last few years the sustainability of this successful industry has become threatened by the faecal pollution of rivers used to irrigate produce that will be consumed raw or after minimal processing. This situation not only poses an enormous risk to the health of the consumer but also to farmers who stand to lose their export licenses. The purpose of this study was to determine the microbial types and loads in river water, irrigation water and on irrigated produce. A baseline study was done on four sites in two Western Cape rivers. These sites were chosen to allow for the sampling of river water, irrigation water and irrigated produce so as to determine whether a link between the use of contaminated irrigation water and the microbial population found on irrigated produce exists. The physico-chemical analyses used in the study consisted of: pH, alkalinity, water temperature, conductivity and chemical oxygen demand. The microbial monitoring included the aerobic colony counts (ACC) and the enumeration of the total coliforms, faecal coliforms, staphylococci, enterococci, and aerobic and anaerobic sporeformers present in the water samples. The presence or absence of the potential pathogens like E. coli, Listeria and Salmonella, was also determined. During the baseline study faecal coliform counts as high as 160 000 organisms.100 mL-1 were noted in the Plankenburg River, while counts as high as 460 000 organisms.100 mL-1 were found in the Mosselbank River. Apart from this, high numbers of staphylococci and intestinal enterococci were often found, while E. coli, Listeria and Salmonella were present in samples from both of these rivers. Based on the results of the baseline study on the two rivers it was decided to do a more intensive study on the microbial load of the river and irrigation water as well as irrigated produce from the Mosselbank site. Lettuce and cabbages from a commercial farmer’s fields were chosen as the irrigated produce. During the warmer summer months, ACC counts in the river samples peaked at 12 8000 000 cfu.mL-1, while faecal coliform counts of 1 600 000 organisms.100 mL-1 were found. The three potential pathogens (E. coli, Listeria and Salmonella) were present in all the river samples taken during this period. While the counts of indicator bacteria in the irrigation water was often lower, faecal coliform counts as high as 1 600 000 organisms.100 mL-1 and several other potential pathogens were found on the irrigated lettuce and cabbage. This could indicate a possible “build-up” of contamination on the produce with the repeated application of the tainted irrigation water. According to guidelines published by DWAF in 2008, water to be used for irrigation should not contain more than 4 000 organisms.100 mL-1 faecal coliforms if it is used for the irrigation of crops that are to be consumed raw or after a minimal processing step, as this would increase the health risk to the consumer. Guidelines published by the South African Department of Health are even stricter and state that raw vegetables and fruit should not contain more than 200 coliform organisms per gram, while E. coli and L. monocytogenes should be absent in one gram, and Salmonella spp. in 25 grams of the produce, respectively. From the data obtained during this study it was evident that the two rivers monitored regularly contained faecal indicators at levels much higher than those proposed in national and international guidelines for safe irrigation, making them unfit for the irrigation of MPF’s. It could be concluded that the rivers investigated during this study contained high levels of faecal contamination. Since some of the pathogens isolated from the river and irrigation water and the irrigated produce, it suggests a carry-over of microbial contamination from the river water to the irrigated produce. This was, however, only done using the traditional international methods and the presence of specific pathogens should in future be confirmed by means of molecular techniques. / AFRIKAANSE OPSOMMING: Landbou is nie net die een van die belangrikste ekonomiese sektore in die Wes-Kaap nie, maar verskaf ook vele werksgeleenthede. Oor die afgelope paar