A GEOMORPHOLOGICAL 
INVESTIGATION OF THE KLIP RIVER 
WETLAND, SOUTH OF 
JOHANNESBURG 
by 
Vanessa Vermaak 
(0107624V) 
A dissertation submitted in partial fulfilment of the 
requirements for the degree of 
M.Sc. in Environmental Sciences 
The University of the Witwatersrand 
 
2009 
Supervisor: Prof. Terence McCarthy 
 
Declaration  
 
I declare that the dissertation which is being submitted for the Degree of Master of Science 
at the University of the Witwatersrand is my own, unaided work and has not been 
previously submitted by me at another University for any degree.  
 
 
 
______________________ 
Vanessa Mary Vermaak 
December 2009 
 i 
A Geomorphological Investigation of the Klip River Wetland, South of 
Johannesburg 
 
Abstract 
The Klip River Wetland is located south of Johannesburg. The city is a major conurbation 
which developed as a result of mining activity. The wetland has, and always had, economic 
importance to the region, firstly as a water source and later as a purifier of polluted water 
reporting from the urban-mining-industrial complex. A previous investigation determined that 
this wetland is in an advanced state of collapse. A network of irrigation canals dug in the 
early part of the 20th century to support agriculture in the area, provided a preferred 
discharge conduit for the sewage pumped into the Klip River. These irrigation canals have 
cut back and avulsed forming an almost continuous channel, with little water flowing through 
the wetland. The wetland?s capability to remove contaminants from the water is therefore 
compromised.  
 
The geomorphological study presented here aims at the quantification and the better 
understanding of:  
? The extent of the degradation 
? The rate of change to the wetland and 
? To offer potential remedial solutions.  
 
Desktop analysis was conducted on the following data: 
? Aerial photography (1938, 1952, 1961, 1984, 2003, 2006) 
? Historical maps from 1917 
? High resolution topography data (LIDAR),  
? Water quality 
? Water discharge  
 
In order to quantify the degradation of the wetland and associated rates, the following 
factors were considered:  
? Wetland area and shape 
? Channel formation, propagation and widening  
 ii 
? Knick point migration.  
 
The major findings of this study are: 
? The analysis of wetland area did not prove as effective a method as the study of the 
channels and knick points. 
? Two major knick points are distinguished. The first is located at the outlet of the 
Olifantsvlei Sewage Works and the second represents a major rapid near Kromvlei. 
The latter has incised 5 m into the wetland. 
? Knick point migration was not detected between 2003 and 2006, except for the rapid 
near Komvlei where the knick point migrated a significant 118 m in two years. 
? Between Kibler Park and Kromvlei the wetland experienced the greatest head cut 
advance of all the sections, most significantly in the periods 1952 to 1961 and 2006 
to 2008. 
? The period of the most significant change for all channel width monitoring points was 
between 2001 and 2003. 
 
Results of this investigation confirm that the current state of the Klip River Wetland is dire. 
Large tracts of wetland have disappeared over the last 100 years. The wetlands that remain 
are not functioning optimally. The importance of the wetland to purify the sewage, industrial 
and mining effluent is paramount. The window of opportunity to save or restore at least 
parts of the wetland, and therefore maintain the purification potential of the wetland, is 
closing. Peat burns will lead to erosion of the soils and therefore the reworking of the 
pollutants into the river system. It is anticipated that if the Klip River Wetland collapses 
completely, the water quality will worsen along the Vaal River and in the Vaal Barrage, 
which will impact negatively on the downstream users.  
 
The proposed solution to protect the wetland from further erosion and head-cut advance is 
to force the water out of the channels and into the wetland. By building a number of weirs in 
the channels, it is possible to dam the water upstream of the weir, thereby forcing the water 
to flood the wetland again. 
 
 iii 
A Geomorphological Investigation of the Klip River Wetland, South of 
Johannesburg 
 
Table of Contents  
 Page 
Abstract ............................................................................................................................... i 
Table of Contents.............................................................................................................. iii 
List of Figures.....................................................................................................................v 
List of Tables....................................................................................................................viii 
1. Introduction ....................................................................................................................1 
2. Aims and Objectives ......................................................................................................2 
3. Literature Review ...........................................................................................................3 
3.1 Klip River Catchment Description ..........................................................................3 
3.1.1 Regional Setting............................................................................................3 
3.1.2 River Systems and Water Resources ...........................................................3 
3.1.3 Climate, Hydrology and Geohydrology .........................................................4 
3.1.4 Vegetation.....................................................................................................9 
3.1.5 Wetland Characteristics and Status ..............................................................9 
3.1.6 Land Uses...................................................................................................11 
3.2 The History of the Klip River Catchment ..............................................................11 
3.3 Water Pollution Sources: Mining ..........................................................................14 
3.3.1 Source of Acid Mine Drainage ....................................................................18 
3.3.2 Transport of Acid Mine Drainage ................................................................19 
3.3.3 Bioaccumulation and the Toxicity of Oxidation Products ............................20 
3.4 Water Pollution Sources: Waste Water................................................................22 
3.5 Wetlands and their Importance ............................................................................29 
4. Methodology.................................................................................................................33 
5. Results .........................................................................................................................35 
5.1 Aerial Photography...............................................................................................36 
5.1.1 Section 1: Lenasia to Olifantsvlei Water Treatment ....................................42 
5.1.2 Section 2: Golden Highway to the N1 .........................................................47 
5.1.3 Section 3: N1 to Kibler Park ........................................................................50 
5.1.4 Section 4: Kibler Park to Kromvlei...............................................................54 
 iv 
5.1.5 Section 5: Zwartkoppies Section.................................................................61 
5.2 Long River Profiles...............................................................................................68 
5.2.1 Section 1: Lenasia to Olifantsvlei Water Treatment ....................................69 
5.2.2 Sections 2 and 3: Golden Highway to Kibler Park ......................................71 
5.2.3 Sections 4 and 5: Kibler Park to Zwartkoppies ...........................................74 
5.3 Digital Terrain Model ............................................................................................79 
5.4 Water Chemistry ..................................................................................................81 
5.5 Water Discharge ..................................................................................................84 
6. Discussion and Conclusions ........................................................................................88 
6.1 Wetland Degradation ...........................................................................................88 
6.2 Rate of Wetland Degradation...............................................................................90 
7. Recommendations .......................................................................................................94 
Reference List ..................................................................................................................97 
Acknowledgements........................................................................................................101 
Appendix 1: Digital Aerial Photography and LIDAR Survey Report 
Appendix 2: LIDAR Data Test 
 v 
A Geomorphological Investigation of the Klip River Wetland, South of 
Johannesburg 
 
List of Figures 
  Page 
Figure 1.1: Location of the Klip River Wetland in relation to industrial, urban and         
mining development in the Johannesburg area. ...............................................1 
Figure 3.1.1: Location of the river system and water resources around the Klip River 
Wetland. ............................................................................................................4 
Figure 3.1.2: Surface geology of the Klip River Catchment. ...............................................6 
Figure 3.1.3: Lithostratigraphy of the Malmani Subgroup. ..................................................7 
Figure 3.1.4: Aeromagnetic anomaly map with dykes inferred. ..........................................8 
Figure 3.1.5: Aeromagnetic map with dykes projected onto Central Rand Goldfield. .........8 
Figure 3.1.6: Open water bodies left from peat mining east of Klipspruit Valley Road. ....10 
Figure 3.1.7: Open water bodies and scars left from peat mining.....................................11 
Figure 3.2.1: The Klip River in relation to the Vaal Barrage and the Vaal Dam. ...............13 
Figure 3.3.1: Acid Mine Drainage during 1938 impacting the Klip River Wetland.............15 
Figure 3.3.2: Total Dissolved Solids concentration map of the Klip River Basin...............16 
Figure 3.4.1: Discharge point of the Olifantsvlei sewage treatment plant ponds. .............23 
Figure 3.4.2: Discharge point of the Bushkoppies sewage treatment plant ponds. ..........24 
Figure 3.4.3: Algal blooms seen at the Hartebeespoort Dam wall. ...................................28 
Figure 3.5.1: Klip River catchment showing distribution of a) pH b) conductivity, c) iron    
and d) sulphate................................................................................................31 
Figure 4.1: Location of monitoring stations along the Klip River. ......................................34 
Figure 5.1.1: The wetland divided into five sections..........................................................35 
Figure 5.1.2: Landsat image showing the water and moisture content of the wetland. ....36 
Figure 5.1.3: Burns in the wetland downstream of the moist section of wetland...............37 
Figure 5.1.4: Locality of channel width measurements. ....................................................39 
Figure 5.1.5: Rate and average rate of change of channel width over time......................41 
Figure 5.1.6: The reed-covered swamp in the Lenasia area.............................................42 
Figure 5.1.7: Gypsum crust accumulated on the banks of the Klip River Wetland at  
Lenasia............................................................................................................43 
 vi 
Figure 5.1.8: The wetland in the Lenasia area..................................................................45 
Figure 5.1.9: Point A: Channel cut after the Olifantvlei discharge point. ...........................46 
Figure 5.1.10: Study site of Arnold (1980) ........................................................................46 
Figure 5.1.11: The wetland between the Golden Highway and the N1.............................48 
Figure 5.1.12: Detail of the wetland at the N1 Highway. ...................................................50 
Figure 5.1.13: The wetland between the N1 Highway and Kibler Park.............................51 
Figure 5.1.14: Field and aerial reconnaissance photos between the N1 Highway and 
Lenasia............................................................................................................53 
Figure 5.1.15: The breakdown of Section 4 as to be discussed from Jackson?s Drift to 
Kromvlei. .........................................................................................................54 
Figure 5.1.16: The wetland at Jackson?s Drift Bridge........................................................55 
Figure 5.1.17: The wetland south of Kibler Park. ..............................................................57 
Figure 5.1.18: Diffuse channels as seen in an aerial survey eroding upstream of the     
drain. ...............................................................................................................58 
Figure 5.1.19: The wetland downstream of Kibler Park. ...................................................60 
Figure 5.1.20: The wetland of Section 5. ..........................................................................62 
Figure 5.1.21: Wetland retreat from the Klip River Railway Station between 1910 and  
2003. ...............................................................................................................63 
Figure 5.1.22: Wetland retreat from the Klip River Railway Station between 1910 and  
1961. ...............................................................................................................64 
Figure 5.1.23: The wetland at Zwartkoppies Pump Station. .............................................66 
Figure 5.1.24: The wetland south of the Zwartkoppies Pump Station...............................66 
Figure 5.1.26: A 2 m high waterfall on bedrock.................................................................67 
Figure 5.1.25: Aerial photograph of the final stages of the formation of a single          
channel in the Zwartkoppies section of the wetland........................................67 
Figure 5.2.2: Long river profile from the road east of the golf course to the Golden   
Highway in relation to 2006 photography. .......................................................70 
Figure 5.2.3: Long river profile from the Golden Highway to the N1 in relation to 2006 
photography. ...................................................................................................72 
Figure 5.2.4: Long river profile from the N1 Highway to Kibler Park in relation to 2006 
photography. ...................................................................................................73 
Figure 5.2.5: Field and aerial reconnaissance photos of the rapids between the N1  
Highway and Lenasia......................................................................................74 
 vii 
Figure 5.2.6: Long river profile from the Kibler Park to Zwartkoppies Pump Station in 
relation to 2006 photography...........................................................................76 
Figure 5.2.7: Knick point migration between 2003 and 2008 at Point 16..........................77 
Figure 5.2.8: Field photos of the knick point eroding through soil at Point 16...................78 
Figure 5.3.1: DTM derived from the contour LIDAR data at 40 m resolution. ...................79 
Figure 5.3.2: DTM derived from the original LIDAR point data at 10 m resolution............80 
Figure 5.3.3: Hillshade at 5? sun angle derived from the DTM at 10 m resolution. ...........81 
Figure 5.4.1: K21 phosphate levels from 1992 to 2006. ...................................................82 
Figure 5.4.2: K21 sulfate levels from 1992 to 2006...........................................................83 
Figure 5.4.3: K21 nitrate levels from 1992 to 2006. ..........................................................83 
Figure 5.4.4: K21 dissolved oxygen levels from 1992 to 2006..........................................84 
Figure 5.5.1: Mean wet and dry season trends for water discharge data at Witkop      
Station (C2H141) between 1977 and 2006.....................................................85 
Figure 5.5.2: Water discharge in February at Witkop Station (C2H141) between            
1979 and 2006. ...............................................................................................86 
Figure 5.5.3: Water discharge in June at Witkop Station (C2H141) between 1977            
and 2006. ........................................................................................................86 
Figure 5.5.4: Water discharge at Witkop Station (C2H141) between 1977 and 2006 
compared to rainfall data (OR Tambo Airport). ...............................................87 
Figure 5.5.5: Water discharge at Witkop Station (C2H141) per month compared to     
rainfall data (OR Tambo Airport). ....................................................................88 
Figure 6.1: Long river profile from Lenasia to the Zwartkoppies Pump Station.................90 
Figure 6.2: Wetland area in (a) 1917, (b) 1938, (c) 1961, and (d) 2006............................91 
Figure 6.3: Channel formation, propagation and widening from the Golden Highway to 
Kibler Park between 1917 and 2003. ..............................................................92 
Figure 6.4: Channel formation, propagation, widening and knick point migration from   
Kibler Park to the Zwartkoppies Pump Station between 1917 and 2006. .......93 
Figure 7.1: Proposed weir locations according to dyke positions......................................96 
 
 viii 
A Geomorphological Investigation of the Klip River Wetland, South of 
Johannesburg 
 
List of Tables 
  Page 
Table 3.4.1: Method to determine trophic status statistics.................................................27 
Table 3.4.2: The Trophic Status classification of the Vaal Dam and the Vaal Barrage. ....28 
Table 5.1.1: Change in wetland area over time.................................................................37 
Table 5.1.2: Channel widths over time along the length of the Klip River Wetland. ..........39 
Table 5.1.3: Rates of change for channel widths for Points A to F....................................41 
Table 5.1.4: Advance rate of the channel south of the Zwartkoppies Pump Station. ........65 
Table 6.1: Rates of head cut advance for Sections 2 to 5.................................................94 
 1 
A Geomorphological Investigation of the Klip River Wetland, South of 
Johannesburg 
 
1. Introduction 
The Klip River Wetland is situated south of the Witwatersrand conurbation in the most 
densely populated province of South Africa, Gauteng (Figure 1.1). It lies between latitudes 
26?10? and 26?25? south and longitudes 27?45? and 28?05? east, at an altitude of about 1750 
m above mean sea level.  
 
 
Figure 1.1: Location of the Klip River Wetland in relation to industrial, urban and mining 
development in the Johannesburg area.
 2 
The Witwatersrand is one of the few great conurbations (population of about 11 million 
people) of the world that is not located on a major river, lake or seashore (Turton et al., 
2006). The existence of this city is solely due to the discovery of gold-bearing reefs of the 
Central Rand Group, which outcrop at surface as part of the greater Witwatersrand 
Goldfield (Figure 1.1). Gold mining began on the Witwatersrand in 1886 and it soon 
became evident that the gold occurrence was of considerable significance. Prospectors 
and miners flocked to the area and townships were established along the length of the 
outcrop. As the mining industry grew, secondary industries were established on the 
Witwatersrand, creating a vast mining?industrial complex. Most of this mining and 
industrial development took place along the southern flank of the Witwatersrand Ridge, 
which forms a major watershed that separates the Vaal River Basin (part of the Orange 
River) to the south from the Limpopo River Basin to the north. The Witwatersrand mining-
 industrial complex is therefore located at the headwaters of two major international river 
basins, the Orange River and Limpopo River. Most of the industrial development is 
located in one sub-basin of the Vaal River system, namely the Klip River Basin, and it is 
this sub-basin that receives the bulk of the run-off from the Witwatersrand mining?
 industrial complex (Figure 1.1). Run-off consists of surface water from rain, treated and 
untreated sewage, treated and untreated discharges from industrial sources, and water 
pumped from mines.  
 
