PART II - SYNTHESES OF INFORMATION ON SELECTED AFRICAN RESERVOIRS (continued)

PARTIE II - SYNTHESES DE L'INFORMATION SUR CERTAINS RESERVOIRS AFRICAINS (continuer)

KAINJI (NIGERIA) (continued)

Table 34
Estimated Income Outlay for the Pilot Fisheries Scheme at Shaguun and Moai - Kainji Lake, Nigeria
(From Ita and Eyo, 1981)

Expected average catch/boat/day (kg)	Total landings by 100 boats per day (kg)	Cost price at an average of 1.50/kg	Selling price at an average of 200/kg	Gross income/day	Gross income per year for 250 working days
20	2 000	3 000.00	4 000.00	1 000.00	250 000.

Table 35
Revolving Inputs to 100 Fishermen
(From Ita and Eyo, 1981)

Equipment	Unit Cost	No. of Fishermen	Total Cost
Boat	600	100	60 000
Engine	800	100	80 000
6 Bundles of Nets	40	100	24 000
50 Cords of Twine	1	100	5 000
		Total	169 000

KARIBA (ZIMBABWE/ZAMBIA)

by

B.E. Marshall
Officer-in-Charge
Lake Kariba Fishery Research Institute
P.O. Box 75
Kariba, Zimbabwe

1. LITERATURE

There is an extensive literature on various aspects of the Kariba ecosystem. Bibliographies were produced by Coche (1971) and Marshall (1979); review papers with wide reference lists include Balon and Coche (1974), McLachlan, A.J. (1974) and Marshall et al., (1982). The more important papers from a fisheries point of view, include:

(i) Physico-chemistry, morphology and hydrology:

Allison (1969), Begg (1969, 1970, 1974a), Bowmaker (1973, 1976), Caulton (1970), Coche (1968, 1969, 1974), Harding (1961, 1964, 1966), McLachlan, S.M. (1970), McLachlan and McLachlan (1971), Mitchell, D.S. (1970, 1973), Robarts and Southall (1977), and Ward (1978, 1979).

(ii) Plankton and Benthos:

Begg (1974, 1976), Bowmaker (1973), Kenmuir (1980a), Lake Kariba Report 35, 36, McLachlan, A.J. (1969, 1969a, 1970, 1970a), McLachlan and McLachlan (1971), Magadza (1980), Mills (1973) and Mitchell, S.A. (1975).

(iii) Macrophytes:

Bowmakers (1973, 1973a), Marshall and Junor (1981), Mitchell, D.S. (1969, 1970), Mitchell and Rose (1979) and Thomas (1974).

(iv) Fish Systematics and Distribution:

Balon, (1974, 1974a, 1975), Begg (1974a), Bell-Cross (1972, 1976), Bowmaker et al., (1978), Jackson (1961), Jubb (1967, 1976, 1976a) and Kenmuir (1975), Mitchell, S.A. (1976, 1978).

(v) Fish Biology and Ecology:

Balon (1971, 1974), Cochrane (1978), Coke (1968), Donnelly (1969), Jackson (1960), Joubert (1975), Kenmuir (1973, 1975a), Lake Kariba Report No. 17, Marshall (1982) and Mitchell S.A. (1976, 1978).

(vi) Population Dynamics, Production and Commercial Statistics:

Balon (1973, 1974a), Bazigos et al. (1975), Cochrane (1978), Junor (1981), Kenmuir (1973a), Lake Kariba Report 17, 38, and 40, Mahon and Balon (1977), Marshall (1979a, 1982a), Minshull (1973), Mitchell, S.A. (1976).

Annual catch statistics are also issued by the Lake Kariba Fisheries Research Institute for the Zimbabwean side, and for Zambia by the Central Statistics Office, Lusaka.

(vii) Socio-economic Aspects:

Balon (1978), Colson (1960, 1971), Scudder (1962, 1972, 1975, 1980), Skehel (1969), and Webster (1975).

2. THE ZAMBEZI RIVER

The middle Zambezi extends from the Victoria Falls to the Cahora Bassa Gorge. It has been greatly modified by the construction of two large dams, Kariba and Cahora Bassa, and proposals have been made to construct three more. This would completely alter the nature of this part of the river, about which very little is known.

