SCOPE 42 - Biogeochemistry of Major World Rivers

6

Biogeochemistry of Major African Rivers: Carbon and Mineral Transport

OLASUMBO MARTINS
Federal University of Agriculture, Abeokuta, Nigeria
and
JEAN-LUC PROBST
Centre de Géochimie de la Surface, CNRS, Strasbourg, France
 
6.1 INTRODUCTION
6.2 PHYSICAL CHARACTERISTICS
6.2.1 RELIEF
6.2.2 GEOLOGY
6.2.3 CLIMATIC SETTING
6.2.4 VEGETATION
6.2.4.1 Tropical rain forest
6.2.4.2 Tropical savannah
6.2.4.3 Steppe and desert
6.2.5 SOILS
6.3 HYDROLOGY
6.4 CASE STUDIES
6.4.1 THE NIGER RIVER
6.4.1.1 Dissolved solutes
6.4.1.2 Suspended sediments
6.4.1.3 Carbon dioxide content
6.4.1.4 Dissolved organic carbon (DOC)
6.4.1.5 Particulate organic carbon (POC)
6.4.1.6 Total organic carbon (TOC)
6.4.1.7 Transport rate of organic carbon
6.4.2 THE ORANGE RIVER
6.4.2.1 River discharge
6.4.2.2 Hydrochemistry
6.5 COMPARATIVE ANALYSIS OF MAJOR AFRICAN RIVERS
6.5.1 DISSOLVED MINERAL SPECIES
6.5.2 TOTAL SUSPENDED SOLIDS (TSS)
6.5.3 ORGANIC CARBON
6.6 CONCLUSION
REFERENCES

6.1 INTRODUCTION

The SCOPE/UNEP River Project has improved our knowledge about the transport of particulate and dissolved matter by African rivers, and more particularly about their carbon transport. In order to model river transport, rivers were classified according to 'type' as delineated by climate, geology, relief, biota and degree of development, rather than by geographic zones. Accordingly, the African rivers belong to two types: (1) tropical forest as in the case of the Zaïre; and (2) tropical savannah as in the case of the Niger, Nile and Orange.

This chapter provides a synthesis of all the investigations on transport in African rivers supported by the SCOPE/UNEP Project. In addition, results of other investigations are included in this synthesis, as appropriate. After a general review of the characteristics of the African continent (relief, geology, soils, vegetation, climate and hydrology), two case studies (Niger and Orange) are presented in detail. A comparative analysis of carbon and mineral transport by the major African rivers provides the basis for the modelling of transport of major world rivers.

6.2 PHYSICAL CHARACTERISTICS

The African continent has a total area of about 30 x 106 km2. Except for the environments surrounding the Sahara Desert, the surface is fairly well dissected by rivers. The drainage systems may be subdivided into peripheral and inland systems. Prominent among the latter are the Chad basin system and part of the East African lake system. The peripheral rivers drain into the Atlantic Ocean (Zaïre, Niger, Orange, Senegal, Gambia), the Indian Ocean (Zambezi, Limpopo), and the Mediterranean Sea (Nile).

6.2.1 RELIEF

The lowlands and highlands within the African continent can be conveniently divided by a line stretching from east of the lower Nile valley southwards across the upper Zaïre basin, terminating south of the Zaïre basin (Figure 6.1). The area northwest of this line constitutes lowland plains whose altitudes range from 150 to 600 m. Its major drainage basins are those of the Zaïre, the middle and lower Nile, the Sénégal, the Niger and the Chad inland drainage system. Higher altitudes are limited to the Atlas mountains, the Ahaggar and the Tibesti massifs, the Highlands of the Cameroonian Range and the Guinean Highlands. Most of the land southeast of the relief line is higher than 1000 m. Exceptions are found in the coastal strip extending southwestwards from Somalia, adjacent to the Indian Ocean. The most prominent feature of the African relief is the rift valley system situated in East Africa. It has two components:

  1. The eastern segment which cuts through Tanzania and Kenya and extends northeastwards along the trenches of the Afar depression, the Red Sea and the Gulf of Aden.
  2. The western rift segment which starts from the upper Nile, runs through Lake Kivu and Malawi southwards and extends to the middle Zambezi and to the Okavango swamps in Botswana.

Figure 6.1 Physical features of the African continent. The bold line demarcates low relief «600m) from high relief (>1000m) areas (after Grove 1978)

6.2.2 GEOLOGY

The African Shield is underlain by Precambrian rocks, covered in places by relatively thin layers of sedimentary rocks (Figure 6.2). A distinction is generally made between the 'older cratons' and the 'younger orogens'. They demarcate drainage basins, forming escarpments of the Air and Ahaggar, the Guinean Highlands and the Cameroon Highlands. They are made up of granites, gneiss' and schists. East of the continent, they are more extensive particularly between Transvaal in the south and the Red Sea hills of Egypt in the north.

Contrary to the latest Precambrian rocks, which have not been intensively folded in most areas, the oldest crystalline rocks have been strongly folded over wide areas. During the lower Paleozoic, Gondwanaland formed a large land mass surrounded by an extensive geosyncline. The rocks were subjected to tectonic movements at about the same time as the Caledonian orogeny, although folding in Africa was less intensive. At about 230 x 106 years ago, the southern part of Gondwanaland was glaciated, as witnessed by tillites and other glacial features in southern Africa. While marine clays and coalbearing continental deposits accumulated in southern and central Africa between the Upper Carboniferous and Lower Jurassic (Karoo Formation), northern Africa was buried by clastic sediments, jointly called the 'Continental Intercalaire', which includes the Nubian sandstone and various porous layers.

Rifting of Gondwanaland started in the Mesozoic, accompanied by the formation of the South Atlantic. This was followed by the appearance of large fractures in East and in West Africa, a fault-trough opened from the Gulf of Guinea extending northeastward into the Sahara region (Grove 1978). Transgression, mainly from the Tethys, flooded extensive areas in the north and west of the continent, depositing thick layers of sandstone and limestone. Further movement of the African continent to the north during the Tertiary reduced the size of the Tethys, deforming the geosyncline, and uplifted the Atlas ranges.