2. Aims and Objectives 
The Klip River Wetland has always been economically important to this region, initially as a 
source of water and of late as a purifier for the mentioned contaminants. A 
geomorphological investigation has already determined that these wetlands are in fact in an 
advanced state of collapse. The study by McCarthy et al. (2007) identified that the network 
of irrigation canals dug in the early part of the 20th century to support the agriculture in the 
area, have provided a preferred discharge conduit for the sewage pumped into the Klip 
River. The network of irrigation canals have cut back and avulsed forming an almost 
continuous channel and therefore, little water is flowing through the wetland. The wetland?s 
capability to remove contaminants from the water is compromised and McCarthy et al. 
(2007) predict that eutrophication problems can be expected down stream in the Vaal River 
Barrage as is experienced in the Hartebeestpoort Dam. This study aims to elaborate on the 
 3 
above research, especially with some newly acquired data, to better understand the extent 
of degradation of the wetland. 
  
The objectives of this research are to:  
? Establish the current status of the Klip River Wetland.  
? Calculate the progression of the degradation. 
? Determine the mechanisms/causes for the degradation. 
? Determine the broader implications to downstream users.  
? Offer potential remedial solutions.  
 
3. Literature Review 
3.1 Klip River Catchment Description 
3.1.1 Regional Setting 
The Klip River Wetland is situated south of Johannesburg, beginning in Lenasia and 
continuing south of Alberton and the Zwartkoppies Pump Station. The wetland lies between 
latitudes 26?10? and 26?25? south and longitudes 27?45? and 28?05? east (Figure 1.1). Thus, 
this wetland falls within the south-central portion of the major urban, industrial economic 
region of South Africa, the Southern Gauteng Metropolitan area, where rapid urbanization is 
taking place. The land use impacting on the wetland varies from urban to rural, industrial as 
well as informal settlements and agriculture. Partridge (1968) commented that this region 
has witnessed the greatest physical development of any area in Africa south of the Sahara.  
The Klip River headwaters are located south of Roodepoort with two main branches 
(Mindalore and Princess Dam) beginning on the continental drainage divide, the 
Witwatersrand ridge, at an altitude of about 1750 meters (Partridge, 1968). From there the 
Klip River gradually descends to the Vaal River, less than 1440 meters above sea level, 
near Vereeniging. The Klip River is therefore classified as part of the Orange/Vaal system 
draining into the Atlantic Ocean (Partridge, 1968).  
 
3.1.2 River Systems and Water Resources 
Impoundments or dams are confined mainly to the upper reaches of the catchment and are 
mostly man-made structures associated with mining activities (Davidson, 2003; Figure 
3.1.1). Other water bodies along the Klip River are sewage works maturation ponds, vlei 
 4 
areas and man-made recreational and farm dams. The tributaries of the Klip River include 
the Klipspruit, Harringtonspruit, Bloubosspruit, Glenvistaspruit, Rietspruit and Natalspruit, 
with the Foriespruit and Varkensfonteinspruit coming in at the lower Klip sub-catchment 
(Davidson, 2003; Figure 3.1.1).  
 
 
Figure 3.1.1: Location of the river system and water resources around the Klip River 
Wetland. 
 
3.1.3 Climate, Hydrology and Geohydrology 
The climate of the area is sub-humid with a tendency towards semi-aridity (Partridge, 
1968). The entire Klip River catchment has a typical Highveld climate with warm to hot 
summers (October ? March) and cool but generally cloudless winter days with cold nights, 
when frost is common (Davidson, 2003). Mean maximum temperatures average 26?C in 
January dropping to an average maximum of around 16?C in June. Precipitation is 
confined to the summer season mostly in the form of convectional downpours of high 
intensity, which favours rapid runoff (Partridge, 1968). The average summer rainfall is 
 5 
between 600-750 mm per annum (Grundling and Marneweck, 1999). Partridge (1968) 
concluded that the highest rainfall occurs in the vicinity of the Witwatersrand ridges and 
high-lying country to the east, where the greatest degree of atmospheric convergence is 
induced by the topography, i.e. at the Klip River headwaters. The average evaporation in 
the area is 1600 ?1700 mm a year with the highest evaporation taking place during the 
summer month of December (176 mm) and lowest rates during the winter month of June 
(70 mm) (Davidson, 2003). The evaporation rate is much greater than the rainfall and thus 
drought and water shortages may occur (Davidson, 2003). Water is transferred into the 
catchment for potable purposes through Inter Basin Transfer schemes (see Section 4.2; 
Davidson, 2003).   
 
According to Partridge (1968) the Klip River Basin, along with other south draining rivers, 
does not contain any major tributaries other than its axial streams when compared to 
north flowing rivers. From the continental divide, the southern flowing rivers generally 
show a lower density than the north flowing rivers (Partridge, 1968). The south flowing 
rivers carry mainly fine sediment, whilst the north flowing rivers carry sand and gravel. 
Hence, the latter lack wetlands. 
 
The Klip River Wetland is located on a plain and is underlain predominantly by dolomite of 
the Malmani Subgroup (Figure 3.1.2). These dolomite rocks form part of the Transvaal 
Supergroup and were deposited approximately 2, 500 m.y. ago in a shallow-marine, epeiric 
sea (Clay, 1981). In the past, natural dolomitic springs of pure water decanted into the Klip 
River Wetland (White, 1957). White (1957) describes natural springs of pure, dolomitic 
water in the Klip River beyond Olifantsvlei, which increases the pH, calcium and magnesium 
content and at the same time lowers the Total Dissolved Solids (TDS), sulphate (SO4) and 
iron (Fe) content. However, probably due to excessive pumping of the dolomitic aquifer, the 
springs no longer flow and contaminants are discharging from the Klip River into the aquifer 
(Kafri and Foster, 1989). 
 6 
 
Figure 3.1.2: Surface geology of the Klip River Catchment (black arrows highlight dykes; after Arnold, 2004). 
 7 
The annual recharge of the local aquifer varies between 75-110 mm per annum and a 
base flow of between 10-25 mm per annum (Grundling and Marneweck, 1999). The 
dolomitic aquifer is divided into sub aquifers on the basis of lithostratigraphy (Figure 
3.1.3). Kafri and Foster (1989) found that the chert-poor units are poor aquifers compared 
to the chert-rich units. Groundwater barriers formed by dykes divide the aquifer into 
compartments, which are linked by flow across the groundwater barriers along the Klip 
River. Two barriers are highlighted (by black arrows) on the surface geology in Figure 
3.1.2. Arnold (2004) and Schweitzer et al., (2004) interpreted dyke positions from 
aeromagnetic data (Figures 3.1.4 and 3.1.5) across the Klip River Wetland, which can be 
used to infer groundwater barriers. Kafri and Foster (1989) found that where the Klip River 
breaches the groundwater barriers the stream transcends in a series of small rapids until 
the groundwater level of the downstream compartment is reached.  
 
TR
 AN
 SV
 AA
 L
 KA
 RO
 O
 TR
 AN
 SV
 AA
 L
 KA
 RO
 O
  
Figure 3.1.3: Lithostratigraphy of the Malmani Subgroup (Kafri and Foster, 1989). 
 8 
 
Figure 3.1.4: Aeromagnetic anomaly map with dykes inferred (Arnold, 2004). 
 
Figure 3.1.5: Aeromagnetic map with dykes projected onto Central Rand Goldfield 
(Schweitzer et al., 2004). 
 9 
3.1.4 Vegetation 
The vegetation within the Klip River catchment falls into three Bioregions, namely Mesic 
Highveld Grassland (the dominant Bioregion), Dry Highveld Grassland and Central 
Bushveld (Mucina and Rutherford, 2006). The Mesic Highveld Grassland is the largest 
grassland region in South Africa with the highest number of vegetation types. These 
grasslands are found mainly in the higher precipitation parts of the Highveld and extend 
northwards along the eastern escarpment (Mucina and Rutherford, 2006). The dominant 
vegetation type of the Klip River catchment is the Rand Highveld Grassland, which is a 
species-rich, wiry, sour grassland alternating with low, sour shrub land on rocky outcrops 
and steeper slopes (Mucina and Rutherford, 2006). This Highveld Grassland is a typical 
grassland of the high inland plateau found along the ridges of the Witwatersrand and the 
dolomite plains of Gauteng mainly between 1 300 to 1 635 m in altitude (Davidson, 2003).  
 
Rogers and Herrera (1986) investigated three vegetation types in the Rietspruit Wetland, 
namely Typha capensis (Bulrush), sedentary and floating Phragmites australis (common 
reed) and Scirpus lucustris. The Klip River Wetland vegetation consists primarily of 
Phragmites australis with lesser Typha capensis and sedges.  
 
3.1.5 Wetland Characteristics and Status 
Wetlands are transitional areas between land and water and are distinguished by wet 
soils, plants that are adapted to wet soils, and a water table depth that maintains these 
characteristics. Since land and water can merge in many ways, there is no single correct 
definition for all purposes. According to Grundling and Marneweck (1999), the Klip River 
Wetland consists of four basins, all of which are classified as valley-bottom fens. The term 
fen is used mainly for minerotrophic peatlands (groundwater fed; H?jek, et al., 2006) and is 
a form of classification. Different from bogs (Ombrotrophic; rain fed), fens are identified on 
the basis of circum-neutral waters rich in dissolved minerals, notably calcium (Ca), and an 
abundance of brown mosses and vascular plant species (Warner, 1996). Wetland systems 
reduce or remove contaminants including organic matter, inorganic matter, trace organics 
and pathogens from the water. Reduction is said to be accomplished by diverse treatment 
mechanisms including sedimentation, filtration, chemical precipitation and adsorption, 
microbial interactions and uptake by vegetation (Watson et al., 1989). 
 
 10 
Peatlands ranging in size from 1 740 000 m? to 41 400 000 m? in the main basin, and with 
depths of between 2 m and 3.5 m, were identified in the Klip River Wetland by Grundling 
and Marneweck (1999). A transect across the wetland south of Kibler Park revealed the 
maximum peat thickness at 4 m (McCarthy and Venter, 2006). All basins of this peatland in 
the catchment of the Klip River can be described as tall emergent (reed?sedge) fens, made 
up of reed-sedge peat with a fibrous to fine-grained texture (Grundling and Marneweck, 
1999). 
 
Peat mining, mainly for the horticulture industry, took place in the wetlands east and west of 
the Klipspruit Valley Road, before operations ceased due to high pollution levels (Grundling 
and Marneweck, 1999). Open water bodies and scars are now left from this peat mining as 
seen in Figures 3.1.6 and 3.1.7. 
 
 
Figure 3.1.6: Open water bodies left from peat mining east of Klipspruit Valley Road 
(McCarthy et al., 2007). 
 11 
 
Figure 3.1.7: Open water bodies and scars left from peat mining. 
 
3.1.6 Land Uses 
The Klip River Wetland is located downstream and around the Witwatersrand mining?
 industrial complex. Land uses include:  
? Urban  
o Sewage Works 
? Mining 
? Industrial 
? Agriculture 
 
3.2 The History of the Klip River Catchment 
The name ?Witwatersrand? is literally translated as ?ridge of white waters?. This name arose 
because of the numerous small springs that were present along the ridge. The mining 
camps that sprung up along the Witwatersrand Ridge during the gold rush were generally 
sited according to this availability of water (Turton et al., 2006). George Harrison discovered 
gold on the ridge early in February 1886 and by October 1886 the city of Johannesburg was 
 12 
officially founded. The first water for the mining town was drawn from two springs called 
Fordsburgspruit and Natalspruit (Turton et al., 2006). Prior to 1896 another source of supply 
was developed north of the mining operations in Roodepoort, on the farm Weltevreden 
(Turton et al., 2006). When this water supply failed to meet the demands, development 
turned to the dolomitic aquifers that underlie the Klip River Wetland, firstly on the farm 
Zuurbekom. The Zuurbekom pump station was erected in 1899 and is still in operation 
today (Turton et al., 2006; Figure 1.1).  The water demand grew dramatically as more mines 
were founded. The water requirements for processing gold ore were approximately 2 000 
litres of water per ton of ore. Between 1886 and 2002, 937 375 000 tons of ore was milled 
and 7 697 666 kg gold was extracted from the Central Rand Goldfield (Handley, 2004).  
 
By 1902 there were three companies responsible for water supply in the Witwatersrand ? 
Johannesburg area namely: the Johannesburg Waterworks Estate and Exploration 
Company, the Braamfontein Company and the Vierfontein Syndicate (Turton et al., 2006). 
The sources of water were from: 
? The dolomites around Zuurbekom.  
? Springs at Doornfontein, Natalspruit, Berea and Parktown.  
? The Klip River Valley pumping station, i.e Zwartkoppies (Turton et al., 2006).  
Turton et al. (2006) documents the first incidence of the Klip River being polluted in 1894, 
which resulted in the death of livestock. An enquiry established that the water in the 
Doornfontein Valley was also polluted by mine effluent (Turton et al., 2006). This was the 
beginning of a substantial water resource management problem that persists today. In 1903 
the Rand Water Board (RWB) was established and given the responsibility of developing a 
secure water supply system. In 1914 the RWB adopted the Vaal River Development 
Scheme. The first phase of the Scheme was the construction of the Vaal Barrage (Figure 
3.2.1) with purification works, pumping stations in Vereeniging and a main pipeline to the 
Witwatersrand being completed in 1923 (Turton et al., 2006). It is noted that the Klip River 
enters the Vaal River upstream of the Barrage, but downstream of the Vaal Dam (Figure 
3.2.1). At this point in time the Witwatersrand was effectively recycling its waste water. The 
Vaal Dam being the second phase of the Scheme was built during the great depression. 
The yield of the Vaal Dam was increased in 1955/6 to result in the capacity of 2 330 million 
m3. This was still insufficient with Pretoria and Vereeniging being incorporated into the Rand 
 13 
Water Board supply area (PWV Triangle), the huge demand was reaching the limits of the 
Vaal?s capacity. This gave rise ultimately to the Lesotho Highlands Project with the first 
water flowing into the Vaal Dam via the Ash River outfall on 8 January 1998 (Turton et al., 
2006). The fact that the water is pumped to the Witwatersrand from the Inter Basin Transfer 
schemes, which is then transferred to the Klip River through the water treatment plants, 
means that this system is highly altered. The RWB eventually stopped pumping water from 
the Barrage due to water quality concerns. The Vaal River upstream of the Barrage is 
however still used as a major recreational area for the Gauteng region and water quality 
remains a concern.  
 
 
Figure 3.2.1: The Klip River in relation to the Vaal Barrage and the Vaal Dam. 
 
 14 
3.3 Water Pollution Sources: Mining 
The first incidence of the Klip River being polluted was documented in 1894 (Turton et al., 
2006). Figure 3.3.1 illustrates the impact of mining on the wetland in aerial photographs of 
1938. The glowing white substance washed from the river into the wetland is mine dump 
sediment entering this ecosystem from the Roodepoort mines. The effect of this effluent on 
the wetland can be seen by the absence of reeds and vegetation in the polluted areas. 
Today, however, the vegetation is fully recovered and no fluorescing white polluted water is 
evident (Arnold, 2004).  
 