jaar word die volhoubaarheid van hierdie suksesvolle industrie egter bedreig deur die fekale kontaminasie van riviere wat gebruik word vir die besproeiing van voedsel wat rou of na ‘n minimale prosesserings stap ingeneem word. Hierdie situasie hou nie net ‘n groot gevaar vir die gesondheid van verbruikers in nie, maar ook vir boere wat hul uitvoerlisensies hierdeur kan verloor. Die doel van hierdie studie was om die ladings en tipes mikrobes in rivier water, besproeiingswater en op besproeide produkte vas te stel. ‘n Basiese studie van vier liggings in twee Wes-Kaapse riviere is gedoen. Hierdie liggings is só gekies dat dit moontlik was om die rivier water, besproeiingswater en die besproeide produkte te monitor, en daar sodoende vasgestel kon word of daar ‘n verhouding is tussen die gebruik van gekontamineerde besproeiingswater en die mikrobe populasie wat op die besproeide produkte aanwesig was. Die fisiko-chemiese analises wat gedurende die studie gedoen is, het pH, alkaliniteit, water temperatuur, geleidingsvermoë en die chemiese suurstof vereiste (COD) ingesluit. Die mikrobiese analises het die aërobe kolonie tellings (ACC) en die enumerasie van die totale kolivorme, fekale kolivorme, staphylococci, enterococci en die aërobe en anaërobe spoorvormers ingesluit. Daar is ook vir die aanwesigheid van potensiële patogene soos E. coli, Listeria en Salmonella getoets. Gedurende die basiese studie is fekale kolovorme tellings van so hoog as 160 000 organismes.100mL-1 in die Plankenburg Rivier aangeteken, terwyl tellings van so hoog as 460 000 organismes.100mL-1 in die Mosselbank Rivier gevind is. Hoë tellings stafielokokki en intestinale enterokokki is gereeld genoteer, terwyl E.coli, Listeria en Salmonella uit die waters van beide hierdie riviere geïsoleer is. Gebaseer op hierdie resultate is daar besluit om ‘n meer intensiewe studie van die rivier, besproeiingswater en die besproeide produkte van die Mosselbank Rivier te doen. Blaarslaai en kool van ‘n kommersiële boer se lande is vir hierdie doel gekies. Gedurende die warmer somer maande het die aërobe kolonie tellings in die rivier ‘n piek van 12 800 000 kve.mL-1 bereik, terwyl fekale kolivorme tellings van 1 600 000 organismes.100mL-1 genoteer is. Die drie potensiële patogene (E. coli, Listeria en Salmonella) was aanwesig in al die monsters wat gedurende hierdie tydperk van die rivierwater geneem is. Alhoewel die tellings indikator bakterieë in die besproeiingswater meestal laag was, is tellings fekale kolivorme van so hoog as 1 600 000 kve.100mL-1 en verskeie ander potensiële patogene op die besproeide blaarslaai en kool gevind. Dit kan dui op ‘n moontlike opbou van kontaminasie op die produkte met die herhaalde besproeiing met gekontamineerde besproeiingswater. Volgens die riglyne wat in 2008 deur DWAF gepubliseer is, mag water wat vir die besproeiing van minimaal geprosesseerdevoedsels gebruik word nie meer as 4 000 organismes.100mL-1 bevat nie, aangesien dit die gesondheid van die gebruiker in gevaar mag stel. Die riglyne van die Suid-Afrikaanse Departement van Gesondheid is selfs strenger en beveel aan dat rou vrugte en groente nie meer as 200 kolivorme en geen L. monocytogenes per gram, en geen Salmonella spp. in 25 g van die produk mag bevat nie. Vanuit die data wat tydens hierdie studie ingesamel is, is dit duidelik dat die twee riviere gereeld fekale indikators bevat het teen vlakke baie hoër as wat in die nasionale en internasionale riglyne aanbeveel word. Hierdie water is dus nie geskik vir die besproeiing van minimaal geprosesseerde produkte nie. Die afleiding kan gemaak word dat die riviere wat tydens hierdie studie gemonitor is, hoë vlakke van fekale kontaminasie bevat het. Aangesien sommige van die patogene vanuit beide die rivier- en besproeiingswater, en vanaf die besproeide produkte geïsoleer is, kan dit dui op ‘n moontlike oordrag van mikrobiese kontaminasie vanuit die rivierwater na die besproeide produkte. Tydens hierdie studie is daar egter net van die tradisionele internasionale metodes gebruik gemaak. Vir toekomstige navorsing word dit aanbeveel dat die aanwesigheid van die spesifieke patogene deur die gebruik van molekulêre metodes bevestig word.