2.1 Pre-Impoundment Studies, Kariba Area

Limited pre-impoundment surveys were carried out in the Kariba area from 1956–58 (Jackson, 1959, 1961). The most striking result of these surveys was the depauperate fish fauna, with only 28 species being collected (Table 1) in contrast to the Upper Zambezi which has 83 (Bowmaker et al., 1978). This was particularly marked with regard to smaller species (15 cm) of which only eight were recorded.

This was attributed to the physical conditions in the river, termed a “sandbank” river and characterized by extreme differences between high and low flow levels. At low water there were extensive sandbanks with few aquatic macrophytes; small fish were thus extremely vulnerable to predation by the tigerfish Hydrocynus vittatus. By contrast the Upper Zambezi was a “reservoir” river with smaller water level fluctuations and abudant macrophytes which enabled small species to escape predation.

Few cichlids were taken and it was recommended that they should be introduced into the new lake (in the event this was unnecessary). It was also recommended that a pelagic planktivore, such as Limnothrissa or Stolothrissa should be introduced as the local species would not colonize open water. This proved to be fully justified.

Jackson acknowledged that these surveys were limited in scope and that much of the Zambezi itself and none of its tributaries had been examined. The list of species found was not comprehensive and other species would undoubtedly occur. He also suggested that small species would become more abundant once the lake was formed because of the cover provided by submerged macrophytes.

2.2 Post-impoundment Studies, Zambezi River

The only post-impoundment studies of the Zambezi below Kariba were carried out as part of a pre-impoundment survey of the proposed Cahora Bassa dam. They included the physico-chemical studies (Hall, Valente and Davies, 1977), zooplankton (Hall, Davies and Jackson, 1976) and a broad, predictive ecological summary (Davies, Hall and Jackson, 1975). Data on the fish populations were provided by Jackson and Rogers (1976).

The main conclusions from this work were that the Zambezi below Kariba was richer in nutrients but it was not clear how much of this was contributed by the Kafue. Since the flow was now regulated by the dam, the river was changing from a “sandbank” type to a “reservoir” type, so becoming a more favourable habitat for small fish. As a consequence, Jackson and Rogers (1976) were able to record 38 species from this part of the river, compared with 28 from the pre-Kariba river (Jackson, 1961) and 41 from Kariba lake itself (Marshall et al., 1982).

2.3 Conclusions

The ecology of the Zambezi is poorly-known when compared to the extensive data collected from Kariba. This is an important ommission because the Zambezi from Kariba to the Mozambique border supports an important fishery on the Zambian side (the Zimbabwean side is a wildlife area). This yields about 1 000 t yr^-1 (Zambian Fisheries Department, pers.comm.), compared to 1 200 t from the Zambian side of Lake Kariba (Bazigos et al., 1975). This highlights the need for further research in the area.

The effects of Kariba on downstream terrestrial ecosystems are not fully understood. Attwell (1970) suggested that Kariba was causing desiccation of the Mana Pools floodplain with consequent effects on vegetation and wildlife ecology. Guy (1980) showed that changes to the flow regime had increased river bank erosion below Kariba.

3. LAKE KARIBA

3.1 Ecosystem Development

Man-made lakes provide an opportunity to observe and understand the development of lake ecosystems. Lake Kariba was the first great man-made lake in Africa and many of the phenomena which occurred after its construction have since been observed in other such lakes. Reviews which summarize this information include McLachlan, A.J. (1974), Balon and Coche (1974) and Marshall et al., (1982).

The lake's history can be divided into three main periods (Coche, 1974). The “filling” phase lasted from December 1958 to September 1963 when the lake reached its maximum level of 487.8 m a.s.l. The “transition” phase lasted until mid-1966 and was characterized by relatively large lake level fluctuations. From 1966 the lake was thought to have stabilized; lake level fluctuations rarely exceeded 3 m yr^-1. In 1974, the mean lake level was raised by 2 m and has been kept at that level since then.