An evaluation of the hydrogeology of the Klip River Basin was carried out in 1989 in order to 
assess the aquifer?s potential for public water supply. It was determined that effluents from 
sewage purification works and gold treatment plants have caused extensive deterioration of 
the quality of surface and groundwater (Kafri and Foster, 1989). High TDS and sulphate 
values were encountered along the Klip River Valley with contamination increasing towards 
the wetland (Figure 3.3.2). This indicates that there is a reversal of flow; while previously the 
wetland was fed by springs (Humphrey, 1910; White, 1957) now water flows out into the 
dolomitic aquifer, probably due to over pumping of the aquifer. Water sampled from a spring 
south of Olifantsvlei in the 1950?s was found to have a low saline content, chlorides and 
sulfates were also low (4 ppm) and nitrogenous compounds were absent (White, 1957). 
Prior to 1957 the TDS were 138 mg/L, and in 1989 Kafri and Foster record this area to fall 
into the > 500 mg/L zone. White (1957) described natural springs of pure, dolomitic water in 
the Klip River beyond the Olifantsvlei, which used to increase the pH, calcium and 
magnesium (Mg) content and at the same time lower the TDS and sulfate and iron (Fe) 
content.  
 
Figure 3.3.3 illustrates that Acid Mine Drainage (AMD) from the Klip River catchment is also 
reporting1 to the Vaal River (DWAF, 2007). TDS values are low up until the Vaal Dam and 
thereafter increase significantly.  
 
 
                                               
1
  Synonymous with ?draining? or ?entering? 
 15 
 
Figure 3.3.1: Acid Mine Drainage during 1938 impacting the Klip River Wetland. Areas 
impacted indicated by red arrows (Arnold, 2004). 
 16 
 
Figure 3.3.2: Total Dissolved Solids concentration map of the Klip River Basin (Kafri and Foster, 1989). 
 17 
 
Figure 3.3.3: Total Dissolved Solids concentration map of the Vaal River Basin (DWAF, 2007). 
 18 
Mine, sewage and industrial effluent has been flowing into the Klip River and the associated 
wetland for over a hundred years. AMD is the water flowing from or caused by surface 
mining, deep mining or refuse piles and is typically highly acidic with elevated levels of 
dissolved metals. The formation of AMD is primarily a function of the geology, hydrology 
and mining technology employed for the mine site. AMD is formed by a series of complex 
chemical and microbial reactions that occur when water comes in contact with pyrite (FeS2), 
as contained in the auriferous conglomerates of the Witwatersrand, refuse or the 
overburden of a mine operation. The resulting water is usually highly acidic and has high 
TDS. The metals stay dissolved in solution until the pH raises to a level where precipitation 
occurs. AMD results when the mineral pyrite is exposed to air and water, resulting in the 
formation of sulfuric acid (H2SO4) and iron hydroxide (Fe(OH)3):  
 
FeS2 + 3.75 O2 + 3.5 H2O ? Fe(OH)3 + 2 H2SO4 
 
The products of AMD formation, acidity and iron, can devastate water resources by lowering 
the pH and coating stream bottoms with iron hydroxide, forming the familiar orange coloured 
Ferric Hydroxide or "yellow boy" common in areas with abandoned mine drainage. 
 
3.3.1 Source of Acid Mine Drainage 
On the Central Rand Goldfield, gold was initially extracted using a mercury amalgam 
method, but later as mining operations went deeper and unoxidized ore containing pyrite 
was encountered, the MacAuthur-Forrest process, using cyanide was implemented (Naicker 
et al., 2003). The cyanide process required fine milling of the ore, after which it was treated 
with a cyanide containing solution, which dissolved the gold. The solution was then 
separated for further processing, while the tailings were pumped to large dumps (slimes 
dams). Both the mercury amalgam and the cyanidation processes are highly selective for 
gold, and other ore minerals, such as pyrite, were unaffected and hence reported to the 
tailings dumps. 
 
A principal environmental concern associated with mine waste results from the oxidation of 
sulfide minerals within the waste materials and mine workings, and the transport and 
release of these. The principal sulfide minerals in mine wastes are pyrite and pyrrhotite 
(FeS), which release elements such as aluminium (Al), arsenic (As), cadmium (Cd), cobalt 
 19 
(Co), copper (Cu), mercury (Hg), nickel (Ni), lead (Pb), and zinc (Zn) to the water flowing 
through the mine waste and into the groundwater below. The Witwatersrand Supergroup 
sediments contain varying proportions of sulphide minerals, the predominant sulphide being 
pyrite. In many of the gold bearing reefs, this mineral may form up to three per cent (by 
mass) of the rock. Other sulphides such as pyrrhotite, arsenopyrite (FeAsS), chalcopyrite 
(CuFeS2), galena (PbS), cobaltite ((Fe,Co)AsS) and gersdorffite NiAsS), occur as traces 
(Schweitzer et al., 2004). Average uranium (U) grades of Witwatersrand orebodies are 
about 250 ppm with uranium predominantly contained in uraninite (UO2) and brannerite 
(UCaCeTiFeO), (Handley, 2004).  
 
Mining of the Witwatersrand Supergroup sediments has produced rock piles, sand and 
slimes dams on surface, and underground backfilled rockpiles and reef remnants in stopes 
and developments2. These all contain pyrite that is constantly exposed to air and water. Due 
to oxidation and subsequent precipitation of secondary iron oxide minerals and mineraloids, 
the rocks and precipitates exhibit red and yellow staining and discoloration.  Water that has 
passed through these precipitates exhibits low pH and high concentrations of metals in 
solution. 
 
3.3.2 Transport of Acid Mine Drainage 
Acidic environments, particularly at pH values less than 3, are very conducive to the mobility 
of contaminants at very high concentrations. The major contaminant that can move beyond 
the neutralization zone is sulfate. Its concentration in solution is not directly pH dependent; 
therefore neutralization will not inhibit its mobility as it does with metals (Deutsch, 1997). 
The dissolved concentration of sulfate is usually limited by the solubility of gypsum 
(CaSO4.H2O). However, the solubility of this mineral can vary depending on the other 
constituents in solution. In the neutralization zone, the solubility of gypsum will decrease as 
the cations are removed from solution (Deutsch, 1997). Because of the mobility of sulfate 
compared with the other constituents in this type of acid plume, sulfate is the best indicator 
of the first arrival of contaminants from a sulfide source. A study by Naicker et al. (2003) 
found ground water within the Central Rand mining district to be heavily contaminated and 
acidified. Naicker, et al (2003) commented that although erosion of dump material into water 
                                               
2
  Refers to underground excavations, especially haulages and cross-cuts. 
 20 
courses contributed to pollution, the major contribution came from rain water which had 
percolated through the dumps, creating polluted groundwater plumes beneath the dumps, 
which were emerging at surface in streams.  
 
Formation of AMD on the Central Rand is dominated by the hundreds of mine tailings 
dumps which are scattered all along the Witwatersrand Ridge and form large footprint 
plumes contaminating the natural water table and associated surface water bodies, i.e. the 
Klip River Catchment. Some of these mine tailings have in the past been reprocessed, or 
are currently being reprocessed due to their relatively high gold values and more efficient 
processing methods. In some cases it takes up to 15 years to reprocess a large dump and 
during reclamation, vegetation cover is removed and exposed faces are subject to wind and 
water erosion and more importantly to oxidation (Mphephu, 2004).  
 
Another major AMD source is the underground workings, which are now flooding and  being 
pumped from East Rand Proprietary Mine (ERPM) in the eastern limit of the Central Rand 
Goldfield. This water is then treated with lime and sent through a High Density Sludge Plant 
where precipitated iron hydroxide is removed. The water is then pumped into the 
Elsburgspruit, which flows into the Natalspruit and Rietspruit through wetlands and the 
Malamani Dolomites to the south of Johannesburg before joining with the Klip River and 
then the Vaal Dam (Figure 3.1.1).  
 
3.3.3 Bioaccumulation and the Toxicity of Oxidation Products 
Metals released from mines and mine sites can harm aquatic biota in adjacent water bodies. 
Metals are present in aqueous environments in a variety of species with various toxicities 
and potential for bioaccumulation. Transformations among these species depend on the 
physical and chemical characteristics of the water body. Metal toxicity can be acute or 
chronic. For example, aluminium can precipitate to form species that result in sudden killing 
of fish in water bodies near mine sites. Other metals, such as mercury, can bioaccumulate, 
leading to chronic toxicity (Blowes et al., 2003). The extent of bioaccumulation and the 
toxicity of metals within natural environments are controlled by a number of factors, among 
which are pH, oxidation?reduction potential, organic carbon content, concentrations and 
compositions of other dissolved species, and the composition of the sediment (Blowes et 
al., 2003). 
 21 
The acidification of wetlands can elevate the concentrations of metals and increase the 
potential bioavailability in aquatic plants and freshwater biota and influences the uptake of 
metals in both submerged and rooted plants. Metals such as iron, copper, cadmium, 
chromium (Cr), lead, mercury, selenium (Se), and nickel can produce reactive oxygen 
species, resulting in lipid peroxidation, DNA damage, depletion of sulfhydryls, and calcium 
homeostasis (Blowes et al., 2003).  
 
Iron is an essential element for metabolic systems, but in iron-rich solutions toxicity can 
develop in both fish and biota. In mine-waste discharge, Fe(III)-sulfate and (oxy)hydroxide 
precipitates can accumulate on the gill epithelium, resulting in clogging and damage, and 
decreasing the available surface area and increasing the diffusion distance for respiratory 
exchange (Blowes et al., 2003).  
 
Several studies considered the bioaccumulation of heavy metals in various components of 
the food chain in the Elsburg/Natalspruit wetland ecosystem which is fed by the water 
pumped out of the Central Rand Gold Mines at ERPM. In one study, the concentration of 
zinc, manganese and nickel in Arundo donax and Typha capensis increased significantly 
under more alkaline water conditions, while the uptake of iron was found to be less effected 
by the pH (Rall and Rall, 1996). The metals did not appear to be detrimental to these plant 
species, as the plants thrived under both acidic and alkaline conditions. One suggestion 
was that the reeds be cropped regularly during the summer growing season, so that a 
substantial portion of the heavy metals present in the plant, and thus in the ecosystem, 
could be effectively removed. At present the reeds are burned occasionally in the winter 
months, with most of the ash and therefore metals, being returned directly into the stream 
ecosystem.  
 
The biological significance of heavy metals in aquatic and semi-aquatic bird eggs is still 
unknown, although potentially toxic metals in the egg shells could induce various 
teratogenic effects and may even be lethal to the developing embryo. High lead 
concentrations detected in adult coots may have a serious impact on the long term health 
status (Rall and Rall, 1996). Heavy metal levels detected in the muscular tissue of fish 
sampled in the Natalspruit were found not be unduly high and would therefore not pose any 
health threat to human life should these fish be consumed.     
 22 
 
A study by Haywood (2004) investigated the use of Xenopus laevis embryos and tadpoles 
as biological indicators for the measurement of biologically available metals and acidity 
found in water associated with AMD within water bodies on the Witwatersrand. Exposure to 
pH 3 and pH 4 showed clear toxicological effects on the growth and development of the 
embryos and tadpoles. Upon exposure to individual bioavailable metals, the relationship 
between metal concentrations and the various physiological responses of the tadpoles was 
different for each of the metals tested. Copper and cadmium affected the hatching of the 
tadpoles whereas zinc and lead affected tadpole survival after hatching. For the field 
research, the physiological responses of X. laevis embryos and tadpoles exposed to water 
collected along the length of the Natalspruit provided a spatial and temporal indication of the 
impact of AMD within the river. The baseline physiological responses of the X. laevis 
embryos and tadpoles exposed to known pH levels and metal concentrations within the 
laboratory were used to correlate the responses of the embryos and tadpoles exposed to 
the water from the Natalspruit River. It was evident that the pH of the water influenced the 
toxicity of the lethal and sub-lethal pollutants within the river. 
 
3.4 Water Pollution Sources: Waste Water 
Another category of water quality problems in the Klip River Catchment relates to the 
effluent drainage from the Rand Water supply and runoff from the informal settlements. A 
number of informal settlements flank the banks of the Klipsruit and the Klip River. Ineffective 
maintenance and infrastructures of water and sewage systems results in raw sewage and 
polluted water entering the river system. Discharge from three sewage treatment plants 
enters the Klip River Wetland (Figure 1.1) namely:  
? Olifantsvlei ? 182 Ml/day 
? Bushkoppies ? 186 Ml/day 
? Goudkoppies ? 126 Ml/day 
 
Consequently, approximately 500 Ml/day flows into the Klip River Wetland from these three 
plants. Figures 3.4.1 and 3.4.2 illustrate the velocity and volumes of water entering the Klip 
River Wetland from the Olifantsvlei and Bushkoppies sewage treatment plants in the dry 
winter month of August (2008) and a wetter November (2007).  
 
 23 
 
Figure 3.4.1: Discharge point of the Olifantsvlei sewage treatment plant ponds into the Klip 
River Wetland (August 2008). 
 24 
 
Figure 3.4.2: Discharge point of the Bushkoppies sewage treatment plant ponds into the 
Klip River Wetland (courtesy of T.S. McCarthy, November 2007). 
 
Bushkoppies Wastewater Treatment Works collects and treats sewage from the southern 
suburbs of Johannesburg, Soweto East and from the industries to the south of 
Johannesburg. The works were commissioned between 1984 and 1985. It consists of a 
head of works with screening, degritting, primary sedimentation, thickeners for raw sludge, 
thickeners for waste activated sludge, bioreactors, final clarification and maturation ponds. 
No sludge is treated at this works. Gravity thickened raw sludge is pumped to 
Goudkoppies Wastewater Treatment Works for digestion, dewatering, solar drying and 
disposal. Gravity thickened waste activated sludge is pumped to Olifantsvlei?s RSHF 
(Regional Sludge Handling Facility) for dewatering and composting (Johannesburg Water, 
2004). 
 
Goudkoppies Wastewater Treatment Works collects and treats sewage from the City 
Centre and the south-eastern areas of Johannesburg. The works was commissioned in 
 25 
1978 and consists of a new head of works with screening, degritting, primary 
sedimentation, raw sludge thickening / acid fermentation, flow balancing, activated sludge 
incorporating the five stage Phoredox process, final clarification, chlorination, waste 
sludge thickening, digestion, dewatering and solar drying of sludge. Waste activated 
sludge is thickened in dissolved air flotation units and mixed with raw, thickened sludge 
and thickened sludge produced at Bushkoppies Works. The mixture is anaerobically 
digested and thereafter dewatered on linear screen/belt press units and solar dried on 
drying beds. The dried sludge is ultimately disposed of on privately owned farmland 
(Johannesburg Water, 2004). 
 