A growing human population accompanied by urbanisation and industrialisation have led to over exploitation and pollution of freshwater resources and have consequently impacted on aquatic ecosystem health. The Swartkops River in the Eastern Cape of South Africa is no exception. It drains a heavily industrialised catchment which has led to deterioration of its water quality due to pollution. Integrated water resources management (IWRM) requires the concurrent sustainable use of water resources and the protection of aquatic ecosystem health. Macroinvertebrates are well known for their ability to reflect the health of the environment in which they live, thus they were used to assess anthropogenic impacts in the Swartkops River for this study. Macroinvertebrate based biomonitoring approaches, including the South African Scoring System version 5 (SASS5); a multimetric approach involving 19 metrics; Chironomidae community assessments and screening of morphological deformities in Chironomidae larvae, were applied at four selected sampling sites to assess environmental water quality in the Swartkops River. Macroinvertebrates were sampled us ing the SASS5 protocols. Chironomidae were mounted and identified as far as practically possible using available keys. Mentum, ligula, mandible, paraligula and antenna in Chironomidae larvae were screened for deformities. Physical and chemical water quality variables were measured at each of the selected sampling sites. All data were subjected to relevant statistical analyses. Of the four sites sampled during the study period, results revealed that water quality at site 1 was the least impacted with highest SASS5 scores, average score per taxa (ASPT) values, richness, diversity, equitability and Ephemeroptera –Plecoptera-Trichoptera (EPT) richness, as well as least incidences of chironomid deformities. Water quality at site 2 was considered the next least impacted with higher SASS5 scores, A SPT values, richness, diversity and equitability, and lower incidences of deformities compared to sites 3 and 4. SASS5 scores and ASPT values revealed that both sites 3 and 4 were critically modified but the multimetric analysis, Chironomidae community assessment and incidences of deformities in Chironomidae larvae indicated that site 3 is the most impacted of the four sampling sites, with least species diversity, richness, equitability and highest incidences of deformities. The study revealed the importance of multicriteria approach to environmental biomonitoring as an integrated water resources management tool, and based on the results, site 3, as the most impacted, could be prioritised for restoration intervention.

This study assesses the presence of selected trace metals and pops in the Baynespruit River, Sobantu, KwaZulu-Natal. Sediments, plants and water samples were collected along the river during dry and wet season. For metal analysis, the extraction was done using acid digestion and then determined on the inductively coupled plasma atomic emission spectroscopy, for total metals in sediments, plants and water. Bio-available metals in sediments were analysed using atomic absorption spectroscopy. POPs were extracted in sediments using soxhlet extraction, which was followed up by evaporation on the rotary evaporator, sample clean up on a GPC column, further sample concentration on the rotary evaporator and later analysed on GC/MS instrument. Metals of concern in both seasons were Cu, Cr, Pb and Zn which exceeded maximum permissible limits for South African soil, Cr was just about the maximum permissible limit. The targeted POPs were not detected on any site during this work. / Environmental Sciences / M. Sc. (Environmental Science)

Gold mining in the Witwatersrand Basin of South Africa has resulted in soil contamination due to the lack of sufficient environmental management plans for the tailings dumps and remnant footprints. Tailings reclamation as a strategy of reducing pollution in the Central Rand, for instance, has resulted in contamination of water systems by acid mine drainage (AMD). After removal of the tailings dumps, remnant material is left over on the tailings footprints and these contain significant amounts of pollutants that were initially in the tailings. Heavy rainfall during summer dissolves primary minerals and later in the dry season, secondary minerals are precipitated as efflorescent crusts on and nearby tailings dumps as well as footprints due to high evaporation. The efflorescent crusts can redissolve when it rains and form acidic, metal and sulphate-rich solutions due to their soluble characteristics. This study aimed to characterise tailings footprints in areas targeted for human settlements and office spaces to assess their potential to release left over toxic elements such arsenic (As), lead (Pb), copper (Cu) and zinc (Zn). The approach to the study involved characterisation of oxidised and unoxidised tailings material and secondary precipitates on both tailings dumps and footprints. This involved determining the mineralogical composition using Powder X-ray Diffraction (PXRD). Dissolution and leaching studies were also conducted on the material followed by determination of constituent elements using inductively