Figure 6.2 Simplified geological map of Africa. The unshaded areas are mainly underlain by sandstones, shales and limestones (after Grove 1978)

6.2.3 CLIMATIC SETTING

Major climatic zones identified of Africa are elucidated in Table 6.1.


Table 6.1 African major climatic zones (after Rodier 1964)


Zone Annual precipitation (mm)

Desert 100 -150
Sub-desert  150 -300
Sahelian 300 - 750
Tropical  750 -1200
Tropical-equatorial  >1200

Severe evapotranspiration leading to loss of water from lakes and rivers is evident in places where high temperatures are combined with low relative humidity. Rate of water loss varies between 750 mm in the more humid areas to more than 2000 mm in and around the Sahara. Over most parts of the continent the potential evapotranspiration is larger than the precipitation rate, at least during a few months of the year.

The Niger, rising from the more humid (precipitation, 1500 mm/year) Fouta Jallon Highlands, flows due northeast into the sub-desert zone, then south-easterly through the Sahelian zone into the tropical area, where it forms its delta. The Orange River drains solely desert and sub-desert climatic zones, where annual precipitation is mostly less than 250 mm. The Nile originates from the wet equatorial zone and flows northwards into sub-desert and desert environments. The Zaïre River drains exclusively areas within the tropical-equatorial zone (1200 mm/year), thereby representing the only truly tropical river in the continent.

6.2.4 VEGETATION

Three major vegetation types can be discerned, namely: the tropical forest, the tropical savannah, and the steppe and desert (Figure 6.3).

6.2.4.1 Tropical rain forest

The tropical rain forest is limited to lowland areas with rainfall throughout the year. The annual total rainfall usually exceeds 1400 mm/year, and temperature ranges between 21 °C and 32 °C. The Niger Delta and wide areas of the Zaïre drainage basin are tropical rain forest regions.

6.2.4.2 Tropical savannah

The savannah belt occupies sub-humid grassland regions extending through the south-central and East African Plateau, to the northern margins of both the Niger and Zaïre, as far as to the south of the Gambia River. Distribution of savannah coincides with peneplains where waterlogging and moisture deficiency alternate.

Figure 6.3 Distribution of major vegetation zones in relation to annual rainfall (after Grove 1978)

6.2.4.3 Steppe and desert

Steppe and desert are restricted to areas where annual precipitation is less than 400 mm/year. Steppe vegetation is located in parts of semi-arid West Africa, in the central Sudan Republic and in Somalia. Also, in the southern part of Africa, gradual merging of the savannah southwards into wooded steppe and thorn scrub has been described.

6.2.5 SOILS

The quality and type of soils depend on the climate and the parent material from which the soils are derived. Thus, African soils are very variable, and for each climatic type there are both highly productive and very poor soils. Also, due to the long-term pattern of land use, a given soil type expected for a particular climatic belt may have been drastically altered, making the overall distribution of soils very complicated.

The vegetation type often dictates the soil potential. Desert soils are generally shallow and lacking in organic material. The driest areas are characterized by sands and pebbles and the upland soils are usually thin and stony. Areas with 250-600 mm/year of rain (semi-arid) are typified by ferruginous soils which are sometimes alkaline, but may also be slightly acidic, if heavily leached.

Dark-brown or black vertisols are distributed on alluvial plains where they are flooded by occasional rains or by large rivers. Such soils are heavy and calcareous, consisting mainly of montmorillonite. They are common in the basin of the White Nile and in some inland drainage areas, such as the southeastern Chad basin. Parts of West Africa, including the Zaïre basin, are covered by red and ferruginous lateritic ironstones; these areas receive between 150 and 1250 mm annual rainfall and the vegetation is of the savannah woodland type. Intensely weathered soils are characteristic of the more humid regions, where high and dense forest vegetation protects the regolith from surface erosion. Such soils, called latosols or ferralitic soils, are red or yellow in color and they may extend for several meters below the surface, although the topsoil which is rich in organic matter may not be more than a few centimeters thick. Replenishment of organic matter is very rapid through decomposition of leaves, branches and tree trunks by micro-organisms.

The best soils are found at higher altitudes of the Kenyan and Ethiopian highlands. These are deep loamy soils derived from basalts and other volcanics of the Tertiary period. They are generally well drained but easily liable to erosion by gullies.

6.3 HYDROLOGY

Most African rivers have highly seasonal flows. However, construction of dams and reservoirs has obliterated most seasonal flow patterns. The Nile is the most severely impacted aspect (Kempe 1983, 1989). The Aswan Dam was constructed between 1960 and 1970. The annual hydrograph revealed between 1960 and 1964 prior to the dam closure a distinct flood peak lasting from July to October (Figure 6.4). In 1968, the flood peak was considerably reduced and the flood period shortened to June and July when water from the reservoir is drained to provide enough water for summer irrigation.

Figure 6.4 Average discharge at Asyut before and after the closure of the Aswan Dam (after Kempe 1983)

Other rivers affected by similar circumstances, although to a lesser extent, are the Niger (with dams in Kainji and Jebba) and the Orange (Le Roux and Verwoerd Dams). Figure 6.5 depicts the Orange River system with locations of major impoundments as well as the year of their closure. It has been ascertained (Hart 1985) that the construction of these impoundments has considerably dampened the seasonal flow patterns. Average monthly discharges, which were highest in late summer (March) before 1959-60, now have a bimodal late summer peak while the annual discharges have been lowered slightly (Figure 6.6). Minimum flows occurred in winter months which are characteristically dry over the catchment.