Olifantsvlei Wastewater Treatment Works collects and treats sewage from 3 outfalls. The 
south-western outfall collects sewage from the western sewer areas of Soweto via the 
inlet screw pump station. The south-eastern outfall obtains sewage from the southern and 
south-eastern areas of Johannesburg. The third outfall collects sewage from Lenasia via 
the Van Wyksrust Pump Station (Johannesburg Water, 2004). The works consists of 3 
units: 
? Unit 1: Was commissioned in 1956, but was decommissioned and abandoned in 
the early 1990?s. 
? Unit 2: Was commissioned in 1973 and was originally an extended aeration plant 
(without primary sedimentation). The unit consists of four activated 
sludge bioreactors and final clarification. Each module was designed to 
treat 30 MI/d. Two reactors were decommissioned in the mid 1990?s. 
? Unit 3: Was commissioned in 1996 and consists of a new head of works with 
screening and de-gritting, primary sedimentation, acid fermentation, flow 
balancing, four stage bioreactors incorporating the four stage 
Johannesburg process and final clarification. The capacity of each 
module is 50 MI/d. 
 
Effluent from Units 2 and 3 are combined and flow through a series of five maturation 
ponds before being discharged into the Klip River. Sludge treatment is carried out at the 
Olifantsvlei Regional Sludge Handling Facility (RSHF), commissioned in 1997. The facility 
has capacity to compost biological sludge produced at Olifantsvlei, Goudkoppies and 
Bushkoppies Works and primary sludge from Olifantsvlei. The facility consists of a sludge 
 26 
dewatering facility (belt presses and linear screens) with liquor treatment, a sludge cake 
and bulking agent blender, a composting area, screens to recover bulking agent, and a 
bulking agent preparation plant. All primary sludge is thickened in acid fermenters / 
thickeners and anaerobically digested before being dewatered on belt presses. Waste 
activated sludge is gravity thickened and could either be digested with the primary sludge, 
or directly dewatered on belt presses. All dewatered sludge is composted or disposed of 
on private farms. Filtrates from the dewatering unit are lime treated to reduce phosphates 
(PO4). After treatment, the lime treated liquors could either be pumped back to the head of 
works or to Bushkoppies Works for treatment and disposal (Johannesburg Water, 2004). 
 
Associated with sewage treatment is nitrogen (N) and phosphates. Nitrogen is the most 
common element in the earth's atmosphere, where it is present as the gas N2. Nitrogen is 
also an important nutrient required for all living things. The amount of nitrate (NO3) present 
in a water body will determine how much algae and plant material will be able to grow. An 
excess of nitrogen will allow aggressive plant and algae growth and contributes to 
eutrophication of a water body. Nitrate is soluble in water, and thus naturally enters bodies 
of water via rain, surface runoff and groundwater. Yet anthropogenic sources can increase 
the amount of nitrate entering a water body. Sewage effluent and fertilizer runoff are the 
most significant sources of anthropogenic nitrate 
 
Phosphorus (P), like nitrogen, is a nutrient required by all life forms. However, phosphorus 
is generally present in very small amounts, thus it is often the limiting nutrient in an 
aquatic system. This means that plants have all of the ingredients necessary to grow, 
except for phosphorus. When phosphorus is added, even in minor amounts, plant growth 
is possible and algae growth results. Because of its role as a limiting nutrient, phosphorus 
should be carefully monitored in areas where eutrophication is a threat. Phosphate can 
also be added to water bodies due to the addition of sewage effluent, fertilizer runoff, and 
urban runoff. For example, sewage that has been through primary and secondary 
treatment contains 5 to 8 ppm phosphate, most of which is a result of household soaps 
and detergents in the wastewater. Because of its role as a limiting nutrient, this additional 
phosphate can allow a tremendous amount of algae growth that would otherwise not 
occur. 
 
 27 
It is possible to determine the trophic state of a lake by measuring the amount of nutrients 
in the water. The following classification terms are used by the Department of Water Affairs: 
National Eutrophication Monitoring Programme (NEMP, 2003):  
? Oligotrophic ? low in nutrients and not productive in terms of aquatic animal and 
plant life.  
? Mesotrophic ? intermediate levels of nutrients, fairly productive in terms of aquatic 
animal and plant life and showing emerging signs of water quality 
problems.  
? Eutrophic ? rich in nutrients, very productive in terms of aquatic animal and plant 
life and showing increasing signs of water quality problems.  
? Hypertrophic ? very high nutrient concentrations where plant growth is determined 
by physical factors. Water quality problems are serious and can be 
continuous.  
 
Table 3.4.1 illustrates the method to determine trophic status statistics and Table 3.4.2 
provides the trophic status classification of the Vaal Dam and the Vaal Barrage as 
determined for the period October 2002 to September 2003. Both the Vaal Dam and the 
Barrage were showing algal productivity in 2002 to 2003. 
 
Table 3.4.1: Method to determine trophic status statistics (NEMP, 2003). 
 
 
 28 
Table 3.4.2: The Trophic Status classification of the Vaal Dam and the Vaal Barrage 
(NEMP, 2003). 
Dam Name Mean TP 
mg/L n TP 
Mean 
Annual 
Chlorophyll 
a (?g/l) 
n Chl
 Percent of 
Time Chl a 
> 30 ?g/l 
Trophic Status
 Vaal Barrage 0.069 10 8.4 6 10 Oligotrophic, significant potential 
and current algal productivity 
Vaal Dam 0.077 46 14.78 16 17 Mesotrophic, significant potential 
and current algal productivity
  
 
The rivers flowing north of the Witwatersrand Ridge (Crocodile River), where little to no 
wetlands are present, also have treated sewage water discharged into them. In contrast to 
the Vaal Barrage, the Hartebeespoort Dam is impacted chronically with eutrophication. The 
Trophic Status classification of South African impoundments determined, for the period 
October 2002 to September 2003, that the Hartebeespoort Dam was hypertrophic with 
significant current algal productivity. According to the North West Environmental 
Management Plan for the Hartebeespoort Dam Remediation (2005) roughly 90% of the 
yearly inflow is derived from the Crocodile River. During the dry season the water of the 
dam is more than 50% treated wastewater from urbanised areas upstream. Every year 
over 170 metric tonnes of phosphorus is discharged into the dam. The immediate cause of 
the unpleasant and often hazardous quality of the water is blue-green algae i.e. 
cyanobacteria. Green, smelling and often toxic algae blooms frequently cover large areas 
of the water surface (Figure 3.4.3). 
 
Figure 3.4.3: Algal blooms seen at the Hartebeespoort Dam wall (North West 
Environmental Management Plan, 2005). 
 29 
The excessive algae and cyanobacteria blooms considerably reduce the recreational 
values and usability of water as well as disturb the natural biological mechanisms in the 
water body. Often, people swimming and water-skiing end up in rashes or even more 
serious health problems. Domestic animals and pets die after consuming a sufficient 
quantity of this toxic water (North West Environmental Management Plan, 2005). The algae 
toxin (microcystin) levels of the water in the Hartebeespoort Dam are regularly so high, 
that according to the international guidelines of World Health Organisation, all uses of the 
water, including swimming and water-skiing, should be banned seasonally (North West 
Environmental Management Plan, 2005). These problems have prevailed in the 
Hartbeespoort Dam ever since the early 1970s. Studies (North West Environmental 
Management Plan, 2005) clearly established that the Hartbeespoort Dam can be saved. 
However, there are no cheap, quick or easy ways to do this. The North West Environmental 
Management Plan for the Hartebeespoort Dam Remediation (2005), predicts a start up 
cost of R2 million and annual running costs of R1.8 million to comply with this 
management programme. 
 
3.5 Wetlands and their Importance 
The value and importance of wetlands has been drawing increasingly more interest over the 
past few decades. Wetlands produce and sustain many diverse life forms, especially 
wetland dependant species as well as rare and endangered species (Begg, 1990). Their 
most important function is the fulfillment of hydrological and hydrochemical functions ? 
intercepting storm runoff and storing storm water, recharging groundwater, removal of 
organic and inorganic nutrients as well as toxic materials (Hammer, 1997).  
 
In 1975 South Africa became the first African signatory to the Ramsar Convention, which 
obligates government of this country to protect designated wetlands (Whyte and Shepherd, 
1990). The strain on future resources of this country (such as freshwater) means that in the 
face of exponential population growth, man?s dependence upon wetlands is steadily 
increasing (Begg, 1990).  
 
As with the proximity to the metropolitan area of Johannesburg, the Klip River Wetland is 
impacted by infrastructure, transport systems, urbanisation and sewage as well as peat 
mining and gold mining activities (Grundling and Marneweck, 1999). Figure 1.1 highlights 
 30 
industrial and mining areas as well as sewage and water works along the catchment. 
Cultivation, especially vegetable farming, on the perimeters of the wetland requires water to 
be abstracted and channelled for irrigation, while sewage works discharge effluent into the 
wetland impacting negatively on the quality of the water and peat (Grundling and 
Marneweck, 1999). 
 
The wetlands of the Klip River have played a vitally important role to ensure the natural 
treatment of polluted mine water on the Central Rand.  It has been observed that as water 
moves south of the catchment towards the Vaal River, element concentrations and 
attributes, such as iron, sulphate, conductivity and pH improve to acceptable and ideal 
levels, versus unacceptable levels near the mines and mine dumps (Figure 3.5.1). This 
dramatic improvement of water quality is attributed to the presence of the large tracts of 
Malmani Dolomites as well as the extensive wetlands in the area (Davidson, 2003). The 
alkaline dolomites lead to an increase in the pH of water, which enhances the precipitation 
of iron and other elements, which become insoluble in water of neutral pH (Davidson, 2003).  
The wetlands serve as a sink for pollution where polluting metals and other inorganic and 
organic constituents are trapped in sediments (Arnold, 2004). Dilution from runoff and 
treated sewage water probably is important in neutralizing the water chemistry. 
 
Water analysis completed by Grundling and Marneweck (1999) in the northwest basin of the 
Klip River Wetland indicated slightly acidic to slightly alkaline (pH: 6.3 - 7.23) levels with the 
water rich in carbonates, because of it?s dolomitic origin, and high sulphate and mercury 
levels, confirming the pollution by the gold mines. Analysis of the main basin further 
downstream indicated that, although sterilised, this wetland is still acting as a natural filter, 
since lower sulphate, mercury and cadmium levels were found (Grundling and Marneweck, 
1999). 
 
 31 
 
Figure 3.5.1: Klip River catchment showing distribution of a) pH b) conductivity, c) iron and 
d) sulphate (Davidson, 2003). 
 32 
Peat sampling in the Klip River Wetland has also recorded these pollution levels. 
Concentrations of heavy metals and phosphorus in the inorganic content of the peat are 
higher in the uppermost reaches of the wetland (closer to the source) than downstream and 
further from the source (McCarthy et al.,2007). McCarthy and Venter (2006) concluded that 
were it not for the Klip River Wetland trapping the pollutants, they would have reported 
directly to the Vaal River system, which would have created a much more widespread 
pollution problem. It was further noted by these authors that this wetland fulfills such an 
important function to the region that it should be rigorously protected.  
 
Hancock (1966) realized the importance of these reed swamps for water purification prior to 
1966 when completing her thesis on the algal ecology of the Klip River in the gold mining 
area of Johannesburg. Proximal to the mine dumps, which is at the source of the Klip River, 
she described the water quality as ?Lethal? (Hancock, 1966, page 49). 
In the Klip River all populations (of algae), from whatever source, are completely 
destroyed in the lethal stretches. 
 
Her conclusion on what conditions brought about the recovery of the stream from 
degradation, caused by the gold-mining industry, were three fold:  
1. The nature of the stream bed (dolomitic limestone) 
2. Dilution (from the natural springs and sewage water) 
3. Reed swamps 
 
The open swamps of Phragmites, she highlights, improve conditions physically by causing 
the slowing down of water, which permits greater sedimentation of suspended materials, 
while the vegetation acts as a mesh to ?flotsam and jetsam?. The chemical improvement 
was attributed to the fact that there is greater degree of neutralization, as the water has a 
longer contact time with the dolomite beds. The rise of pH then causes the precipitation of 
iron-hydroxide. Biological effects were seen in the increase of bacteria, nitrification, and 
algal populations. This in turn, during photosynthetic activity, further raised the pH, 
oxygenated the water, and together with the increase in heterotrophs, added further organic 
material to the stream.  
 33 
4. Methodology 
Johannesburg has an extensive aerial photography collection and the Klip River Wetland 
was analysed using photographs going as far back as 1938. By studying the changes seen 
in successive photographs, including 1938 (job 129), 1952 (job 314), 1961 (job 438), 1984 
(job 881) 2003 and 2006, the aim was to determine the changes to the wetland over time 
and associated rates of change. The photos were georeferenced in GIS (Geographic 
Information Systems; ArcMAP Version 9.2) to allow convenient comparisons. 
 
Historical maps of the area were also anlaysed. The Klip River Wetland was included in the 
area geologically mapped between 1910 and 1915 by Mellor, E.T. Although Arnold (2004) 
analysed the wetland as recorded by Mellor, these maps will be analysed again, now with a 
better insight of the system, especially in relation to channel formation. Goldman maps of 
the 1890?s (Goldmann, 1895-6) were also considered for analysis but unfortunately this 
section of the Witwatersrand was not mapped in detail. Historical reports and descriptions of 
the study area were analysed in conjunction with the aerial photographs. 
 
In 2006 a LIDAR survey was undertaken from April to May 2006 by Airborne Laser 
Solutions (ALS) to produce digital orthophotos with 20 cm pixels, and 25 cm contours for the 
City of Johannesburg. Light Detection and Ranging (LIDAR) is a remote sensing system 
used to collect topographic data. This survey was carried out using an aircraft mounted 
LIDAR system that ?scanned? the ground below with a 33kHz laser resulting in a dense 
DTM of the ground surface and objects above the ground. The flying height of the aircraft 
was 1 100 m above the ground level with a swathe width of approximately 700 m. The 
project survey report is provided in Appendix 1. This topographic data was analysed to 
document topographic changes along the wetlands length, i.e. a long river profile. A test 
was conducted along the N1 Highway to determine the noise (error) in the dataset 
(Appendix 2).  
 
The Department of Water Affairs and Forestry (DWAF) and Rand Water Board have 
monitoring points along the length of the Klip River (Figure 4.1). Water chemistry and 
discharge data from these points were evaluated to determine secular changes in water 
 34 
quality. Where water chemistry and discharge data is sparse, technical reports by the 
hydrological research units and results of Ph.D. and M.Sc. theses were used. 
 
 
Figure 4.1: Location of monitoring stations along the Klip River (grey areas indicate 
industrial land use). 
 35 
5. Results 
The extent of the Klip River Wetland is vast and the impacts along it variable, thus the 
wetland has been divided into five sections (Figure 5.1.1): 
? Section 1: Lenasia to Olifantsvlei Water Treatment (Golden Highway) 
? Section 2: Golden Highway to the N1 
? Section 3: N1 to Kibler Park 
? Section 4: Kibler Park to Kromvlei 
? Section 5: Zwartkoppies Section 
 
 
Figure 5.1.1: The wetland divided into five sections. 
 
 36 
5.1 Aerial Photography  
Attempts were made to quantify the extent of the wetland over time to determine the rate of 
wetland deterioration. This proved to be subjective using aerial photography only. Wetlands 
have historically been one of the most difficult ecosystems to quantify. A Landsat image of 
the wetland was sourced to determine if satellite multi-spectral imagery could be used to 
quantify the extent of the wetland. The Landsat image (Figure 5.1.2) identifies water as 
black. Using this methodology the wetlands (or peat) only contain moisture in the western, 
very proximal section totalling 57.45 ha. Without further analysis it is uncertain whether this 
represents peat moisture or flow of water across the wetland. The 2003 photography shows 
burns from the winter fires downstream of this moist zone (Figure 5.1.3), potentially 
confirming this moisture.  
 