coupled plasma optical emission spectroscopy (ICP-OES) and sulphates using ion chromatography (IC). The leaching solutions used included rainwater; dilute sulphuric acid at pH of 3.0 (a common leachate in such acidic soils); as well as plant exudates such as oxalic and citric acids. The leachate solutions were used to correlate the mineralogical composition of secondary precipitates and tailings footprints. Potential implications on humans following any accidental ingestion of the tailings or contaminated soils were assessed using gastric juices. The ecological risk factors and risk index together with the model to evaluate daily intake and different pathways to humans were used to assess the toxicity caused by exposure to contaminants in the materials. The experimental work was augmented by computer simulations based on geochemical modelling (using the PHREEQC geochemical modelling code) to determine the speciation of elements (and thus their potential lability and bioavailability), dissolution and formation of secondary mineral precipitates in the tailings dumps and footprints. The findings of the PXRD study showed that the mineralogy of the tailings and footprints was dominated by quartz (SiO2) and some minor minerals such as pyrite (FeS2), pyrophyllite (Al2Si4O10(OH)2), chlorite (Mg,Fe)3(Si,Al)4O10), mica (K(Mg,Fe)3 AlSi3O10(F,OH)2) while that of secondary precipitates was dominated by jarosite (KFe3+ 3(OH)6 (SO4)2), goethite (FeOOH), melanterite (FeSO4.7H2O) and gypsum (CaSO4.2H2O). Minerals obtained for the secondary precipitates were corroborated by geochemical modelling. Leaching results using rainwater with pH ranges from 3.5 to 3.9 showed that trace elements are released very slowly from tailings dumps and footprints and in small concentrations during rainy seasons as follows: As (1.5 mg/L-4.5 mg/L), Pb (3.5 mg/L-5.5 mg/L), Cu (4 mg/L-4.8 mg/L) and Zn (23 mg/L-44 mg/L). The release and mobility of Cu, Pb, Zn and As occurs quite markedly when secondary precipitates dissolve, making the immediate impacted environment unfavourable for plant growth and any habits in the vicinity. This was substantiated by simulated dissolutions and assessment of the resulting elemental speciation that pointed to the elements being distributed in bioavailable forms, implying potential uptake by plants (such as vegetables that may be cultivated on such impacted soils). The model was used to evaluate the daily intake and different exposure pathways and the results showed that children may daily intake 48.4 mg kg-1 day-1 and adults‟ 32.8 mg kg-1 day-1 . After 5 years (1825 days) of exposure more harm may be experienced and findings shows that kids are the most victims on these contaminated sites compared to adults. Both children and adults may absorb low levels of these toxic elements daily and after long time of exposure it may cause disease like cancer in their body which may lead to death. Pathways may be through inhalation and accidentally ingesting tailings soil that contain toxic elements. Drawing from the above findings, it will be important that tailings footprints that have been earmarked as land for development (residential or office space) be thoroughly assessed for potential release of toxic elements and high levels of acidity. Further reclamation aimed at reducing these hazards can then be implemented. / College of Agriculture and Environmental Sciences / M. Sc. (Environmental Science)

The study aimed at investigating the factors that affect the efficient management of wastes produced by food outlets in Pretoria, South Africa. This was with a view of providing empirical data and information with respect to the causative factors responsible for the visible environmental pollution by wastes and materials from food outlets. Data was gathered from a representative sample of 332 food outlets on socio-economic, demographic, environmental, sanitary and health related variables that are related to waste management and efficiency in the disposal of solid and liquid wastes. Data entry and analysis was performed using statistical procedures such as frequency tables, Pearson’s chi-square tests of association and binary logistic regression analysis in the statistical package STATA version 10. Odds ratios estimated from logistic regression analysis were used for identifying key factors that affect efficiency in the proper disposal of waste. Results obtained showed that 18 % of the 332 food outlets in the study were generally inefficient in waste disposal. Based on odds ratios estimated from binary logistic regression analysis, wrong perception (a factor of 10.88), failure to provide trash cans to customers (a factor of 3.15), the operation of food outlets by non-owners or managers (a factor of 2.33), and failure to practice source reduction of waste (a factor of 2.25) are the top 4 influential factors that affect the proper management of waste at the 332 food outlets in the study. / Environmental Sciences / M. Sc. (Environmental Management)