Figure 6.5 The Orange River system, showing the position of major impoundments (with year of closure) and hydrological stations (with record span) (after Hart 1985)

Above its reservoirs, the Zambezi River shows a distinct seasonal hydrograph, as evidenced from the 59-year record (Figure 6.7). Peak discharges occur between March and June while low flows are generally recorded between August and January .The variability in water discharge is relatively low during the dry period, but increases rapidly at rising water (Borchert and Kempe 1985). The discharge ratio is about 10, which is similar to values for some large semi-arid-savannah rivers, like the Niger and the Orange Rivers (see below).

Table 6.2 summarizes discharge characteristics of some major rivers. The average annual discharge of the Zaïre is estimated at 1300 km3/year for an 80- year period (Probst and Tardy 1987). This corresponds to a specific discharge of 12.81/km2/s. Of particular interest is the discharge ratio between high flows and low flows, or seasonality index (SI), which is only 1.7; this is very low when compared to that of the Niger and Orange Rivers (13 and 8 respectively). Low SI values may be due to regularity of rainfall ( as in the case of the Zaïre); it may also be a direct effect of damming, for example in the case of the Nile (SI = 2.6).

Figure 6.6 Hydrological data for selected Orange River stations (after Hart 1985): (a) interannual fluctuations; (b) monthly discharges since the record began unti11960; (c) monthly discharges during the period 1960-70; (d) monthly discharges since closure of PK Le Roux Dam

In Figure 6.8, the succession of dry and humid hydroclimatic periods is evident in the historical discharge record. As described by Faure and Gac (1981), the discharges of the Sahelian rivers (Senegal, Chari and Niger)' experience synchronous fluctuations during the last 80 years. Fluctuation amplitudes decrease in direction to the Equator, i.e. from the Niger, Senegal and Chari to the Zaïre which presents the smallest interannual fluctuations (Probst and Tardy 1987).

If one compares the discharge fluctuations of African rivers flowing in the northern hemisphere (Senegal, Niger, Chari and Nile) with those flowing in the southern hemisphere (Orange, Limpopo and Zambezi), one can detect some lags and even oppositions between the main hydroclimatic periods. The humid periods of 1930 and 1960, which have affected the rivers of the northern hemisphere, are opposed to a dry period in rivers of the southern hemisphere. In contrast, the dry event of the year 1970 in the northern hemisphere corresponds to a humid interval in the southern hemisphere. Nevertheless, it is remarkable that the drought period of 1940-45 has affected more or less all African rivers, both in the northern and southern hemispheres.

Figure 6.7 The Zambezi monthly discharge record at Victoria Falls, 59-year average (after Borchert and Kempe 1985)

6.4 CASE STUDIES

6.4.1 THE NIGER RIVER

The Niger River was studied during the 1980-81 hydrological year within the framework of the SCOPE/UNEP Program. The main aims were to establish the seasonal fluxes of inorganic and organic components, as well as to derive the total annual fluxes of these constituents through regularly collected samples.

Table 6.2 Discharge characteristics of some major rivers


River (station)  Drainage  area
 (106 km2)
Interannual discharge (m3/s)
Specific discharge (l/s/km2) References 
Max. Min.  Mean

Zaïre 3.5 57200 32750 41134 11.8 Nkounkou and  
(Brazzaville)    Probst (1987)
Nile 3.0 1960 752 1220 0.4 Kempe (1983)
(Assgut)
Niger 1.2 17797 1335 4886 4.1 Martins (1982) 
(Lokoja)
Orange 1.0 500 63 360 0.3 Hart (1985)
(Vioolsdrip)
Senegal 0.27 2000 0.19 330 1.2 Gac and Kane
(Bakel)     (1986a,b)
Zambezi 0.54 5354 1273 2429  4.5 Probst and
(Matundo-Cals)     Tardy (1987)

Figure 6.8 Interannual fluctuations of the mean annual discharges (m3/s) for major African rivers¾filtered data (five year arithmetic mean) (after Probst and Tardy 1987)

6.4.1.1 Dissolved solutes

Table 6.3 depicts the geochemical variables analysed on the Niger River at Lokoja. Calcium and magnesium are the dominant ions, the water therefore corresponds to the rock-dominated type. The evolution of seasonality of major ions (Figure 6.9) reveals initial dilution at the beginning of the rainy season (May to June) and a subsequent salinity increase as dissolved salts leached from the floodplain and from adjacent river systems are drained into the main stream. This phenomenon is a combined effect of the discharge hydrograph and the nature of the river valley. The ratio between the lowest and highest water stages at Lokoja is 1/13, and the river valley is very broad. A large portion of the floodplain lies above the mean water stage for more than five months of the year; even the main river course appears strongly braided, the river having dropped its solid load abruptly at the start of the dry season. Other minor rivers often dry up completely, or have their water volumes substantially reduced. Intense evapotranspiration concentrates salts in the upper few centimeters of the soil layer and isolated water bodies may even form miniature salt flats. The influence of such concentrated solutions is evident when floodplain water drains into the main channel. The influx of concentrated salts from the floodplains as the river stage rises produces a leaching or flushing effect.

Figure 6.9 Seasonal variations of major ions and total suspended sediments (TSS) in relation to river discharge (Q) of the Niger at Lokoja (after Martins 1983)

With decreasing water volume and increasing rate of evapotranspiration, the concentration of dissolved salts rises throughout the dry season. At this point, it is not possible to distinguish between concentration increase due to evapotranspiration and that due to infiltrating groundwater. The latter process seems to be predominant towards the end of the dry season.

6.4.1.2 Suspended sediments

Similarly, the concentrations of suspended sediments are related to the discharge characteristics of the river at the sampling station. As depicted in Figure 6.9, it is evident that the temporal variation in discharge does not coincide with that of the total suspended solids (TSS), the peak concentration of which was attained well before the maximum discharge. Concentration is highest in July (134 mg/l), dropping to about 80 mg/l during peak discharge. According to Kat tan et al. (1987), this lag between maximum TSS and peak discharge may be attributed to the remobilization of sediments deposited in the riverbed and floodplain. It therefore seems that large amounts of fines are transported during the first rains, as flow increases initially. Subsequent floods would detach sediments of larger diameter, which are generally carried in the lower water column. The secondary TSS peak is probably an effect of atmospheric dust input between December and January, as is generally witnessed throughout West Africa.