The first result of the lake's formation was an immediate increase in dissolved nutrients and conductivity rose from about 55 μS cm^-1 in 1958 (Zambezi River) to about 120 μS cm^-1 in 1959. These nutrients were leached from newly-flooded land as well as decaying organic matter.

During the filling phase Lake Kariba was eutrophic and was characterized by the explosive growth of aquatic organisms (Jackson, 1960; Balinsky and James, 1960; Harding, 1964, 1966). The most important of these was the fern Salvinia molesta which, by 1962, covered about 22 percent of the lake's surface (Mitchell, 1969).

Salvinia had an important role in retaining nutrients during this phase; from 1969 to 1962 conductivity declined even though Salvinia was expanding (Fig. 1). At its peak Mitchell (1973) estimated that 7 925 t of nitrogen and 417 t of phosphorus was retained in the weed mats.

Salvinia decline in 1963 when the lake reached its maximum level and conditions became less suitable, and again in 1964 when the lake level was dropped some 7 m. At first there was an increase in nutrients in the water, shown by nitrate-nitrogen (Fig. 2), but this was reduced through spilling. Conductivity declined throughout the filling phase and was only 75 μS cm^-1 in 1963.

During the transition phase conductivity steadily rose to about 85 μS cm^-1 in 1966. Salvinia coverage also rose to about 15 percent in 1967 and lake level fluctuations were gradually reduced. This period saw the establishment of aquatic macrophytes. Benthic fauna were not investigated during the stabilization period with the exception of the freshwater mussels (Kenmuir, 1980).

Although Kariba is presently in its “stabilization” phase, the situation is still dynamic. Changes in the inshore fish population have occurred and may continue. The length of time (25 year) is still too short, and the scale of environmental changes, such as water inflows and outflows or fluctuations in water level too great to permit true stability to have developed.

3.2 Factors Affecting Fish Production in Lake Kariba

3.2.1 Morphology and hydrology

The main morphometric features of some African man-made lakes are given in Table 2. Lake Kariba is one of the largest, exceeded in size by only Volta and Nasser-Nubia. However, Kariba has the greatest mean depth and is less dendritic (indicated by shoreline development). This suggests that Kariba is the least favourable for the development of inshore fish stocks and most suited to a pelagic fishery.

The Zambezi River supplies about 80 percent of Kariba's water (Table 3) and about 14 percent comes from the secondary rivers. This is important because the Zambezi is relatively poor in nutrients (Coche, 1974) and is the major influence on the chemistry of the lake. A further influence is the rapid replacement time; the outflow: volume ratio is about 1:3, so the replacement time is around three years. The effects of this will be discussed in more detail in subsequent sections.

3.2.2 Temperature and oxygen

Detailed description of the temperature and oxygen regime of Lake Kariba may be found in Harding (1966), Coche (1968, 1974), Begg (1970, 1974) and the following account is drawn from them.

The lake is warm and monomictic with temperatures ranging from over 30°C at the surface in summer to about 21°C in winter. The thermocline develops in about mid-August descending from 10 m in September to 30 m in May (Fig.2). Stratification breaks down in late June and the lake is truly isothermal for only about two months. This pattern began in the lake's first year and has remained constant since then (Fig.3). The hypolimnion was strongly deoxygenated to begin with and hydrogen sulphide (H₂S) occurred throughout. From about 1963 deoxygenation became less intense and H₂S is rarely found (Lake Kariba Fisheries Research Institute, unpubl.data).

This sequence of events is linked to the lake's early history. Nutrients leached from flooded land and vegetation led to the deoxygenation and production of H₂S in the hypolimnion. This was gradually reduced as the lake stabilized, and nutrients were either lost through the outflow or taken up by biological communities. Similar processes occurred in other large lakes, particularly Lake Volta (Viner, 1970) where possibly because of more abundant vegetation, even surface waters were deoxygenated. This has also been recorded from small Zimbabwean reservoirs where deoxygenation seems more persistent and was found in lakes over 60 years old (Mitchell and Marshall, 1974). This may be because in a small lake a much greater proportion of the water body is in contact with the bottom and that autochthonous nutrient supplies will be proportionately greater.