Surface water pollution is prevalent in numerous areas of central Roodepoort mainly due to gold mining activities. The surface water quality for the Bosmontspruit, Russell’s Stream and the New Canada Dam was assessed from October 2010 to March 2011. Physical, chemical and biological characteristics of the water were determined for 8 monitoring points and the results obtained were compared with the In-stream water quality guidelines for the Klip River catchment and the South African Water Quality Guidelines. A trend noticed throughout the sampling period was the non-compliance in the levels of total dissolved solids (TDS) and dissolved oxygen. The results indicated that concentrations of iron, aluminium, nickel, manganese and potassium were above the limit across the Bosmontspruit and Russell’s stream. There was also significant evidence of excessive faecal coliform and ammonium pollution in the Bosmontspruit. During the monitoring period it was noted that water from these streams were utilised for crop irrigation, bathing, livestock and human consumption and may pose a health hazard due to poor water quality. / Environmental Sciences / M.Sc. (Environmental Science)

Car washing consumes large quantities of water and results in large amounts of wastewater effluent being generated, but has received little attention as a potential source of water pollution globally. The study investigated the water use, wastewater effluent quality and toxicity of selected carwashes in City of Tshwane. Ten carwashes in the City of Tshwane were purposively selected and analysis done on a variety of water quality aspects including microbiological, toxicity, BOD,COD, oil and grease, anionic surfactants, sulphates, phosphates and heavy metals. The results obtained show that oil and grease exceeded the WHO and National Water Act standards of 2.5mg/L in all sites except CW2 and Diesel range organics were significantly high in some sites (CW4, 7 & 8) at 60.5, 40 and 48.8mg/L. COD and BOD of the waste water exceeded the WHO standards of 30mg/L and 60mg/L in all sites except CW2. Bacteria contamination was very high in all wastewater samples and all samples exhibited (mild to high) toxicity to Vibrio fischeri and high toxicity to Daphnia pulex except CW2. These results suggest that wastewater from carwashes is high in some pollutants that may interfere with the receiving environment and municipal treatment systems and there is a need for responsible authorities to regulate the quality of effluent discharged into the environment / Environmental Sciences / M.Sc. (Environmental Science)

Thesis submitted in fulfilment of the requirements for the degree Doctor of Technology: Biomedical Technology in the Faculty of Health and Wellness Sciences at the Cape Peninsula University of Technology 2008 / The quality of South Africa’s water sources is fast deteriorating due to an influx of pollutants from industrial and agricultural areas. In addition, urbanisation has led to the establishment of informal settlements along river systems. This study focuses on the importance of maintaining water quality and the management of water resources in order to ensure its sustainability in South Africa. The primary aim of this study was to determine the extent of bacterial contamination in three rivers namely the Berg-, Plankenburg- and Diep Rivers in the Western Cape, South Africa and to investigate the application of a bioremediation system as a possible treatment technology. Several aspects contributing to the contamination were addressed and different approaches were studied and reviewed. In all three rivers, four sampling sites were identified, which were sampled over a period of 9 to 12 months. Contamination levels for the three rivers were evaluated by applying various enumeration techniques, which could provide an accurate indication of the planktonic bacterial pollution load in the river systems. The Most Probable Number (MPN) technique was used to determine the level of faecal coliforms and E. coli. The highest MPN, faecal coliform and E. coli counts of 3.5 x 107 micro-organisms/100 m , 3.5 x 107 micro-organisms/100 m and 1.7 x 107 micro-organisms/100 m , respectively, were recorded at Site B2 in week 37 in the Berg River. Results showed that in all the river water sampled and evaluated, the total MPN count mostly exceeded the maximum limit of 2000 micro-organisms/100 m (SABS, 1984) stipulated for river water throughout the study period. The heterotrophic plate count (HPC) method was used to determine the number of culturable micro-organisms in planktonic samples, while the flow cytometry (FCM) and epifluorescence microscopy (EM) with different fluorochromes (Acridine orange and BacLight™ Live/Dead stain) were employed to evaluate total bacterial counts in planktonic (water) samples. The highest HPC at the various sites sampled was 1.04 x 106 micro-organisms/m (Berg River, Site B2), 7.9 x 104 micro-organisms/m (Plankenbrug River, Site A) and 1.7 x 105 micro-organisms/m (Diep River, Site B). Total cell counts as high as 3.7 x 107 micro-organism/m (Berg River, Site B2), 5.5 x 108 micro-organism/m (Plankenburg River, Site D) and 2.5 x 109 micro-organisms/m (Diep River, Site B) were obtained by the FCM technique, which were significantly (p < 0.05) higher than the total counts obtained by epifluorescence microscopy. The results thus show that the FCM technique was the most reliable method for determining the total cell count in river