6.4.1.3 Carbon dioxide content

It is now generally observed that irrespective of lithology, climate or latitude, rivers have a higher partial pressure of CO2 than ambient atmosphere (Kempe 1982). An explanation is offered by in situ oxidation of organic matter within the riverine environment. However, results obtained from the Niger seem to indicate that terrestrial (soil) CO2 is prevalent during at least part of the seasonal cycle.

The average pressure of CO2 in the river is about seven times higher than the atmospheric value, i.e. 2400 ppm as against 330 ppm. Highest pCO2 (6360 ppm) was observed in June at the start of the wet season (Figure 6.10). This also coincides with the period of strong erosion. Data on plankton primary production and respiration would be needed to evaluate the extent of metabolic input of CO2 in the river.

Figure 6.10 Fluctuations of pCO2 and DOC in the Niger River water during the water year 1980/81 (after Martins 1983)

6.4.1.4 Dissolved organic carbon (DOC)

Dissolved organic carbon levels range between 2.1 mg/1 and 6.6 mg/l, the average value being 3.5 mg/l. This is well below the median value (5 mg/l) commonly encountered in most world rivers (Meybeck 1982). So far, tropical rivers are known to have values between 2 and 15 mg/l.

The distribution of DOC with time is shown in Figure 6.10. The maximum concentration was attained at rising water in June and July. Thereafter, the concentration decreases gradually, reaching its lowest level between March and April. The pattern of the seasonal variation of DOC seems to indicate that its source is partly terrestrial. Dissolved organics are leached from the soil profile during times of heavy surface erosion. After the floods, input of soil organics strongly declines and DOC level is probably maintained by phytoplankton. Autochthonous organic carbon input possibly occurs in form of extracellular exudation, especially during algal photosynthesis.

6.4.1.5 Particulate organic carbon (POC)

Particulate organic carbon constitutes between 2% and 9% of the total suspended sediments. This is equivalent to concentrations from 1 mg/1 to 4-6 mg/l. Season fluctuations, which are similar to that of the suspended sediments, depict maximum values at rising and high water periods. POC contents rose from about 2.8 mg/l to 4.6 mg/l at rising water stage, while concentrations remain between 1 mg/1 and 2.3 mg/l (average 1.8 mg/l) at low flow.

If the source of POC in the river were autochthonous, then low concentrations of particulate organic matter may have resulted from in situ decomposition since this fraction is mostly labile. Dilution of organic matter by clay minerals and other inorganic colloids is, however, thought to explain the low proportion of allochthonous POC. The ratio of organic to in organic materials would therefore be similar to that of soil; soils from semi-arid environment are known for their low organic carbon content.

When the possibility of both autochthonous and allochthonous sources of POC is considered within the hydrological year, in situ organic carbon production should be dominant in the dry season, since tributary river inputs are negligible throughout this period. High abundance of phytoplankton has been reported in parts of the basin (Adeniji 1973) at low water level and at a time when transparency of water is high. At high water turbidity (low transparency) a significant drop in phytoplankton quantity is recorded.

6.4.1.6 Total organic carbon (TOC)

TOC contents range from 3.5 mg/l to 10.8 mg/l (average 5.8 mg/l) at Lokoja station. Highest values were recorded between June and September, when the erosive force of the river is greatest; between January and May concentrations did not rise above 5.0 mg/l. As shown in Table 6.3. DOC is generally greater than POC. This is further revealed by the DOC/TOC ratio which is higher than 0.5 in the river. Due to the enhanced POC content in the tributary rivers, DOC/TOC values fall below 0.5 particularly in June and September when their inputs are prominent. Thus, the smaller tributary rivers seem to behave differently compared to the main stream, with respect to the ratio of dissolved to particulate organic carbon.

Table 6.3 Constituents of Niger River discharge


Dates (Days of the year)

 River disch.
(m3/s)

 Ca2+

 Ma2+

 Na+

 K+  

HCO3-

SO42-  

Cl-  

SiO2 (mg/l)