 
Figure 5.1.2: Landsat image showing the water and moisture content of the wetland. 
 
 37 
 
Figure 5.1.3: Burns in the wetland (2003) downstream of the moist section of wetland. 
 
It would be expected that with the diversion of water from the wetland into channels, as was 
discovered by McCarthy et al. (2007), the wetland area would reduce in size over time. A 
portion of the geological map of the Witwatersrand mapped between 1910 and 1915 by E.T. 
Mellor covers the Klip River Wetland. Using the georeferenced aerial photography and 
augmenting it with Mellor?s mapping (Mellor, 1917) it was possible with GIS to quantify the 
wetland area at particular periods of time (Table 5.1.1).  
 
Table 5.1.1: Change in wetland area over time. 
 
1917 1938
 1 456 569
 2 105 47
 3 199 39
 4 221
 5 194 415
 Wetland Area (ha)
 2006
 546
 56
 78
 Section
 465>634
 38 
In Section 1 the wetland area increased in size between 1917 and 1938 (Table 5.1.1). This 
is confirmed by the increased rate of peat accumulation since the establishment of 
Johannesburg, as described by McCarthy and Venter (2006). Subsequently, there was a 
slight decrease in wetland size between 1938 and 2006 which can, in part, be attributed to 
peat mining.  
 
The wetland in Section 2 decreased in size by 58 ha over 20 years (1917 to 1938). The 
wetland area then increased marginally (9ha) over the period 1938 to 2006. In Section 3 the 
wetland decreased in size by 179 ha in 20 years and then doubled in size between 1938 
and 2006. A combined area measurement was taken for Sections 4 and 5. Unfortunately 
Mellor?s mapping only extends through to a part of Section 5 and so the minimum area was 
calculated for 1917 at 634 ha. The wetland area decreased by more than 170 ha between 
1917 and 1938 (a similar reduction as Section 3 over the same time period) and a further 50 
ha by 2006.  
 
To summarise, with the exception of Section 1, the wetland decreased in size by 41% 
between 1917 and 2006. Sections 2 and 3 increased in extent by 9 and 39 ha, respectively 
between 1938 and 2006. It is evident from this summation that the measurement of wetland 
area is not an effective way to quantify the extent of the wetland degradation from 
canalisation. This might be due to the wetland receiving sufficient water from rainfall and 
local runoff to remain functioning, albeit not optimally. 
 
Channel formation, propagation and widening along with knick point migration are 
potentially better indicators of wetland deterioration. At six points (A to F) along the 
wetland?s length the channel width was measured off aerial photography and with the use of 
GIS between 1938 and 2006 (Figure 5.1.4 and Table 5.1.2). 
 39 
 
Figure 5.1.4: Locality of channel width measurements. 
 
Table 5.1.2: Channel widths over time along the length of the Klip River Wetland. 
1938 1961 Approx. 2001 2003 2006
 A 6 19 24 25 27
 B 2.5 3 6 8 8
 C 2 2.5 8 10 10
 D 5 5 13 15 15
 E 0 1 11 13 13
 F 0 3 5 9 9
 A - Downstream of Olifantsvlei Water Treatment Plant
 B - Downstream of Golden Highway
 C - Downstream of N1
 D - Downstream of Bushkoppies Water Treatment Plant
 E - Downstream of R82
 F - At Zwartkopjes Pump Station
 Channel Width (m)Point
  
 40 
The first channel to be considered is at Point A (Figure 5.1.4), which today flows under the 
Golden Highway. This channel was mapped initially by Mellor in 1917 as a major crossing of 
the wetland. At Point A only a channel was developed and no marsh or reed beds were 
recorded (A in Figure 5.1.4). This channel increased in breadth by 21 m between 1938 (6 
m) and 2006 (27 m) (Table 5.1.2). The channel width at Point B increased only 5 m over the 
same time period (Table 5.1.2), even though Points A and B are in close proximity. Channel 
widths were measured at two locations (C and D in Figure 5.1.4) in Section 3 and were 
calculated to have increased by about 8 to 10 m between 1938 and 2006 (Table 5.1.2). 
Both, Sections 4 and 5 had no channels present in 1938 but measurements at E and F 
show channels about 9 to 13 m wide in 2006 (Table 5.1.2). The most noticeable change in 
Section 4 (Point E) is recorded between 1961 and 2003 where the channel width increased 
from 1 m to 11 m (Table 5.1.2).  
 
The channel at Point A, downstream of the Olifantsvlei Sewage Works, had the greatest 
growth with a 21 m widening over 68 years. Point E in Section 4 exhibits the second-largest 
increase of width, and Point A the largest at 13 m over the same time period. The intervals 
between the channel width measurements are not consistent. Consequently, the rates of 
change where calculated (Table 5.1.3) to provide a better indication of which channels are 
the worst impacted and over which periods of time. The channel at Point A widens at the 
greatest rate over periods 1938 to1961 and 2003 to 2006 (Table 5.1.3 and Figure 5.1.5). 
For the period 1961 to 2001 the channel at Point E enlarges at the greatest rate (Table 
5.1.3 and Figure 5.1.5). Point F however, changes at the greatest rate with a widening of 
2m/year between 2001 and 2003 (Table 5.1.3 and Figure 5.1.5). When calculating the 
average rate of change for the entire period, 1938 to 2006 and 1938 to 2003, Point F 
experiences the greatest rate of change (Table 5.1.3 and Figure 5.1.5).  
 
 41 
Table 5.1.3: Rates of change for channel widths for Points A to F. 
1938 to 
1961
 1961 to 
2001
 2001 to 
2003
 2003 to 
2006
 Average 
1938 to 2006
 Average 1938 
to 2003
 A 0.57 0.13 0.50 0.67 0.46 0.40
 B 0.02 0.08 1.00 0.00 0.27 0.37
 C 0.02 0.14 1.00 0.00 0.29 0.39
 D 0.00 0.20 1.00 0.00 0.30 0.40
 E 0.04 0.25 1.00 0.00 0.32 0.43
 F 0.13 0.05 2.00 0.00 0.55 0.73
 Point
 Rate of Channel Width Change (m/year)
  
 
0.00
 0.50
 1.00
 1.50
 2.00
 2.50
 A B C D E F
 Points
 Ra
 te
  
o
 f C
 ha
 n
 ge
  
(m
 /y
 ea
 r)
 1938 to 1961
 1961 to 2001
 2001 to 2003
 2003 to 2006
 Average 1938 to 2006
 Average 1938 to 2003
  
Figure 5.1.5: Rate and average rate (m/year) of change of channel width over time. 
 
It is important to consider the above changes in the wetland area and channel formation and 
widening in conjunction with the analysis of the aerial photographs, which is outlined in the 
following. 
 
 42 
5.1.1 Section 1: Lenasia to Olifantsvlei Water Treatment 
The Klip River Wetland first occurs in the Lenasia area of Johannesburg (Figure 5.1.1). The 
wetland in this reach is between 450 and 500 m in width and is a reed-covered swamp 
dominated by Phragmites australis (Figure 5.1.6). McCarthy et al. (2007) recorded flow 
velocities in Section 1 between 0 and 0.03 m/s, with flow depth rarely exceeding 20 cm. 
Peat in this area is about 4 m thick with an upper rhizome layer of 30 to 50 cm (McCarthy et 
al., 2007). McCarthy et al. (2007) also commented that although heavy metal concentrations 
in the peat are high, the vegetation does not appear to be adversely impacted. This section 
of the wetland is described as the best approximation of the appearance and hydrological 
functioning of the Klip River Wetland prior to the development of the Witwatersrand mining 
and industrial complex (McCarthy et al., 2007).  
 
 
Figure 5.1.6: The reed-covered swamp in the Lenasia area (450 m width) dominated by 
Phragmites australis (courtesy of T.S. McCarthy, November 2007). 
 43 
Anthropogenic impacts in this section appear to be minor, apart from the golf course 
development in the far west (excluded from analysis). Peat mining has also occurred in this 
section of the wetland east of Klipspruit Valley Road and has left large open water bodies 
(Figures 3.1.3 and 3.1.4). Olifantsvlei Sewage Works have been established in the lowest 
reaches of Section 1. McCarthy (op. cited) describes a contact dam from the treatment 
works, which has been constructed in the wetland itself.  
 
Prior to the summer rains in 2008, a white crust was seen on the banks of the Lenasia 
section of wetland (Figure 5.1.7). This occurrence is described in Naicker et al. (2003) as 
being attributable to the effect of capillarity and surface evaporation. Metal-rich ground 
water, which is drawn up by capillarity evaporates on the soil surface and produces a 
gypsum crust, which is enriched in metal sulfates. This effect is most noticeable in winter 
when evaporation exceeds rainfall.  
 
 
Figure 5.1.7: Gypsum crust accumulated on the banks of the Klip River Wetland at Lenasia. 
 
 44 
White (1957, page 13) also described this section of wetland: 
Soon after crossing the Johannesburg ? Potchefstroom Road the Klip River widens out 
considerably and flows through four or five miles, firstly of marshy land, and then of reed 
beds, until it is joined by the Klipspruit tributary. Small, good springs of clear water enter 
the Klip River in the marshy area. These springs are from dolomitic areas and, while their 
general tendency is to dilute the mine dump contributions, they increase the calcium 
content and raise the pH, with consequent fall in the iron content.  
 
A study in 1980 used the Klip River Wetland upstream and downstream of the Golden 
Highway to formulate an optimisation model for sewage treatment plants (Arnold, 1980). 
Arnold (1980) highlighted the presence of a large vlei at the confluence with the Klipspruit 
and upstream of the Golden Highway. A portion of the Klipspruit had been diverted into an 
open pond area which discharged via a sluice gate back into the main river immediately 
upstream of the bridge. This ponded area is the contact dam described by McCarthy et al. 
(2007).  
 
The geological mapping by Mellor (1917) does not illustrate any channels in the wetland in 
Section 1, except near the road crossing where Olifantsvlei is now built (B in Figure 5.1.8a). 
Examination of aerial photographs further support that in 1938 and 1961 channels were still 
absent in Section 1. However, as mentioned previously the channel at the road crossing 
increased in width by 21 m between 1938 and 2006 (Table 5.1.2 and Figure 5.1.9). The 
1938 and 1961 aerial photography shows the wetland to consist of an extensive reed bed 
with irrigation canals cut along the northern and southern banks (C and D in Figures 5.1.8b 
and c). These canals are still apparent in the 2006 photography (D in Figure 5.1.8d) but they 
have been heavily encroached by Phragmites reeds (McCarthy et al., 2007). At C (Figure 
5.1.8d) discharge into the Klipspruit tributary from the Goudkoppies Sewage Works appears 
to have led to the widening of an old irrigation furrow. This furrow was present prior to 1938 
but channelizing of the water only occurred after 1961 (Figures 5.1.8a to d).  
 
The building of the Olifantsvlei Treatment Works and the construction of the contact dam in 
the wetland occurred prior to 1961 (E in Figures 5.1.8c and d) but after 1957 (White, 1957). 
The maturation ponds built to the south of the wetlands were constructed after 1961 (F in 
Figure 5.1.8a to d). A portion of these ponds were evident by 1984, which is illustrated in 
Figure 5.1.10 (Arnold, 1980). 
 45 
 
Figure 5.1.8: The wetland in the Lenasia area in (a) 1917, (b) 1938, (c) 1961 and (d) 2006.
 46 
 
Figure 5.1.9: Point A: Channel cut after the Olifantvlei discharge point (Golden Highway in 
background; August 2008). 
 
Figure 5.1.10: Study site of Arnold (1980), showing the extent of the wetlands at this time, 
the maturation ponds and his sampling points (after Arnold, 1980). 
 47 
5.1.2 Section 2: Golden Highway to the N1 
Section 2 of the Klip River Wetland has been highly impacted by farming. In 1917, Mellor?s 
mapping depicts a continuous wetland after a short section of channel (Figure 5.1.11a). By 
1938 large tracts of the wetland were drained by furrows for development of agricultural 
land (A and B in Figures 5.1.11a to c). This is confirmed by the extent that the wetland 
decreased, 58 ha over 20 years (1917 to 1938; Table 5.1.1). White (1957) does not depict 
wetlands in this section either. Arnold (1980, page 59), however, does illustrate and 
describe wetlands in 1980 (Figure 5.1.10).  
Downstream of the Golden Highway bridge the stream follows a clear channel for about 
700 meters until the first heavily reeded vlei. The channel through this vlei is fairly well 
defined, although some deviations into the reeds occur, and after about a kilometer of 
reeds the river channel becomes clear again. A second major vlei is encountered about 
3.75 kilometers from the bridge. Here the channel gradually disappears as it proceeds 
along the southern side of the vlei discharging water at numerous points into the reeds on 
the left bank. Finally the channel becomes a small earth canal which continues to spill 
water over its left bank with very little flow eventually continuing in the canal. The river flow 
is thus well dispersed throughout the reeds.  
 
 48 
 
Figure 5.1.11: The wetland between the Golden Highway and the N1 in (a) 1917, (b) 1938 
and (c) 2006. 
 49 
Floodplain wetlands reappear further downstream in 1938. Point C in Figure 5.1.11 
illustrates that this area has increased in size between 1917 and 1938 and continues to 
grow into 2006. This is likely due to pooling of water behind a road in 1938 and the N1 after 
1961 (Figure 5.1.12). 
 
According to the geological map (Mellor, 1917) Section 1 of the wetland had a single 
channel in 1917 (Figure 5.1.11a). Potentially farming was occurring in these parts by 1917 
and this channel actually represents a furrow cut by the farmers. The channel of 1917 
however, does not match any of the 1938 furrows. Furrows were cut by 1938 on the north 
and south banks of the wetland (D and E in Figures 5.1.11b and c). The furrow to the south 
becomes the preferred channel by 1984. The central channel present in 1938 (Figure 
5.1.12a) propagates through the wetland and connects with the south bank furrow after 
1961 (Figure 5.1.12b). The bridge for the N1 Highway was built to accommodate this central 
channel in 1979 (Arnold, 1980). The southern branch of this channel propagates into the 
wetland by about 200 m between 1938 and 1961 (Figure 5.1.12b). This branch extends a 
further 770 m by 2003 to connect with the furrow on the southern bank (Figure 5.1.12c). The 
northern branch of the channel in 1938 grows over by 1961 (Figure 5.1.12b). 
 
 50 
 
Figure 5.1.12: Detail of the wetland at the N1 Highway in (a) 1938, (b) 1961, (c) 2003 and 
(d) 2006. 
 
5.1.3 Section 3: N1 to Kibler Park 
Similar to Section 2, Section 3 is highly impacted by farming prior to 1938. The wetlands 
were completely drained with the use of furrows (A to D in Figure 5.1.13) to convert the 
wetland into agricultural land (E to H in Figure 5.1.13). In 1917 this was a continuous section 
of wetland almost 200 ha in extent and by 1938 only 39 ha remained. The wetland 
decreased in size by 179 ha in 20 years (Table 5.1.1). White does not depict wetlands in 
this section in 1957 either. The central channel in Figure 5.1.12 and I in Figure 5.1.13 
(discussed in the previous section) is found to be derived from a furrow cut on the northern 
bank of the wetland prior to 1938 (A in Figures 5.1.13b to d). The furrows have therefore 
become connected between sections and the water has been channelised away from the 
wetland. The channel depicted in 1917 is the preferred channel today (I in Figure 5.1.13).  
 