water samples. This technique makes use of computer software whereas epifluorescence microscopy involves manual counting which may lead to human error. In addition, the impact of residential, agricultural and industrial areas situated along these rivers was also investigated. Even though exact point sources of pollution could not be determined, it was found that all the sources, such as the storm water drainage pipes, the industrial as well as the agricultural areas, could contribute to increased MPN, heterotrophic and total bacterial counts. This study also aimed at investigating and comparing the microbial contamination levels at various sites in the Plankenburg and Diep Rivers in the Western Cape, South Africa. Sampling of sites along the Plankenburg River started in June 2004 and continued for a period of one year until June 2005. Sampling of the Diep River sites started in March 2005 and continued for a period of nine months until November 2005. Faecal coliform (FC) and E. coli (EC) counts were determined by means of the Most Probable Number technique, the number of culturable cells were determined using the heterotrophic plate count (HPC) technique and total microbial counts were evaluated by Flow cytometric analysis (FCM). The highest microbial counts for the Plankenburg River were observed at site B where the highest MPN, FC, E. coli and total FCM counts of 9.2 x 106 (week 14), 3.5 x 106 (week 39) and 3.5 x 106 micro-organisms/100 m (week 39) and 2.1 x 108 micro-organisms/m (weeks 1 and 39) respectively, were recorded. The highest HPC recorded for the Plankenburg River was 7.9 x 106 micro-organisms/100 m (week 44, site A). Site B is situated close to an informal settlement where waste effluents from storm water drainage pipes enter the river system. In addition, other possible contamination sources included agricultural (site A) and industrial (site C) areas bordering the Plankenburg River. The highest total MPN, FC and E. coli counts in the Diep River were 5.4 x 106 (week 23) and 1.6 x 106 micro-organisms/100 m [FC and E. coli, respectively (both in week 23)], recorded at site B. The highest HPC and total FCM counts of 1.7 x 107 micro-organisms/100 m (week 14) and 2.5 x 109 microorganisms/ m (week 23), respectively, were also recorded at site B. This site was identified as the most contaminated site along the Diep River and served as an accumulation point for waste effluents from the residential and industrial areas, which included paint and machine manufacturers. Other sources situated along the Diep River included storage and maintenance facilities for steel containers, a waste water treatment plant and an oil-refinery. Most of the bacterial counts obtained for the Plankenburg and Diep Rivers exceeded the accepted maximum limit for river water for most of the sampling period. Bacterial species from the Berg- and Plankenburg Rivers were isolated and identified. The presence of various Enterobacteriaceae species isolated at all the sites in both rivers confirmed faecal contamination of these water sources over the entire sampling period. Opportunistic pathogens such as Klebsiella sp., Serratia sp., Enterobacter sp., Shewanella sp., Aeromonas sp., Pseudomonas sp., Acinetobacter sp. and Citrobacter freundii as well as pathogens such as Bacillus cereus and B. anthracis were also identified in both river systems. All the respective articles are presented in the required format of the journal in which the article has been published or submitted to.

Thesis submitted in fulfilment of the requirements for the degree Doctor of Technology: Biomedical Technology in the Faculty of Health and Wellness Sciences at the Cape Peninsula University of Technology 2008 / River systems can become contaminated with micro-organisms and metals and the routine monitoring of these rivers is essential to control the occurrence of these contaminants in water bodies. This study was aimed at investigating the metal contamination levels in the Berg-, Plankenburg- and Diep Rivers in the Western Cape, South Africa, followed by the remediation of these rivers, using bioreactor systems. Sampling sites were identified and samples [water, sediment and biofilm (leaves, rocks and glass, etc.)] were collected along the Berg- and Plankenburg Rivers from May 2004 to May 2005 and for the Diep River, from February 2005 to November 2005. The concentrations of aluminium (Al), copper (Cu), iron (Fe), manganese (Mn), nickel (Ni), lead (Pb) and zinc (Zn) were determined using the nitric acid digestion method and analysed by Inductively Coupled Plasma Atomic Emission Spectrometry (ICP-AES). For the Berg River, the highest concentrations in water samples were recorded for Al, Mn and Fe at the agricultural area (Site A – chapter 2). In the sediment and biofilm samples, the highest metal concentrations were once again recorded for Al and Fe. The concentrations of Al and Fe were significantly higher (p < 0.05) than than Cu, Zn, Pb, Ni and Mn in water, sediment and biofilm samples, and were mostly higher than the quality guidelines recommended by the Department of Water Affairs and Forestry (DWAF, 1996) and the Canadian Council for the Ministers of the Environment (CCME, 2001). Possible sources of