 TDS

 TSS

 POC

 PON

 DOC

 TOC

 DOC
/TOC 


136   1980

 1477

 6.2

 1.3

 1.6

1.2

 36.0

 0.62

 0.6

 20.4

 68.0

 11.5

 1.0

 0.06

 2.70

 3.70

 0.74

162

 2269

 6.2

 1.9

 2.4

 0.8

 33.5

 0.93

 0.7

 15.0

 61.4

 65.0

 2.8

 0.18

 6.60

 9.40

 0.69

204

 4050

 5.4

 1.8

 1.8

 1.2

 30.5

 0.77

 0.5

 14.2

 56.2

 135.0

 4.6

 0.30

 6.15

 10.8

 0.57 

225

 10200

 4.8

 1.2 

1.6

 1.5

 30.5

 0.48

 0.7

 15.0

 55.8

 83.6

 2.9

 0.20

 3.60

 6.50

 0.55 

245

 17797

 5.4

 1.6

 1.2

 1.2

 33.0

 0.09

 1.0

 12.5

 57.2

 80.4

 3.2

 0.25

 3.50

 6.70

 0.53 

265

 12908

 5.6

 2.3

 2.0

 0.8

 33.6

 0.19

 0.9

 15.0

 60.3

 72.0

 2.0

 0.15

 3.75

 5.80

 0.63 

305

 4955

 6.2

 2.4

 2.5

 0.8

 37.2

 tr

 1.1

 15.0

 65.2

 85.0

 2.3

 0.20

 3.50

 5.80

 0.58

341

 1958

 6.8

 2.6

 3.0

 1.2

 41.5

 0.96

 0.7

 15.0

 71.8

 93.2

 2.0

 0.15

 3.00

 5.00

 0.60

13      1981

 1730

 7.0

 2.8

 2.7

 1.2

 42.7

 0.62

 0.7

 8.8

 66.5

 80.4

 1.7

 0.18

 2.50

 4.20

 0.58

42

 1868

 8.0

 2.6

 3.0

 0.8

 42.7

 2.10

 0.9

 11.0

 71.1

 53.3

 1.4

 0.20

 2.10

 3.50

 0.60

78

 1635

 8.6

 2.4

 3.0

 0.4

 43.9

 1.40

 0.7

 8.0

 68.4

 36.6

 2.0

 0.20

 2.15

 4.20

 0.50

106

 1335

 8.4

 2.4

 2.0

 1.2

 41.5

 1.90

 0.7

 25.0

 83.1

 28.0

 2.0

 0.20

 2.95

 5.00

 0.59

134

 1335

 5.4

 2.4

 2.3

 0.8

 33.5

 1.10

 0.6

 15.0

 61.1

 40.0

 n.d.

 n.d.

 n.d.

 n.d.

 n.d.


Figure 6.11 Plots of DOC/TOC (a) and TOC (b) versus total suspended solids for the Niger (after Martins 1983)

The relationship between TOC and TSS (Figure 6.11) is positive and highly significant. This proves the importance of particulate organics, whose increased concentration with increasing suspended solid concentration further accentuates this development.

6.4.1.7 Transport rate of organic carbon

The Niger exports an estimated TOC load of 1.2 x 106 t/year of which 0.66 x 106 t are attributed to POC and 0.53 x 106 t to DOC. The specific rate of annual transport is 0.44 t/km2 for DOC and 0.5 t/km2 in the case of POC. Thus, the specific transport rate of TOC in the Niger drainage basin is about 1 t/km2/year.

6.4.2 THE ORANGE RIVER

The river drains 1.02 x 106 km2 of semi-arid to arid areas; its average discharge of 360 m3/s constitutes over 20% of the mean annual runoff of South Africa. As many as four man-made lakes (Figure 6.5) have been constructed within the river drainage basin, mainly in the upper catchment area; the total storage capacity of the lakes is 900 x 106 m3.

6.4.2.1 River discharge

Figures 6.6(a)-(d) represent the hydrological data for selected stations of the basin. Like other semi-arid rivers, considerable variation exists in the interannual discharge configuration. However, the construction of dams has reduced the amplitude of the seasonal variation curves (see Section 6.2 above).

6.4.2.2 Hydrochemistry

Interannual variation of physico-chemical parameters is portrayed in Figure 6.12; seasonal fluctuations are severely depressed in most variables, this has been attributed to the stabilizing influence of 2.9 km3 of water in Lake Le Roux stretching back up the now drowned Orange River, and covering an area of 128 km2 (Hart 1985) .According to the author, the reservoir was operated at between 80% and 90% of its full supply level during the first eight months of the study. Discharge waters were therefore well oxygenated, and altitude corrected saturation values mostly varied between 70% and 80%. Conductivity, pH and alkalinity showed conservative variation of around 15-17 mS/m, 8.0 and 1.3 meq/l respectively.

There is a marked decrease in total suspended sediment concentration from about 145 mg/l in August 1981 to about 40 mg/l in January 1984. Corresponding decline in particulate organic carbon (POC) concentration from around 1.3 mg/l to 0.6 mg/l was also recorded. DOC values varied only slightly between 2 mg/l and 3 mg/l during the same period.

Figure 6.12 Seasonal variations in selected hydrological, physical and chemical attributes of PK Le Roux Dam discharges during three years (after Hart 1985)

6.5 COMPARATIVE ANALYSIS OF MAJOR AFRICAN RIVERS

6.5.1 DISSOLVED MINERAL SPECIES

Solute contents of rivers are naturally derived from three sources: atmospheric (cations and CO2), lithologic (rock weathering) and groundwater infiltration (in artesian situations or dry period seepage) .

Where wind direction is favorable, the magnitude of precipitation input of dissolved substances will be a factor of the geometry of the drainage basin. Thus, atmospheric input of sea salts in a drainage basin like that of the Niger, which is narrow towards the outlet, may not be as much as that of the Zaïre which broadens out near the coast.

Chemical weathering of rocks, including terrestrial dust, still remains the main source of dissolved substances, its rate being under the control of several factors among which topography, vegetation cover and humidity are primary .

The rate of groundwater infiltration/seepage is difficult to assess. It may play a vital role in highly seasonal rivers during low flow periods. This is the case in some semi-arid and arid rivers and streams, where dry season flows are maintained only by springs and groundwater seepage.

Table 6.4 gives the average solute concentrations of major African rivers. The Nile has the highest dissolved salt content followed by other arid rivers: the Orange and the Zambezi. Rivers from semi-arid environments, such as the Niger and the Senegal, have moderate solute levels, while the Zaïre River water is by far the most dilute.

It should be noted that most African rivers cannot be said to be in the natural state any longer, since the chemical composition of their waters has been altered to various degrees. The total dissolved solids (TDS) of the Niger have been elevated against those of the pre-Kainji reservoir era by about 10% (Martins 1983); this value may have increased slightly since the inception of the Jebba Dam. Kempe (1983) estimated an increase of about 33% in TDS for the Nile at Cairo as a result of the Aswan Dam, irrigation and industrial activities. Although similar data do not exist for other rivers, different authors have reported the presence of dams in other drainage basins: the Orange (Hart 1985); the Zambezi (Borchert and Kempe 1985); and the Senegal (Gac and Kane 1986a,b).

Dissolved inorganic carbon, present mostly as the bicarbonate ion, constitutes more than 50% of the TDS in the Niger, Senegal and Zambezi Rivers. It should be noted that an important fraction of the bicarbonate ions measured in river water, is supplied by the atmospheric CO2 which is consumed by silicate and carbonate rock weathering. Nkounkou and Probst (1987) have estimated that 76% of the bicarbonate ions exported by the Zaïre River may be attributed to the atmospheric CO2 consumption by rock weathering.