 51 
 
Figure 5.1.13: The wetland between the N1 Highway and Kibler Park in (a) 1917, (b) 1938, 
(c) 1961 and (d) 2006. 
 52 
As a result of Mellor?s attention to detail, a pond was illustrated in 1917at Point J, Section 3 
(Figure 5.1.13a). This ponding is not evident at all by 1938. A fork in a channel mapped by 
Mellor (1917) is evident in 1938, 1961 and 2006 (C in Figure 5.1.13). Channels mapped by 
Mellor in 1917 match furrows in 1938, for example Point K in Figure 5.1.13.  
 
Flow has become confined to the irrigation canal along the northern bank at Point D 
(Figures 5.1.13b to d). This irrigation canal propagated upstream and southwards by 500 m 
between 1938 and 1961 to connect with another irrigation furrow on the south bank (L in 
Figure 5.1.13). Toward Kibler Park, agriculture has declined over time. Between 1961 and 
1984 the irrigation canals became dysfunctional and reeds have over grown some of these 
furrows as they still hold water (Figures 5.1.14a and b). A gypsum crust (discussed in 
Section 5.1.1) is also evident between the N1 and Kibler Park (Figure 5.1.14c), which gives 
an indication of the mine pollutants in the groundwater and soils. The Bushkoppies Sewage 
Works, built after 1961, decants east of the N1 into the wetland. Unlike the Olifantsvlei 
Sewage Works outlet, no single channel has formed. The former sewage works discharge 
diffuses through the wetland in multiple small channels (Figure 5.1.14d). 
 
 
 53 
ba
 c Bushkoppies
 Inlet
 Diffuse Flow
 d
  
Figure 5.1.14: Field and aerial reconnaissance photos between the N1 Highway and 
Lenasia (d is courtesy of T.S. McCarthy, November 2007). 
 
 54 
5.1.4 Section 4: Kibler Park to Kromvlei 
Section 4 represents the largest portion of the Klip River Wetland and has been negatively 
affected tremendously since 1917. For ease of discussion Section 4 has been split into 
three parts as seen in Figure 5.1.15.  
 
Figure 5.1.16
 Figure 5.1.17
 Figure 5.1.19
  
Figure 5.1.15: The breakdown of Section 4 as to be discussed from Jackson?s Drift (Kibler 
Park) to Kromvlei.  
 
Near Jackson?s Drift, Mellor (1917) mapped the channel in Section 3 to continue under the 
bridge for a short distance downstream until the flow dispersed into the wetlands of Section 
4 (A in Figure 5.1.16a). This channel on the northern bank in 1917 evolves into the main 
channel in 1938 and this remains the case today (A in Figure 5.1.16e). The wetland south of 
Jackson?s Drift was completely drained by 1938 for agriculture (B in Figure 5.1.16b) using 
furrows cut to the south (C and D in Figure 5.1.16b to d). Furrows cut into the north and 
south bank prior to 1938 represent the main water conduits today (D and E in Figure 
5.1.16). 
 55 
 
Figure 5.1.16: The wetland at Jackson?s Drift Bridge in (a) 1917, (b) 1938, (c) 1961, (d) 
2003 and (e) 2006.  
 
 56 
East (downstream) of the Jackson?s Drift Bridge the 1917 mapping illustrates a channel free 
wetland (Figure 5.1.17a). By 1938 several irrigation canals were cut on the north and south 
banks (A to C in Figure 5.1.17b). Similar to Sections 2 and 3, this was done to drain the 
wetland for agriculture (D in Figures 5.1.17b and c). By 2003 most of these irrigation canals 
are derelict and farming is not very prevalent. Another means of draining the wetland was to 
cut a diagonal drain across the wetland, which then collected water that flowed through the 
wetland and diverted it into a canal on the bank. The drain at E (Figure 5.1.17) was cut prior 
to 1938 and was overgrown in 1961 and 1984. The 2003 and 2006 photographs show a 
number of diffuse channels which eroded head-ward into the wetland, upstream of the drain 
(F in Figures 5.1.17d and e, Figure 5.1.18) and by 2003 the drain had connected to the 
north bank irrigation canal (Figure 5.1.17). In 2006 this was the main water conduit 
canalizing water from Jackson?s Drift through the irrigation canals across the wetland (in the 
drain) and into a south bank furrow. 
 
 57 
 
Figure 5.1.17: The wetland south of Kibler Park in (a) 1917, (b) 1938, (c) 1961, (d) 2003 
and (e) 2006.  
 
 58 
 
Figure 5.1.18: Diffuse channels as seen in an aerial survey eroding upstream of the drain 
(courtesy of T.S. McCarthy, November 2007).  
 
The eastern section of the wetland, toward Kromvlei, was an expansive wetland in 1917 
covered by Phragmites and with no channels evident (Figure 5.1.19a). In 1917 the wetland 
was in places more than 900 m in width. It appears that flow was distributed across the 350 
to 900 m width up until 1961 (Figure 5.1.19d), much like the wetlands in Lenasia (Section 
5.1.1). As seen in Sections 2 and 3, entire portions of the wetland were drained completely 
between 1917 and 1938 (A and B in Figure 5.1.19) by irrigation canals on the north and 
south banks (C to G in Figure 5.1.19b). Many of these canals were supplied water by drains 
cut across the wetland (H to J in Figure 5.1.19). The drain at H was cut prior to 1938 and, 
although it was overgrown in the 1952 photographs, it propagated head-ward by 400 m in 
1961. This feeder channel eroded head-ward from the drain to almost connect to the 
northern bank irrigation canal by 1984 (Figure 5.1.19e). The feeder channel, however, 
continued to erode head-ward and bypassed the northern irrigation canal to connect to 
another drain at Point I prior to 2003 (Figure 5.1.19f). Between 1938 and 1952 another large 
diagonal drain was cut across the wetland to divert the water and drain the wetland for 
 59 
irrigation in the east as seen at K (Figure 5.1.19). A south bank furrow cut prior to 1938 to 
divert all the flow for irrigation to the south of Kromvlei develops into major conduit for the 
entire Section 4 wetland (L in Figures 5.1.19b to g). This furrow was cut in order to drain the 
wetland for agriculture downstream of Kromvlei in the south. Through the connection of all 
these drains and irrigation furrows a continuous channel has formed throughout Section 4. 
A definite impact on this section of wetland is observed where soil is exposed and no reeds 
are growing (M in Figures 5.1.19f and g). Another illustration of the impact is the decrease in 
wetland area by more than 219 ha (Table 5.1.1). 
 
 60 
a
 b
 c
 d
 e
 f
 g
  
Figure 5.1.19: The wetland downstream of Kibler Park in (a) 1917, (b) 1938, (c) 1952, (d) 1961, (e) 1984, (f) 2003 and (g) 2006.  
 61 
5.1.5 Section 5: Zwartkoppies Section 
The geological map from 1917 (Mellor) does not cover the full extent of the Zwartkoppies 
Section. In 1917 a channel had developed near the Zwartkoppies Pump Station and 
continued beyond the mapped extent (A in Figure 5.1.20). By 1938 the wetland had already 
shrunk in extent and was drained in sections for agriculture (B in Figure 5.1.20) by irrigation 
canals (C and D in Figure 5.1.20). Other anthropogenic impacts included an elevated road 
(bridge; E in Figure 5.1.20) across the wetland to the pump station and an underground 
pipeline across the wetland (F in Figure 5.1.20). Both of these were constructed prior to 
1938. A number of openings were incorporated under the road in order to allow through flow 
of water. McCarthy et al. (2007) identified that one of the culverts under the roadway is 
today deeply incised and is carrying all of the discharge, with the remaining culverts no 
longer functioning. It is also evident that the reed bed in the south of this section of wetland 
is receding upstream (G in Figure 5.1.20). The wetland was described by Humphrey (1910) 
as a series of extensive marshes that extended to within 1.6 km of the Klip River Railway 
Station (south of Zwartkoppies Pump Station). Today the wetland terminates more than 6.5 
km from this railway station, which means that the wetland has retreated 5 km upstream 
(Figure 5.1.21). The 1938 photography does not extend to the railway station. However, it is 
suggested that this retreat had occurred by 1938. By 1961 the southern extent of the 
wetland was completely canalized and there was very little evidence of the wetlands (Figure 
5.1.22). In the field, peat burns are evident. As a result of the burns, the soil becomes a 
clayey ash, which is then washed into the river system. Remaining are donga-like gullies in 
which bedrock is often exposed (McCarthy et al., 2007).  
 
 
 62 
 
Figure 5.1.20: The wetland of Section 5 in (a) 1917, (b) 1938 and (c) 2003 (a and b have the same scale as c). 
 63 
 
Figure 5.1.21: Wetland retreat from the Klip River Railway Station between 1910 and 2003. 
 64 
 
Figure 5.1.22: Wetland retreat from the Klip River Railway Station between 1910 and 1961. 
 65 
After Kromvlei the wetland is channelized at the south bank as it enters the Zwartkoppies 
Section (A in Figure 5.1.23). From 1938 to 1952 the channel cuts down and extends 
downstream 200 m in 14 years. The channel propogates a further 20 m/year between 1952 
and 1961, and advances downstream a swift 1 200 m between 1961 and 1984, a rate of 57 
m/year. The channel pirates an irrigation canal on the western bank by 1984 (B in Figure 
5.1.23). At the same time a channel to the south (C in Figure 5.1.23) was propagating 
upstream. It is unclear whether the two streams connect by 1984 but by 2003 the channels 
had broken though the road (D in Figure 5.1.23) to connect. Today there is one continuous 
channel throughout this section of wetland. The channels are still flanked by reed beds, but 
these appear to experience only limited inundation.  
 
South of the Zwartkoppies Pump Station another channel was advancing upstream toward 
the pipeline and the channels to the north (A in Figure 5.1.24). Table 5.1.4 provides the rate 
of head-cutting of this southern channel.  
 
Table 5.1.4: Advance rate of the channel south of the Zwartkoppies Pump Station. 
Years Advance of head-cutting
 Rate of 
head-
 cutting
 1938 - 1952 127m 9m/year
 1952 - 1961 124m 14m/year
 1961 - 1984 912m 43m/year
 1984 - 2003 784m 37m/year
  
 
The incising channel intersected a bedrock barrier in around 1984, resulting in a 2 m high 
waterfall (C in Figure 5.1.24). The last remaining unchannelised portion of the Zwartkoppies 
section (B in Figure 5.1.24) will soon converge to one channel as seen in Figure 5.1.25. 
 
 66 
 
Figure 5.1.23: The wetland at Zwartkoppies Pump Station in (a) 1938, (b) 1952, (c) 1961, (d) 1984 and (e) 2003. 
 
 
Figure 5.1.24: The wetland south of the Zwartkoppies Pump Station in (a) 1938, (b) 1952, (c) 1961, (d) 1984 and (e) 2003. 
 
 67 
 
Figure 5.1.26: A 2 m high waterfall on bedrock (courtesy of T.S. McCarthy, November 2007). 
 
Figure 5.1.25: Aerial photograph of the final stages of the formation of a single channel in the 
Zwartkoppies section of the wetland. 
 68 
5.2 Long River Profiles 
Arnold (2004) used 1: 10 000 orthophoto contours to determine the long river profile for the Klip 
River Wetland (Figure 5.2.1). The knick points observed in the field and documented by McCarthy 
et al. (2007) however, were not evident in the profile. A higher resolution topography dataset was 
required. The LIDAR data flown in 2006 produced 25 cm contours and this was used to document 
topographic changes along the wetlands length, i.e. a long river profile was constructed. A profile 
was delineated through the channel as well as the wetland (Sections 4 and 5) to determine the 
gradients of each. This is crucial in determining the reference gradient for the Klip River Wetland, 
which in turn will determine the gradient that must be attained in any remediation efforts. 
 
 
1420
 1470
 1520
 1570
 1620
 1670
 1720
 0102030405060708090100
 Distance km
 El
 ev
 at
 io
 n
  
(m
 am
 sl
 )
 Karroo
 Ventersdorp
 Chunisport
 Pretoria
 Witw atersrand
 Turffontein
 Legend:
 Wetlands
 Tributaries
 Rietspruit
 Klipspruit
 Harringtonspruit
 Dyke
 Princess Dam
 Klipriviersberg Lavas
 Karoo Supergroup
 Ventersdorp Supergroup
 Chuniespoort Group
 Pretoria Group
 Witwatersrand Supergroup
 Transvaal
 Supergroup
 Wetlands
 Tributaries
 Dyke
 El
 ev
 at
 io
 n
  
(m
 am
 sl
 )
  
Figure 5.2.1: Long river profile as recorded from 1 : 10 000 orthophotos (after Arnold, 2004). 
 69 
5.2.1 Section 1: Lenasia to Olifantsvlei Water Treatment 
The profile of Section 1 deduced from the LIDAR data reveals that the gradient increases 
downstream of the entry of the Klipspruit tributary (Figure 5.2.2). A cutting down of 3 m is 
recorded in the profile at the outflow of the Olifantsvlei Sewage Works (Figure 5.2.2) as large 
quantities of treated sewage water with little sediment load is discharged into the wetland (Figure 
3.4.1). The peat mining and contact dam can also be identified in the profile (Figure 5.2.2). Less 
clear is the groundwater barrier (dyke) depicted by Kafri and Foster (1989). The gradient of the 
wetland west of the Klipspruit Valley Road (unchannelised section of the wetland), i.e. a reference 
gradient for the Klip River Wetland, is 1 : 0.001. West of the Klipspruit Valley Road the gradient 
increases to 1 : 0.004.
 70 
 
Figure 5.2.2: Long river profile from the road east of the golf course to the Golden Highway in relation to 2006 
photography.
 71 
5.2.2 Sections 2 and 3: Golden Highway to Kibler Park 
Three rapids are evident in Section 2 (Points 1 to 3 in Figure 5.2.3). The gradient of the long 
river profile increases to 1 : 0.0015 in this section and cutting down is evident at Rapid 2 (1.5 
m) and 3 (1 m), (Figure 5.2.3).  
 
The largest down cutting in the profile of Section 3 can be seen at Rapid 4 (Figure 5.2.4). After 
being piped under an old road, the river cuts down by 1.8 m (Figures 5.2.4 and 5.2.5a). Rapids 4, 
5, 6 and 8 can be related to groundwater barriers (dykes) as defined by Kafri and Foster (1989) 
and Schweitzer et al; (2004, Figure 5.2.4). Three rapids (5, 7 and 8) were evident in 2003 and in 
2006 and no movement was recorded in the three years. The other rapids evident in 2006 could 
not been seen on the 2003 photography due to the low resolution. After 1984 a bridge was built at 
Rapid 9 for access to a home. This bridge has caused a dam to form south of the main channel 
(Figures 5.2.5b and c). A cut in the profile of 1 m has formed at this bridge (9 in Figure 5.2.4 and 
Figure 5.2.5d). Between Sections 2 and 3 an increase in gradient to 1 : 0.002 is observed.
 72 
B1
 2
 3
 1528
 1530
 1532
 1534
 1536
 1538
 208002180022800238002480025800
 Distance Along the Profile
 E
 l
 e
 v
 a
 t
 i
 o
 n
  
(
 m
 a
 m
 s
 l
 )
 Golden 
Highway
 N1
 1m
 1.5m
 E
 l
 e
 v
 a
 t
 i
 o
 n
  
(
 m
 a
 m
 s
 l
 )
 E
 l
 e
 v
 a
 t
 i
 o
 n
  
(
 m
 a
 m
 s
 l
 )
 E
 l
 e
 v
 a
 t
 i
 o
 n
  
(
 m
 a
 m
 s
 l
 )
 E
 l
 e
 v
 a
 t
 i
 o
 n
  
(
 m
 a
 m
 s
 l
 )
 E
 l
 e
 v
 a
 t
 i
 o
 n
  
(
 m
 a
 m
 s
 l
 )
  
Figure 5.2.3: Long river profile from the Golden Highway to the N1 in relation to 2006 photography. 
 73 
  
Figure 5.2.4: Long river profile from the N1 Highway to Kibler Park in relation to 2006 photography.
 74 
 
Figure 5.2.5: Field and aerial reconnaissance photos of the rapids between the N1 Highway and 
Lenasia (c courtesy of T.S. McCarthy, November 2007). 
 