contamination in the Berg River could be due to the leaching or improper discarding of household waste from the informal- and established residential areas, as well as the improper discarding of pesticides at the agricultural area. For both the Plankenburg and Diep Rivers the Al and Fe concentrations were higher than all the other metals analysed for in sediment and water samples. The highest concentrations recorded in the Plankenburg River was 13.6 mg.l-1 (water - Week 18, Site B) and 15 018 mg.kg-1 (sediment - Week 1, Site C) for Al and 48 mg.l-1 (water - Week 43, Site A) and 14 363.8 mg.kg-1 (sediment - Week 1, Site A) for Fe. The highest concentrations recorded in the Diep River was 4 mg.l-1 (water - Week 1, Site A) and 19 179 mg.kg-1 (sediment - Week 1, Site C) for Al and 513 mg.l-1 (water - Week 27, Site A) and 106 379.5 mg.kg-1 (sediment - Week 9, Site C) for Fe. For most of the metals analysed the concentrations were higher than the recommended water quality guidelines as stipulated by the Department of Water Affairs and Forestry (DWAF, 1996b), the Canadian Council for the Ministers of the Environment (CCME, 2001) and the ‘World average’ (Martin and Windom, 1991). Point sources of pollution could not conclusively be identified, but the industrial and residential areas could have influenced the increased concentrations. Metal concentrations should be routinely monitored and the guidelines should be updated and revised based on the current state of the rivers and pollution influences. Micro-organisms isolated from flow cells after exposure to varying metal concentrations were investigated for possible metal-tolerance. A site where high metal concentrations were recorded along the Plankenburg River was investigated. The micro-organisms isolated from the flow cells were cultured and identified using the Polymerase Chain Reaction (PCR) technique, in conjunction with universal 16SrRNA primers. The phylogeny of the representative organisms in GenBank, were analysed using the Neighbour-joining algorithm in Clustal X. After exposure, the channels were stained with the LIVE/DEAD BacLightTM viability probe and visualised using Epifluorescence Microscopy. The results revealed that when exposed to the highest concentrations of Al (900 mg.l-1), Fe (1000 mg.l-1), Cu (10 mg.l-1) and Mn (80 mg.l-1), the percentage of dead cells increased, and when exposed to the lowest concentrations of Al (10 mg.l-1), Cu (0.5 mg.l-1), Mn (1.5 mg.l-1) and Zn (0.5 mg.l-1), no significant differences could be distinguished between live an dead cells. When exposed to the highest concentrations of Zn (40 mg.l-1) and Ni (20 mg.l-1), no significant differences between the live and dead cell percentages, were observed. The phylogenetic tree showed that a diverse group of organisms were isolated from the flow cells and that some of the isolates exhibited multiple metal resistance (Stenotrophomonas maltophilia strain 776, Bacillus sp. ZH6, Staphylococcus sp. MOLA:313, Pseudomonas sp. and Delftia tsuruhatensis strain A90 exhibited tolerance to Zn, Ni, Cu, Al, Fe), while other isolates were resistant to specific metals (Comamonas testosteroni WDL7, Microbacterium sp. PAO-12 and Sphingomonas sp. 8b-1 exhibited tolerance to Cu, Ni and Zn, respectively, while Kocuria kristinae strain 6J-5b and Micrococcus sp. TPR14 exhibited tolerance to Mn). The efficiency of two laboratory-scale and one on-site bioreactor system was evaluated to determine their ability to reduce metal concentrations in river water samples. The laboratory-scale bioreactors were run for a two-week and a three-week period and the on-site bioreactor for a period of ten weeks. Water (all three bioreactors) and bioballs (bioreactor two and on-site bioreactor) were collected, digested with 55% nitric acid and analysed using ICP-AES. The final concentrations for Al, Ni and Zn (bioreactor one) and Mn (bioreactor two), decreased to below their recommended concentrations in water samples. In the on-site, six-tank bioreactor system, the concentrations for Fe, Cu, Mn and Ni decreased, but still exceeded the recommended concentrations. The concentrations recorded in the biofilm suspensions removed from the bioballs collected from bioreactor two and the on-site bioreactor, revealed concentrations higher than those recorded in the corresponding water samples for all the metals analysed, except Fe. The bioballs were shown to be efficient for biofilm attachment and subsequent metal accumulation. The species diversity of the organisms isolated from the bioreactor (bioreactor two) experiment after three days (initial) differed from the organisms isolated after 15 days (final). Hydrogenophaga sp., Ochrobactrum sp, Corynebacterium sp., Chelatobater sp. and Brevundimonas sp. were present only at the start of the bioreactor experiment. The surviving populations present both in the beginning and at the end of the bioreactor experiment belonged predominantly to the genera, Pseudomonas and Bacillus. Metal-tolerant organisms, such as Bacillus, Pseudomonas, Micrococcus and Stenotrophomonas, amongst others, could possibly be utilised to increase the efficiency of the bioreactors. The bioreactor system should however, be optimised further to improve its efficacy.