Table 6.4. Average solute concentrations of major African rivers


Rivers  Precipitation (mm) Runoff
(mm)
RIP (%)  TDS (mg/l) TSS  (mg/l)  Transport 
106 t
 TSS/TDS

TDS
TSS
Zaïre 1520 338 22 28 37 36 .6 48 1.39
Niger 1140 124 11 67 127 14 .0 25

.4

1.86
Nile 510 47 9 318 54 11 .8 0.18
Senegal  650 48 7 42 196 0 .4 1 .9 2.44
Orange 380  15 4 140 57 1 .6 0 .7 0.44
Zambezi  1020 157 15 113 90 25 .2 20 0.80
Gambia  1100 219 20 17 19

.5

0 .08 0 .09 1.10

Relationship between alkalinity and TDS shows a strong positive correlation in the Niger (Figure 6.13) thereby accentuating the predominance of bicarbonates. Seasonal variation of dissolved elements has been reported in recent years for the Nile (Soliman 1983) and the Orange (Hart 1985). Noteworthy is the contrast in ionic variations between the Senegal and the Niger Rivers. Salinity increases between December and June, dropping sharply towards the end of July in the Senegal River. It remains stable throughout the high water period in August to November. This is slightly different from the conditions of the Niger, where ionic concentrations increase between May through August, as river discharge increases initially. A gradual rise in salinity was observed throughout the peak flood (September and October). The highest concentration level (Figure 6.9) was recorded in the dry season (February to April).

The dissolved load transport of major African rivers is included in Table 6.4. The values presented for the Niger, Senegal, Zambezi and Gambia Rivers are averages computed from monthly samples covering one year . Values for the Orange River spanned over three years of monthly sampling.


Specific transports 
(t/km2 /year)
Organic Carbon concentration (mg/l) Specific transports  
(t/km2 /year)
References 

Chem.
 Mech.
DOC
POC 
TOC
DOC
POC
TOC
9 .9 13 .8 8.5 1.1 9.6 2 .9 0.8 3 .7 Nkounkou and Probst (1987)
11 20 .5 3.5 3.7 7.2 0 .4 0.5 0 .9 Martins (1983) 
3 .9 0 .7 - - - - - - Schamp (1983)
1 .8

4

.4 - 1.5 - - - - Gac and Kane (1986a,b)
1 .6 0 .7 2.3 0.9 3.2 0 .03 0.01 .04 Hart (1985), 
Hall et al.(1977)
21 16 .7 - - - -

-

-  Milliman and 
Meade (1983)
1 .0 1 .1 2.4 1.1 3.5 0 .3 0.1 0 .4 Lô (1984) Lesack et al. 
(1985)

The quantity of dissolved materials exported through the respective drainage basins is mainly controlled by discharge. Thus, the Zaïre exports about 37 x 106 t of dissolved materials into the Atlantic Ocean. If the discharge rate of 225 km3/year (as suggested by Ambroggi 1980) is used to calculate the transport rate of dissolved elements by the Zambezi, then the river exports a staggering 25 million tonnes of dissolved load into the Indian Ocean. It is, however, noteworthy that investigations by Borchert and Kempe (1985) put the water volume of the river at less than 100 km3/year. The Gambia has the least dissolved load (80 000t/year).

The specific transport rate seems to be highest in the Zambezi basin (21 t/km2/year), followed by that of the Niger (11 t/km2/year). The Gambia, Orange, Senegal and Nile basins have values ranging between 1 and 5 t/km2/ year, while the Zaïre has a rate of 9.9 t/km2/year.

6.5.2 TOTAL SUSPENDED SOLIDS (TSS)

It is now common knowledge that the quality and quantity of suspended solids are determined largely by topography, climate, vegetation, geology, size and water volume of the drainage basin. However, the scope of influence of any individual process in nature cannot be delineated with much success since most of these factors exercise their impact simultaneously. To further complicate the issue, human factors, such as land use and large-scale urbanization, have been superimposed on the natural conditions so that it is no longer realistic to talk of a pristine situation in any major African river basin.

Figure 6.13 Relationship between alkalinity and total dissolved solids (TDS) in the Niger

The concentration of TSS, the suspended load and the specific transport rate for individual rivers are contained in Table 6.4. Nkounkou and Probst (1987) reported values for the Zaïre ranging from 12 mg/l to 58 mg/l, results from sporadic sampling carried out by different authors. An average TSS concentration of 37 mg/l was suggested. From data collected during the 1980/ 81 water year, Martins (1988) estimated an average of 127 mg/l for the Niger River, having integrated values obtained from both the Niger (at Lokoja) and Benue (at Umaisha). The Senegal seems to have the highest concentration (196 mg/l) as against 19.5 mg/l in the Gambia. The Nile, Orange and Zambezi have values of 54 mg/l, 57 mg/l and 90 mg/l respectively.

It is remarkable that suspended concentrations attain a maximum before the peak discharge in rivers like the Sanaga (Nouvelot 1972), the Niger (Martins 1982), the Senegal (Gac and Kane 1986a) and the Gambia (Lô 1984). This relationship between river discharge and sediment concentration is more pronounced in semi-arid rivers with highly seasonal flows, and broad floodplains than in tropical-equatorial rivers like the Zaïre (Figure 6.14) .

It is equally important to note that a greater proportion of the total load is transported in the solid phase in the Zaïre, Niger, Senegal and Gambia Rivers (Table 6.4). Most of the particulate load of the Nile is being deposited in the Aswan Dam (Schamp 1983), such that only 14.5% of its total load is particulate. Also, the low proportion of particulate load in the Orange and Zambezi Rivers may be attributed to the presence of large reservoirs in their respective drainage basins.