The importance of the temperature and oxygen cycles to fish production lies in the effects they may have on fish distribution. Deoxygenation may also stress fish populations. In Kariba inshore fish occur down to about 15 m (Coke, 1968) but this is probably not linked to the thermocline. As observed by divers, the pelagic Limnothrissa, however, may go below the thermocline during the day which they would not be able to do if the hypolimnion was anaerobic. There have been no reports of fish kills caused by seiche movements or other disruptions of the thermocline in Kariba.

3.2.3 Water chemistry

Basic information on chemistry of Kariba is available in Coche (1968, 1974) as well as Harding (1961, 1966) and Mitchell (1970, 1973). Chemical processes in estuarine zones were investigated by Caulton (1970) and Bowmaker (1976). Limiting factors were determined from algal bioassays by Robarts and Southall (1977).

Lake Kariba is a slightly alkaline, oligotrophic lake with a low potential fish production. This can be assessed by the Morphoedaphic Index (MEI) as applied to a number of African lakes (Henderson and Welcomme, 1974). The MEI is calculated as conductivity (μS cm^-1)/mean depth (m) and the higher the MEI the higher the potential production, based on the premise that shallow lakes are more productive than deep ones of a similar chemical status. Henderson and Welcomme derived a relationship between MEI and yield in 17 heavily-exploited lakes, in which

Y = 14.3136 MEI^0.4581

where Y = yield in kg ha^-1

Applying this to Kariba it can be seen that it has the lowest potential yield of the large African man-made lakes as well as some smaller Zimbabwean lakes (Table 4).

The most important limiting factors in Kariba are phosphorus and nitrogen as the lake is low in both (Figs.4 and 5). A further indication of its low productivity is given by a comparison of Kariba with the average ionic composition of African and World fresh water (Table 5). The reasons for this low potential productivity lie in the nature of the inflows and outflows. The Zambezi is generally low in nutrients (Table 6), as much of its nutrient is lost on the Barotse floodplain, above the Victoria Falls (Mitchell, 1973). This is also shown by the low phosphorus loading to the lake (Fig.6). The smaller inflowing rivers are generally richer than the Zambezi but as they only contribute 14 percent of the lake's water, their influence is only felt locally (Mitchell, 1973; Marshall, 1982).

The loss of nutrients through the outflow is also important. No detailed nutrient budget has yet been undertaken but a preliminary estimate made by Coche (1974) suggested that nutrient losses may be higher than imports (Table 7). This data was collected in 1965, when the lake was still in its post-filling stage and it seems unlikely that this is now the case. There is an urgent need for a detailed study of Kariba's nutrient budget.

3.2.4 Species composition

About 40 species of fish occur in Lake Kariba, compared with 28 that were recorded before the lake was formed. Much interest was aroused by the assertion that 13 species had invaded the lake from the Upper Zambezi above Victoria Falls which were considered to have been a barrier to downstream migration. There was some speculation about why the other 50 Upper Zambezi species had not been able to colonize the lake (Balon, 1971a, 1973, 1974, 1974a).

If Balon's views were correct then “not even one of the world's greatest waterfalls is a barrier to zoogeographically distinct fish populations” (Bowmaker et al., 1978) so it is necessary to investigate this claim. A list of fish species known to occur in Lake Kariba or the Middle Zambezi is given in Table 8. A number of these are widely distributed in the Middle Zambezi tributaries on the Zimbabwean side (Jubb, 1961, 1967; Bell-Cross, 1972, 1976) whence they colonized the lake. These include Marcusenius macrolepidotus, three small Barbus spp. (paludinosus, unitaeniatius and lineomaculatus) and Labeo cylindricus.

Only eight species may therefore have invaded from the Upper Zambezi. Barbus poechii was found in two north bank tributaries, the Kalomo River above Kariba (Balon, 1974) and the Lusito River below (Balon, 1974a). It may well have occurred in other tributaries and reached the lake from them. The same could have applied to two of the cichlids, Oreochromis andersoni and Serranochromis robustus, which both occurred in the Kalomo River. The latter has also been widely translocated in Zimbabwe and is found in many Kariba tributaries (Toots and Bowmaker, 1976).