5.2.3 Sections 4 and 5: Kibler Park to Zwartkoppies  
The long river profile was drawn for Sections 4 and 5 combined (Figure 5.2.6). A profile was also 
constructed through the wetland to determine the difference in gradient between the wetland 
and the channel. In the west, where the wetland and channel are at the same elevation, the 
gradient is approximately 1 : 0.0026. In the east, where the channel cuts through the wetland by 
more than 5 m, the gradient increases to more than 1 : 0.0036. The rapids at Point 14 correlate 
with a groundwater barrier (dyke) as described by Kafri and Foster (1989). Rapid?s 13, 14, 16 
and 17 were evident in 2003 and 2006. After the rapids at 15 (Figure 5.2.6) the channel incises 
through the wetland and the difference in elevation between the bed of the channel and the 
wetland surface increases from 2 m to 5 m over about a kilometer distance. A major knick point 
 75 
is highlighted at Point 16 by a cut of over 4 m. According to the aerial photography this knick 
point eroded back 85 m in three years between 2003 and 2006 (Figure 5.2.7). Field 
reconnaissance has shown that this rapid has migrated a further 118 m between 2006 and 
2008 (Figure 5.2.7), which means that the rate of knick point migration is increasing. Across 
the entire wetland this is the only knick point which is migrating at a significant rate. This is likely 
due to the fact that the knick point is eroding through soil and not bedrock (Figure 5.2.8).The 
Rapids at 17 are also significant as they have incised into the wetland about 2 m. 
Communications with Terrance McCarthy (2007) indicated that peat burns are occurring around 
this knick point as well. McCarthy et al., (2007) calculated that the water surface near Rapid 17 
has lowered by 2.5 m since 1983. Rapid?s 8 and 9 from the 2003 photographs could be related 
to the dykes interpreted by Schweitzer et al., (2004).  
 
 76 
 
Figure 5.2.6: Long river profile from the Kibler Park to Zwartkoppies Pump Station in relation to 2006 photography.
 77 
 
Figure 5.2.7: Knick point migration between 2003 and 2008 at Point 16. 
 78 
 
Figure 5.2.8: Field photos of the knick point eroding through soil at Point 16 (Figure 5.2.6). 
 
 79 
5.3 Digital Terrain Model 
A Digital Terrain Model (DTM) was created from the LIDAR contour data and from the original 
point data. The point data were obtained because the resolution of the contour data was 
insufficient (Figure 5.3.1) and gaps in the dataset became evident in the DTM. The gaps became 
more prominent as the cell size was reduced and the resolution increased. 
 
 
Figure 5.3.1: DTM derived from the contour LIDAR data at 40 m resolution. 
 
The LIDAR point data were previously separated into ?ground? and ?non-ground? data using 
algorithms tailored to suit the project as a whole (Appendix 1). The majority of the wetland data 
were within the ?non-ground data?. The two datasets were merged in order to produce a complete 
high resolution DTM with a 10 m pixel size (Figure 5.3.2). Open water bodies such as dams or 
ponds appear to have been blanked out of the dataset. This is possibly due to the back scatter or 
absorption of the laser strikes.  
 
 80 
 
Figure 5.3.2: DTM derived from the original LIDAR point data at 10 m resolution. 
 
Hillshading was applied to the DTM raster dataset, which predicts where shadows exist in the 
DTM, depending on the origin (direction) of the light source and the elevations that exist. A very 
low sun angle was chosen (5?) in order to identify slight variances in this floodplain terrain (Figure 
5.3.3). The hillshade of the 10 m DTM was able to provide some insight into the topographical 
constraints on the wetland. The Black Reef to the north of the wetland, which is extremely clear in 
the hillshade, is acting as a physical constraint on the northern limit of the wetland. Another 
topographical constriction, possibly a dyke, is evident running under the Golden Highway. This 
could explain why even in Mellor?s time there was no wetland in this segment. Groundwater 
barriers as described by Kafri and Foster (1989) are not manifested in the DTM or the hillshade.  
 
 81 
Black
  Reef
 Dyke?
  
Figure 5.3.3: Hillshade at 5? sun angle derived from the DTM at 10 m resolution. 
 
5.4 Water Chemistry 
Often referred to as the ?kidneys? of the environment, wetlands filter the waters of rivers and 
streams, therefore reducing pollution. The vegetation in wetlands removes phosphates and other 
plant nutrients washed in, thereby slowing the growth of algae and aquatic weeds wetlands. With 
the degradation and diminishing extent of the Klip River Wetland, it would be expected that the 
water quality downstream should also deteriorate.  
 
As mentioned before the major pollutants entering the Klip River are sewage and acid mine 
drainage. The chemical signatures of these two pollution sources are phosphates, nitrates and 
sulfates. These chemical components are monitored by both Rand Water and DWAF on a regular 
basis. The Rand Water Monitoring Station (K21; Figure 4.1) data covers the period 1992 to 2006. 
Phosphates (Figure 5.4.1) and sulfates (Figure 5.4.2) were found to decrease marginally over this 
time period while nitrate levels increased (Figure 5.4.3). The sulfate levels are likely to decrease 
 82 
due to the dilution effect of the sewage effluent. Dissolved oxygen was also considered, and is 
seen to be decreasing over time (Figure 5.4.3). Low dissolved oxygen levels indicate an 
excessive demand on the oxygen of the system. The build up of organic material from human 
activities is one source of oxygen depletion. Micro-organisms in the stream consume oxygen as 
they decompose sewage, urban and agricultural runoff. Acid mine drainage is also known to 
produce direct chemical demands on oxygen in the water. The only indication that water quality is 
deteriorating over time is through the increased nitrate levels and decreasing oxygen levels.  
 
0
 0.2
 0.4
 0.6
 0.8
 1
 1.2
 1.4
 1.6
 1.8
 2
 06-Apr-92 01-Jan-95 27-Sep-97 23-Jun-00 20-Mar-03 14-Dec-05
 Date
 PO
 4 
(m
 g/
 L)
 Unacceptable
  
Figure 5.4.1: K21 phosphate levels from 1992 to 2006 (acceptability according to DWAF 
domestic water quality standards). 
 83 
0
 50
 100
 150
 200
 250
 300
 350
 400
 06-Apr-92 01-Jan-95 27-Sep-97 23-Jun-00 20-Mar-03 14-Dec-05
 Date
 SO
 4 
(m
 g/
 L)
 SO4 (SO4 mg/l)
 Linear (SO4 (SO4 mg/l))
  
Figure 5.4.2: K21 sulfate levels from 1992 to 2006. 
 
0
 5
 10
 15
 20
 19-Sep-91 15-Jun-94 11-Mar-97 06-Dec-99 01-Sep-02 28-May-05
 Date
 NO
 3 
(m
 g/
 L)
 NO3 (NO3 mg/l)
 Linear (NO3 (NO3 mg/l))
 Unacceptable
  
Figure 5.4.3: K21 nitrate levels from 1992 to 2006 (acceptability according to DWAF domestic 
water quality standards). 
 84 
0
 5
 10
 15
 20
 25
 06-Apr-92 01-Jan-95 27-Sep-97 23-Jun-00 20-Mar-03 14-Dec-05
 Date
 Di
 ss
 o
 lv
 ed
  
O
 x
 yg
 en
  
(D
 O
  
m
 g/
 l O
 2)
 Disolved Oxygen (DO mg/l O2)
 Linear (Disolved Oxygen (DO mg/l O2))
  
Figure 5.4.4: K21 dissolved oxygen levels from 1992 to 2006. 
 
From the aerial photography analysis it is evident that the Klip River Wetland experienced the 
majority of its degradation prior to 1992 and thus these results are inconclusive. Attempts were 
made to obtain data prior to 1992. White (1957) provides some water quality analysis of the Klip 
River above the confluence of the Natalspruit (comparable to Station K21) during the dry season. 
The range of sulphate levels for example where recorded at 80 to 800 ppm in 1957. White (1957) 
did, however, comment that the Klip and Suikerbosrand Rivers were in 1957 contaminating the 
Vaal River though heavy minerals from mining activities at their headwaters.  
 
5.5 Water Discharge 
With the deterioration of the wetlands, it would be expected that the water discharge in the Klip 
River would increase (similar to the argument for the water chemistry discussion). This is due to 
the fact that wetlands act as a flood attenuation mechanism and slow the release of the water into 
the system. Water that previously was forced to flow through the wetland is evidently (from 
Section 5.1) being directed straight into the interconnected furrows and transported quickly out of 
the system.  
 85 
Water discharge data were supplied by DWAF for the Witkop Station downstream of where the 
Rietspruit enters the Klip River. The dataset spans from 1977 to 2006. However, the data was 
sparse from 2004 onwards. From Figures 5.5.1 to 5.5.3 it is evident that an increase in discharge 
over time occurred at this station. It is difficult to attribute this increase in discharge to the 
decreasing impact of the wetlands only. With the rapid growth of the city of Johannesburg it is also 
logical that the sewage input increases over time. Urban expansion has also caused decreased 
infiltration and faster runoff due to low infiltration surfaces (i.e. roads, gutters, etc). It is however, 
noted that the flood peaks in February and March (rainy season) have increased over time and 
this can possibly be attributed to the decreased residence time in the wetlands and reduced 
capacity of the wetlands for flood attenuation. Another consideration  
 
5
 7
 9
 11
 13
 15
 17
 19
 21
 1977 1982 1987 1992 1997 2002
 Year
 Su
 rfa
 ce
  
W
 at
 er
  
Di
 sc
 ha
 rg
 e 
(m
 3/
 s)
 Mean Wet Season (October to March)
 Mean Dry Season (April to September)
  
Figure 5.5.1: Mean wet and dry season trends for water discharge data at Witkop Station 
(C2H141) between 1977 and 2006. 
 
 86 
February
 0
 5
 10
 15
 20
 25
 30
 35
 40
 45
 19
 79
 19
 80
 19
 81
 19
 82
 19
 83
 19
 84
 19
 85
 19
 86
 19
 87
 19
 88
 19
 89
 19
 90
 19
 91
 19
 92
 19
 93
 19
 94
 19
 95
 19
 96
 19
 97
 19
 98
 19
 99
 20
 00
 20
 01
 20
 02
 20
 03
 20
 04
 20
 05
 20
 06
 Year
 Su
 rfa
 ce
  
W
 at
 er
  
Di
 sc
 ha
 rg
 e 
(m
 3/
 s)
  
Figure 5.5.2: Water discharge in February at Witkop Station (C2H141) between 1979 and 2006. 
 
June
 0
 2
 4
 6
 8
 10
 12
 14
 16
 18
 20
 19
 77
 19
 78
 19
 79
 19
 80
 19
 81
 19
 82
 19
 83
 19
 84
 19
 85
 19
 86
 19
 87
 19
 88
 19
 89
 19
 90
 19
 91
 19
 92
 19
 93
 19
 94
 19
 95
 19
 96
 19
 97
 19
 98
 19
 99
 20
 00
 20
 01
 20
 02
 20
 03
 20
 04
 20
 05
 20
 06
 20
 07
 Year
 Su
 rfa
 ce
  
W
 at
 er
  
Di
 sc
 ha
 rg
 e 
(m
 3/
 s)
  
Figure 5.5.3: Water discharge in June at Witkop Station (C2H141) between 1977 and 2006. 
 87 
The Witkop discharge data were also compared to rainfall data recorded at OR Tambo Airport 
(Figure 5.5.4). Although the rainfall data is from a separate catchment the two data sets follow 
similar trends. Around 2006 the discharge increases significantly, with no relationship to the 
rainfall data being evident.  
 
Rainfall & Discharge
 0
 20
 40
 60
 80
 100
 120
 140
 160
 19
 78
 19
 79
 19
 80
 19
 81
 19
 82
 19
 83
 19
 84
 19
 85
 19
 86
 19
 87
 19
 88
 19
 89
 19
 90
 19
 91
 19
 92
 19
 93
 19
 94
 19
 95
 19
 96
 19
 97
 19
 98
 19
 99
 20
 00
 20
 01
 20
 02
 20
 03
 20
 04
 20
 05
 20
 06
 20
 07
 20
 08
 20
 09
 Year
 Ra
 in
 fa
 ll 
(m
 m
 )
 0
 5
 10
 15
 20
 25
 Di
 sc
 ha
 rg
 e 
(m
 3 /s
 )
 Rainfall
 Discharge
  
Figure 5.5.4: Water discharge at Witkop Station (C2H141) between 1977 and 2006 compared to 
rainfall data (OR Tambo Airport). 
 
Figure 5.5.5 considers the change in rainfall and discharge data per month during the periods 
1978 to 1987, 1988 to 1997 and 1998 to 2006. The rainfall trends remain consistent over all three 
of these time periods. The discharge data however, indicate that discharge is increasing over 
time, most noticeably in the dry winter months.  
 88 
Rainfall vs Discharge
 0
 20
 40
 60
 80
 100
 120
 140
 160
 OCT NOV DEC JAN FEB MAR APR MAY JUN JUL AUG SEP
 Month
 Ra
 in
 fa
 ll 
(m
 m
 )
 0.000
 2.000
 4.000
 6.000
 8.000
 10.000
 12.000
 14.000
 16.000
 18.000
 20.000
 Di
 sc
 ha
 rg
 e 
(m
 3 /s
 )
 1978-1987 Rainfall
 1988-1997 Rainfall
 1998-2006 Rainfall
 1978-1987 Discharge
 1988-1997 Discharge
 1998-2006 Discharge
  
Figure 5.5.5: Water discharge at Witkop Station (C2H141) per month compared to rainfall data 
(OR Tambo Airport). 
 
6. Discussion and Conclusions 
6.1 Wetland Degradation 
Prior to 1938, water from the Klip River Wetland was used extensively for agriculture. The 
agricultural practise was to drain the wetland by means of canals and furrows to then develop the 
land. Furrows, canals and drains were also used to irrigate farm lands. The majority of the 
irrigation canals incising along the banks of the wetland and the drains that cut across the wetland 
have now connected to form a single main channel diverting water away from the wetland. 
Analysis of aerial photographs reveals that only one small section (500 m long) of the wetland 
near Zwartkoppies Pump Station has not formed a single channel yet. The furrows and drains cut 
in the early part of the century have provided the perfect conduit for the large quantities of sewage 
and runoff funnelled into the wetland. The growing city of Johannesburg has caused the sewage 
influx in the river to increase resulting in the pirating of the otherwise harmless channels.  
 