Figure 6.14 Comparison of variations of suspended sediment concentration (C or Cex) with discharge (Q or Qex) for (a) the Senegal (after Kattan et al. 1987); (b) Gambia (after Lô 1984); and (c) Zaïre (after Kinga-Mouzeo 1986)

6.5.3 ORGANIC CARBON

River export of organic matter from the continents to the oceans forms an important component in the global carbon cycle. Total organic carbon (TOC) transport has been estimated at 400 x 106 t/year (Meybeck 1982). The river flux of particulate organic carbon (POC) equals 180 x 106 t/year.

The average organic carbon concentrations for major African rivers are presented in Table 6.4. Dissolved organic carbon (DOC) values range between 2.3 mg/l in the Orange River to 8.5 mg/l in the Zaïre. There seems to be an increase in DOC values from the semi-arid to the tropical rain forest belt. This trend is no more recognizable in the particulate organic carbon (POC) concentration: the Orange still has the lowest concentration (0.9 mg/l), while the Niger, equally a semi-arid river, has the highest value (3.7 mg/l) .This may imply that different mechanisms regulate the release of the two fractions of carbon within the riverine system. Several mechanisms have been postulated to explain the concentration of DOC within a hydrological year. Some of these are: leaching of organics from floodplain soils; leaching and heterotrophic processing of newly flooded terrestrial vegetation; and primary production and subsequent processing of the fixed carbon in the water on the floodplain (Lesack et al. 1985).

Figure 6.15 Relationship between POC weight percentages and total suspended sediment concentrations (Cs) for some major African rivers. The points represent average values and the hatched areas represent point clusters, corresponding to several individual sample values. 1, Senegal 1983; 2 and 3, Gambia 1983 and 1982 (after Lô 1984); 4, Zaïre canyon; Zaïre River and estuary; 7, Niger (higher POC value); 8, Benue; 9, Sanaga; 10, Dilamba and 11, Ogooue (after Cadée 1984); 6, Zaïre at Brazzaville (after Kinga-Mouzeo 1986); 7, Niger (lower POC value) (after Martins 1983); 12, Orange (after Hart 1985)

In Figure 6.15, the relationship of particulate organic carbon percentage versus total suspended sediments is examined. Generally, POC contents decrease with increasing sediment concentration, which points to the fact that dilution of particulate organic matter by inorganic mineral substances plays an important role in most African rivers. In the case of the Zaïre, first results as presented by Kinga-Mouzeo (1986) reveal an increase in POC with increasing suspended sediment concentrations. This may indicate that the bulk of the suspended sediments in the Zaïre is organic.

6.6 CONCLUSION

Africa is an 'old continent' where the relief has been largely reduced and the soil profiles are leached, developing large areas of lateritic covers with iron duricrusts. Consequently, the dissolved and particulate specific river transports are very low in comparison to those of other continents. In addition, the river transports are influenced by anthropogenic activities and, particularly, by dam constructions which modify the hydrological regimes of rivers, trap a great quantity of river suspended sediments and increase river water salinity.

The specific transports of large African rivers range from 0.7 to 20.5 t/km2/ year for the particulate materials, and from 1 to 21 t/km2/year for the dissolved materials. The ratios between TSS and TDS are very low and range between 1 and 2, except for the Nile, Orange and Zambezi Rivers where the suspended sediments are retained by dams. For the Senegal (tropical river), the ratio reaches 2.4, whereas the recent construction of dams could probably affect the particulate loads and decrease the TSS/TDS ratio.

Nevertheless, the TSS/TDS ratio seems to be higher for rivers of tropical savannah (Senegal, Niger) than for rivers of the tropical forest (Zaïre). It is noteworthy that the peak of the suspended sediment concentrations reaches its maximum before the peak discharges, as in the tropical savannah rivers, or that it may correspond to the low water period, as in the Zaïre River. These lags may be attributed to an important contribution of autochthonous materials which are remobilized during the first rise in water level.

With regard to the organic materials, the POC percentage of suspended matter decreases when the TSS increases, going from the rivers of the tropical forest (Zaïre) to the rivers of the tropical savannah (Senegal and Orange). This relationship for African rivers follows the general world major river pattern. DOC and POC contents (except DOC in the Zaïre River), are lower than in the world's major rivers of other types (e.g. subarctic, temperate). Consequently, the specific organic transport rates are also lower than in other regions, ranging from 0.04 to 3.7 t/km2/year of TOC.

For the future, it would be important to select some target rivers on the basis of the different 'type'. For Africa, it would be judicious to focus on the Zaïre as a river of the tropical forest type and the Niger or the Senegal as rivers of tropical savannah. The fluctuations of dissolved and suspended loads would have to be monitored for a period of at least 10 years, in order to be able to appreciate the influences of man's activities and global climatic changes.

REFERENCES

Adeniji, H. A. (1973) Preliminary investigation into the composition and seasonal variation of the plankton in Kainji Lake, Nigeria. In: Ackerman, W. C., White, G. F. and Worthington, E. B. (Eds) Man-Made Lakes: Their Problems and Environmental Effects. Geophys. Monograph Ser., Vol. 17, pp. 617-19. 

Ambroggi, R. P. (1980) Water. Sci. Amer. (Sept.), 91-101.

Borchert, G. and Kempe, S. (1985) A Zambezi aqueduct. In: Degens, E. T., Kempe, S. and Herrera, R. (Eds) Transport of Carbon and Minerals in Major World Rivers, Pt. 3. Mitt. Geol.-Paläont. Inst. Univ. Hamburg. SCOPE/UNEP Sonderbd. 58, pp. 443-57.

Cadée, G. C. (1984) Particulate and dissolved organic carbon chlorophyll a in the Zaïre river, estuary and plume. Neth. J. Sea Res. 17(2-4),426-40.

Faure, H. and Gac, J. Y. (1981) Will the Sahelian drought end in 1985? Nature 291, 475-8.

Gac, J. Y. and Kane, A. (1986a) Le fleuve Sénégal. I- Bilan hydrologique et flux continentaux de matières particulaires à l'embouchure. Sci. Geol. Bull. 39(1), 99- 130.

Gac, J. Y. and Kane, A. (1986b) Le fleuve Sénégal. II- Flux continentaux de matières dissoutes a I'embouchure. Sci. Geol. Bull. 39(2), 151-72.

Grove, A. T. (1978) Africa, Oxford Univ. Press, 337pp.

Hall, A., Maria, I., Valente, C. B. S. and Davies, B. R. (1977) The Zambezi river in Mosambique: the physico-chemical status of the middle and lower Zambezi prior to the closure of the Cabora Bassa dam. Freshwater Biol. 7, 187-206.

Hart, R. C. (1985) Aspects of the hydrogeochemistry of the upper Orange river. In: Degens, E. T., Kempe, S. and Herrera, R. (Eds) Transport of Carbon and Minerals in Major World Rivers, Pt. 3. Mitt. Geol.-Paläont. Inst. Univ. Hamburg, SCOPE/ UNEP Sonderbd. 58, pp. 435-42.

Kat tan, Z., Gac, J. Y. and Probst, J. L. (1987) Suspended sediment load and mechanical erosion in the Sénégal basin. Estimation of the surface runoff concentration and relative contributions of channel and slope erosion. J. Hydrol. 92,59-76.

Kempe, S. (1982) Long-term records of CO2 pressure fluctuations in fresh waters. In: Degens, E. T. (Ed.) Transport of Carbon and Minerals in Major World Rivers, Pt.1. Mitt. Geol./Paläont. Inst. Univ. Hamburg, SCOPE/UNEP Sonderbd. 52, 91- 332.

Kempe, S. (1983) Impact of Aswan High Dam on water chemistry of the Nile. In: Degens, E. T., Kempe, S. and Soliman, H. (Eds) Transport of Carbon and Minerals in Major World Rivers, Pt. 2. Mitt Geol.-Paläont. Inst. Univ. Hamburg, SCOPE/ UNEP Sonderbd. 55, pp. 401-23.

Kempe, S. (1989) Reservoir case study: Lake Nasser. In: Meybeck, M., Chapman D. and Helmer, R. (Eds) Global Freshwater Quality, A First Assessment, Ch. 2, pp. 243-52, GEMS, WHO, UNEP, Blackwell, Oxford.

Kinga-Mouzeo (1986) Transport particulaire actuel du fleuve Congo et de quelques affluents. Enregistrement quaternaire dans l'éventail détritique profond (sédimentologie, minéralogie et géochimie). Thèse doct., Univ. Perpignan, 261pp.

Lesack, L. F. W., Melack, J. M. and Hecky, R. E. (1985) Transport of organic carbon in the Gambia River, West Africa. In: Degens, E. T., Kempe, S. and Herrera, R. (Eds) Transport of Carbon and Minerals in Major World Rivers, Pt. 3. Mitt. Geol.- Paläont. Inst. Univ. Hamburg, SCOPE/UNEP Sonderbd. 58, pp. 431-5.

Lô, H. (1984) Le bassin de la Gambie: contribution à l'hydrologie et à la dynamique fluviale en milieu tropical humide africain. Thèse 3ème cycle, Univ. Nancy II, 396pp.

Martins, O. (1982) Geochemistry of the Niger River. In: Degens, E. T. (Ed.) Transport of Carbon and Minerals in Major World Rivers, Pt. 1. Mitt. Geol.- Paläont. Inst. Univ. Hamburg, SCOPE/UNEP Sonderbd. 52, pp. 397-418.

Martins, O. (1983) Transport of carbon in the Niger River. In: Degens, E. T., Kempe, S. and Soliman, H. (Eds) Transport of Carbon and Minerals in Major World Rivers, Pt. 2. Mitt. Geol.-Paläont. Inst. Univ. Hamburg, SCOPE/UNEP Sonderbd. 55, pp. 435-49.

Martins, O. (1988) Flux of particulate inorganic matter through the Niger River into the Atlantic Ocean. Neth. J. Sea Res. 22/2, 91-7.

Meybeck, M. (1982) Carbon, nitrogen, and phosphorus transport by world rivers. Amer. J. Sci. 282, 401-50.

Milliman, J. D. and Meade, R. A. (1983) World-wide delivery of river sediments to the oceans. J. Geol. 91, 1-21.

Nkounkou, R. R. and Probst, J. L. (1987) Hydrology and geochemistry of the Congo river system. In: Degens, E. T., Kempe, S. and Gan Wei-Bin (Eds) Transport of Carbon and Minerals in Major World Rivers, Pt. 4. Mitt. Geol.-Paläont. Inst. Univ. Hamburg, SCOPE/UNEP Sonderbd. 64, pp. 483-508.

Nouvelot, J. F. (1972) Le régime des transports solides en suspension dans divers cours d'eau du Cameroun de 1969 à 1971. Cah. ORSTOM, série Hydrol. IX(l), 47-74.

Probst, J. L. and Tardy, Y. (1987) Long range streamflow and world continental runoff fluctuations since the beginning of this century. J. Hydrol. 94, 289-311.

Rodier, J. (1964) Régimes hydrologiques de l'Afrique Noire à l'Ouest du Congo. Mem. ORSTOM No.6, 137pp.

Schamp, H. (1983) Sadd el-Ali, der Hochdamm von Assuan, I, II. Geowiss. in unserer Zeit, 1/2,51-59, 1/3, 73-85.

Soliman, H. A. (1983) Nile research at the Asyut University. In: Degens, E. T., Kempe, S. and Soliman, H. (Eds) Transport of Carbon and Minerals in Major World Rivers, Pt. 2. Mitt. Geol.-Paläont. Inst. Univ. Hamburg, SCOPE/UNEP Sonderbd. 55, 385-99.

Back to Table of Contents

 

The electronic version of this publication has been prepared at
the M S Swaminathan Research Foundation, Chennai, India.