 89 
The only section of wetland which has not become channalised is in the Lenasia section west of 
the Klipspruit Valley Road. This section of wetland, however, is not pristine. The salt crust on the 
wetland banks indicates that the water quality is poor and metal rich, due to mine discharge. Up 
until 1961 the flow was across the entire width of the wetland at Kromvlei, similar to Lenasia. This 
is no longer the case as a channel on the south bank cuts through the wetland, incised by more 
than 5 m. Peat burns have been documented in Sections 4 and 5 of the wetland. Subsequently, 
soils containing heavy metals are being washed into the system and transported to the Vaal 
River.  
 
With the degradation of the wetland, it was expected that the downstream water quality would 
deteriorate and that water discharge would increase. The only indications that water quality is 
deteriorating over time are increased nitrate and decreased oxygen levels. Since the wetland 
experienced the majority of its decline prior to 1992, the extent of the water quality data is 
insufficient to determine when the water quality was affected by the lack of water purification. An 
increase in discharge over time has occurred at the Witkop Station. It is difficult to attribute this 
increase in discharge to the decreasing impact of the wetlands only. With the rapid growth of the 
city of Johannesburg it is also logical that sewage input increased over time and ideally sewage 
discharge data needs to be acquired to better understand this relationship. 
 
When considering the long river profile from Lenasia to the Zwartkoppies Pump Station, two major 
knick points are identified. The first is located at the outlet of the Olifantsvlei Sewage Works and 
the second represents a major rapid near Kromvlei. The latter has incised by 5 m into the wetland 
(Figure 5.2.8). Channel incision at the Olifantsvlei outlet is also significant (3 m). This channel has 
widened over time but the knick point has not migrated because it currently rests on bedrock (a 
resistant dyke as determined from the LIDAR). If it were to migrate upstream it would intersect the 
weir and then the man-made pond, which means the unchannelised wetlands at Lenasia will be 
protected from this knick point.  
The knick point of concern is therefore in Kromvlei and the dashed line in Figure 6.1 illustrates the 
expected incision potential, assuming that this knick point does not intersect a resistant bedrock 
barrier.  
 90 
Sections 1 to 5
 1480
 1490
 1500
 1510
 1520
 1530
 1540
 1550
 1560
 3800880013800188002380028800
 Distance along Profile
 El
 ev
 at
 io
 n
  
(m
 am
 sl
 )
 Olifantsvlei Outlet 
into the Klip River
 Major Knickpoint
 El
 ev
 at
 io
 n
  
(m
 am
 sl
 )
  
Figure 6.1: Long river profile from Lenasia to the Zwartkoppies Pump Station. 
 
6.2 Rate of Wetland Degradation 
Measuring the change in wetland area over time is not an effective method of quantifying the 
extent of wetland degradation and associated rates of change. Although, wetland area does 
decrease dramatically between 1917 and 1938 (Sections 2 to 5 in Figure 6.2), some wetland 
areas increase between 1938 and 2006 (Sections 2 and 3 in Figure 6.2). The wetland is therefore 
receiving sufficient moisture from rain, surface runoff and groundwater for the majority of the reed 
beds to remain intact. Channel formation, propagation and widening along with knick point 
migration prove to be the best methods of quantifying the collapse of the wetland (Figures 6.3 and 
6.4). Furrows active in 1938 and 1961 are today derelict (Figures 6.3 and 6.4). No knick point 
migration is recorded between 2003 and 2006 except near Kromvlei at Point 16 (Figures 6.3 and 
6.4). This knick point migrated 118 m in two years. This is likely due to the fact that this knick point 
is eroding through soil and not bedrock. It was also determined that groundwater barriers (dykes), 
as identified by Kafri and Foster (1989) and Schweitzer et al. (2004), could be correlated to rapids. 
 
 
 91 
 
Figure 6.2: Wetland area in (a) 1917, (b) 1938, (c) 1961, and (d) 2006. 
 92 
 
Figure 6.3: Channel formation, propagation and widening from the Golden Highway to Kibler Park between 1917 and 2003. 
 93 
 
Figure 6.4: Channel formation, propagation, widening and knick point migration from Kibler Park to the Zwartkoppies Pump Station between 
1917 and 2006. 
 94 
The channel downstream of the Olifantsvlei Sewage Works widened most significantly (21 m), 
when compared to the remaining five channel width measuring stations, due to large amounts of 
sewage inflow. Point F near the Zwartkoppies Pump Station widened at the greatest rate between 
1938 and 2006 (0.73 m/year, Figure 6.4). The period of most significant change for all monitoring 
points was between 2001 and 2003 (Tables 5.1.2 and 5.1.3, Figures 6.3 and 6.4).  
 
The rate of head cut advance over time was also variable (Table 6.1). Between Kibler Park and 
Kromvlei the wetland has experienced the greatest head cut advance of all the sections, most 
significantly in the periods 1952 to 1961 (56 m/year) and 2006 to 2008 (59 m/year). The wetlands 
at the Zwartkoppies Pump Station exhibit significant head cut advance, after 1961 (52 m/year).  
 
Table 6.1: Rates of head cut advance for Sections 2 to 5 (See Figure 6.3 and 6.4 for locality). 
1938 to 1952 1952 to 1961 1961 to 1984 1984 to 2003 2003 to 2006 2006 to 2008
 2
 3
 4 56 14 59
 5 14 20 52
 5 9 14 40 41
 Section Rate of Head Cut Advance (m/year)
 9 18
 22
  
 
The wetland described by Humphrey (1910) that extended to within 1.6 km of the Klip River 
Railway Station (south of Zwartkoppies Pump Station) retreated 5 km upstream by 1938. This 
equates to a retreat of 180 m/year.  
7. Recommendations 
Results of this investigation confirm that the current state of the Klip River Wetland is dire. Large 
tracts of wetland have disappeared over the last 100 years. The wetlands that remain are not 
functioning optimally. The importance of the wetland to purify sewage, industrial and mining 
effluent is paramount. The window of opportunity to save or restore at least parts of the wetland, 
and therefore maintain the purification potential of the wetland, is closing. Peat burns will lead to 
erosion of the soils and therefore the reworking of the pollutants into the river system. It is 
anticipated that if the Klip River Wetland collapses completely, the water quality will worsen along 
the Vaal River and in the Vaal Barrage, which will impact negatively on downstream users.  
 
 95 
The proposed solution to protect the wetland from further erosion and head-cut advance is to 
force the water out of the channels and into the wetland. By building a number of weirs in the 
channels, it is possible to dam the water upstream of the weir, thereby forcing the water to flood 
the wetland again. The choice of weir location and design is key to the success of this kind of 
intervention. It is essential that the weirs are constructed on resistant strata, as water will follow a 
path of least resistance and will undermine the weir built on any rock which is easily eroded, for 
example dolomite. The weirs should ideally be located on resistant dykes which cut across the 
wetland and penetrate through the entire dolomite sequence. From the long river profile analysis it 
is determined that the gradient attained by the current functioning and unchannelised wetland 
(Lenasia area, Section 1 in Figure 5.1.8) is between 1 : 0.001 and 1 : 0.002. It is recommended 
that the installation of the weirs aims at attaining an equivalent gradient. Four weirs are proposed 
at dykes projected from regional aeromagnetic data (Figure 7.1). It is recommended that a 
detailed magnetic survey is carried out to confirm the location of these dykes before weir 
construction is considered.  
 
In order to better understand the correlation of the deterioration of the wetland and changes in the 
water discharge from the wetland, discharge data from the sewage works should be acquired and 
analysed. The use of satellite multi-spectral imagery to quantify and investigate the Klip River 
Wetland and the true extent of its moisture content should also be investigated further. This 
analysis might be useful in better determining the rate of change, since the Landsat program is 
the longest running enterprise for acquisition of imagery of Earth from space, with the first Landsat 
satellite having been launched in 1972. 
 
 96 
 
Figure 7.1: Proposed weir locations according to dyke positions. 
 97 
Reference List 
 
Arnold, R.W., 1980. Modelling water quality in the upper Klip River. Dissertation, School of 
Civil Engineering, University of the Witwatersrand, Johannesburg, 137pp. 
 
Arnold, V., 2004. Geomorphology of the Klip River. B.Sc. (Hons) Dissertation, School of 
Geosciences, University of the Witwatersrand, Johannesburg, 66pp. 
 
Begg, G., 1990. Policy proposals for the wetlands of Natal and KwaZulu. Natal Town and 
Regional Planning Report. Volume 75, 86pp.  
 
Blowes, D.W., Ptacek, C.J., Jambor, J.L. and Weisner, C.G., 2003. The geochemistry of 
Acid Mine Drainage. Treatise on Geochemistry, p 149 - 204.  
 
Clay, A.N., 1981. The geology of the Malmani Dolomite Subgroup in the Carletonville Area, 
Transvaal. Doctorate of Philosophy, Faculty of Science, University of the Witwatersrand, 
Johannesburg, 169pp. 
 
Davidson, C., 2003. Catchment diagnostic framework for the Klip River Catchment, Vaal 
Barrage, October 1998 ? September 1999. Master of Science: University of the 
Witwatersrand, Johannesburg, 154pp.  
 
Deutch, W.J., 1997. Groundwater Geochemistry: fundamentals and applications to 
contamination. Lewis Publishers, 221pp.  
 
DWAF, Project Steering Committee Meeting 3., 2007. Integrated water quality management 
plan for the Vaal River system. Presentation. 
 
Goldmann, C.S., 1895-6. South African Mines; Their position, results and developments: 
Together with an account of diamond, land and finance and kindred concepts. Effingham 
Wilson & Co. London. 3 Volumes.  
 
 98 
Grundling, P.L. and Marneweck, G., 1999. Mapping, characterisation and monitoring of the 
Highveld Peatlands. Wetland Consulting Services Report No. 28/99, 7pp.  
 
H?jeka, M., Hors?kb, M., H?jkov?a, P. and D?t?c, D., 2006. Habitat diversity of central 
European fens in relation to environmental gradients and an effort to standardise fen 
terminology in ecological studies. 23pp. 
 
Hammer, D.A., 1997. Creating freshwater wetlands. Lewis Publishers: New York, 406pp.   
 
Hancock, F.D., 1966. A first study on the algal ecology of the Klip River in the gold mining 
area of Johannesburg. Doctorate of Philosophy. Faculty of Science: University of the 
Witwatersrand, Johannesburg, 101pp. 
 
Handley, J.R.F., 2004. Historic overview of the Witwatersrand Goldfields. Handley, Howick, 
240pp. 
 
Haywood, L. K., 2004. Tadpoles as biological indicators of aquatic health: assessing the 
mining industries? impact on water quality. Unpubl. M.Sc. Thesis, Faculty of Science: 
University of the Witwatersrand, Johannesburg, 163pp. 
 
Humphrey, W.A., 1910. The geology of the valley of the Klip River. Report of the Geological 
Survey (Transvaal) for the year 1909. Union of South Africa Mines Department, p 91 - 97.  
 
Johannesburg Water, 2004. Johannesburg Water Operations. 4pp. 
 
Kafri, U. and Foster, M.J.B., 1989. Hydrogeology of the Malmani Dolomite in the Klip River 
and Natalspruit Basins, South Africa. Environmental, Geology and Water Science 13(2), p 
153 ? 166. 
 
McCarthy, T.S., Arnold, V., Venter, J.S. and Ellery, W.N., 2007. The Collapse of 
Johannesburg?s Klip River Wetland. South African Journal of Science 103, 
October/November 2007, p 1 ? 7.  
 
 99 
McCarthy, T.S. and Venter, J.S., 2006. Increasing pollution levels on the Witwatersrand 
recorded in the peat deposits of the Klip River Wetland. South African Journal of Science 
102, January/February 2006, p 27 ? 34.  
 
Mellor, E.T.,1917. The geological map of the Witwatersrand Goldfield. Union of South Africa 
Department of Mines and Industries, Geological Survey. (1 : 60 000) Maps 1 to 3. 
 
Mphephu, N.F., 2004. Geotechnical, environmental evaluation of mining impacts on the 
Central Rand. Thesis to Faculty of Science for Doctor of Philosophy: University of the 
Witwatersrand, Johannesburg, 222pp. 
 
Mucina, L., and Rutherford, M.C., 2006. The vegetation of South Africa, Lesotho and 
Swaziland. Publishers: South African National Biodiversity Institute, 807pp. 
 
Naicker, K., Cukrowska, E. and McCarthy, T.S., 2003. Acid Mine Drainage arising from gold 
mining activity in Johannesburg, South Africa and environs. Environmental Pollution 122 
(2003), p 29-40. 
 
NEMP, 2003. DWAF: National Eutrophication Monitoring Programme Annual Assessment 
of Eutrophication in South Africa (2003), 4pp. 
 
North West Environmental Management Plan for the Hartebeespoort Dam Remediation., 
2005. North West Provincial Government. Environmental Management Plan Series 5, 15pp.  
 
Partridge, T.C., 1968. Some geomorphic units in the Transvaal and their significance in 
physical development. Thesis to Faculty of Science for Doctor of Philosophy: Department of 
Geography: University of Natal. Volume1.  
 
Rall, V.E. and Rall, J.L.,1996. Ecological health assessment of the Elsburgspruit / 
Natalspruit Catchment.  ERPM Internal Report. 
 
Rogers, K.H. and Herrera, G., 1986. Sulphate Reduction in the Rietspruit Vleis. Department 
of Botany: University of the Witwatersrand, Johannesburg. 
 100 
 
Schweitzer, J.K., Kramers, C.P., Stewart, R.S., Crafford, J., Arnold, V. and Khumalo, B., 
2004. A strategic water management plan for the prevention of water ingress into the 
underground workings of the Witwatersrand mining basin. Shango Solutions contributions to 
Ferret Mining Consultancy Report to Council for Geoscience: Summary (15pp); Task 2 
(72pp; accompanying Appendices 121pp), Task 3 (89pp; accompanying Appendices 
147pp). 
 
Turton, A., Schultz, C., Buckle, H., Kgomongoe, M., Malungani, T. and Drackner, M., 2006. 
Gold, Scorched Earth and Water: The hydropolotics of Johannesburg. Water Resources 
Development, Vol. 22, No. 2, p 313 ? 335.  
 
Warner B.G., 1996. Vertical gradients in peatlands. In Wetlands, Environmental Gradients, 
Boundaries, and Buffers, ed. Mulamoottil, G., Warner, B.G., and McBean, E.A. Lewis 
Publishers: New York, p 45 - 65. 
 
Watson, J.T., Sherwood, S. C., Kadlec, R.H., Knight, R.L., Whitehouse, A.E., 1989. 
Performance expectations and loading rates for constructed wetlands. In: Hammer, D.A. 
(Ed.), Constructed Wetlands for Wastewater Treatment. Lewis, Chelsea, p 319?351. 
 
White, E.G., 1957. Chemical aspects of stream pollution on the Witwatersrand. Doctorate of 
Philosophy. Faculty of Science: University of the Witwatersrand, Johannesburg, 148pp. 
 
Whyte, C.R. and Shepherd, J.K., 1990. Mkomazi wetland inventory. Town and Regional 
Planning Supplementary Report. Volume 46, 44pp. 
 
 
 101 
Acknowledgements 
 
The author wishes to thank the following persons:  
City of Johannesburg Corporate GIS for the 2006 Aerial photography and LIDAR data. 
Michael Silberbauer and Marica Erasmus from Department of Water Affairs and Forestry 
(DWAF) for providing water quality data. 
 
Paula Ogilvie and Matt Mowbry for assistance with Digital Terrain Modelling. 
 
Dr. Shawn Letts for his assistance with the filtering of the long river profiles. 
 
Shango Solutions, especially Jochen Schweitzer, for the financial support and technical 
assistance.