Of the remaining Upper Zambezi cichlids only Serranochromis macrocephalus is at all common and is most abundant in the western portion of the lake. This part of the river was not surveyed by Jackson (1961) and it was probably overlooked. Both S. robustus and S. macrocephalus were also taken from a dam on the Zongwe River, a Kariba tributary (J.F.R.O. Annual Report, 1961), whence they could have reached the lake. The same applies to S. giardi and S. carlottae. There is also a strong possibility that they were introduced from the Kafue in 1959–61 when some 26 t of cichlids were introduced from Chilanga, Zambia (Jubb, 1976). It was originally intended to introduce only Oreochromis macrochir and Tilapia rendalli but the stocks were known to be impure and may have included the other cichlids which all occur throughout the Kafue.

Two species remain; Schilbe mystus is rare in Kariba but was known from the Middle Zambezi (Jubb, 1967) and was unlikely to have invaded Lake Kariba from above the Victoria Falls. Alestes lateralis is abundant in the lake and was considered to have invaded from the Upper Zambezi (Balon, 1971a). It has been suggested that it is not A. lateralis but a species related to A. humilis (Jubb, 1976; Bowmaker et al., 1978) but no details have been given and the species must continue to be called A. lateralis for the present. Jubb (1976) claims that specimens of the species were taken from the Middle Zambezi in 1956 and are now in the Queen Victoria Museum, Harare. These specimens were found to be mislabelled A. imberi. This species may therefore be a true invader from the Upper Zambezi. If this is so, it may well have come through the Victoria Falls power station (Jubb, 1976a) in the same way as Limnothrissa miodon are known to have passed through the Kariba turbines (Kenmuir, 1975) and moved down the Zambezi to colonize Lake Cahora Bassa.

This discussion on the species composition of Kariba highlights two problems. The first is the lack of an adequate pre-impoundment survey which was acknowledged by Jackson (1961) who predicted that new species would later be found in Lake Kariba with further collecting. The second is the unusual political circumstances of the area. Kariba is the only African lake under control of two countries which, until recently, had no diplomatic relations and were virtually at war. There was little contact between workers in the two countries; if there had been, the confusion over origin of “Upper Zambezi” species in the lake would not have arisen. This has been a constant problem in the management of Kariba and one which has not yet been fully overcome.

3.2.5 Distribution and changes in fish populations

Since most Lake Kariba species are of riverine origin, it was predicted (Jackson, 1961) and later shown (Coke, 1968; Balon, 1974) that their distribution was limited to shallow water. This was clearly demonstrated by Coke (1968) who set gillnets in a cleared area and found that fish were most abundant in shallow water (Fig.7). Their numbers decline rapidly, however, and none were taken below 15 m. Balon (1974) considered the limit of fish distribution to be 20 m; this means that only 30 percent of the lake bottom could be inhabited by indigenous fish.

An important factor apparently influencing the distribution of these fish is the availability of submerged macrophytes. So far there is very little data on these plants in Kariba but they are limited to about 12 m depth (Fig.8) with their greatest abudance between 2–6 m (Kenmuir, 1980). Other workers have also noted the correspondence between vegetation and fish abundance (Table 9). This is because the vegetation provides cover for fish fry as well being a substrate for invertebrates (McLachlan, 1969a; Mitchell, 1967). Submerged vegetation was slow to develop in Kariba, possibly inhibited by Salvinia mats or lake level fluctuations (Mitchell, 1968). Vegetation is also limited by the lake's morphometry as it is relatively deep and has extensive rocky areas. This is a futher constraint to fish production in Kariba. In its early years Salvinia mats and drowned trees provided cover and a substrate for invertebrates (McLachlan, 1970; Mitchell, 1976). This was of importance while the lake was unstable and before submerged macrophytes had become established.

As previously discussed, most of the riverine fish are restricted to water 15–20 m deep. Some fish do occur in deeper water, notably eels Anguilla nebulosa labiata (Balon, 1975) and occasionally Hippopotamyrus discorhynchus and Mormyrus longirostris (J.D. Langerman, unpubl.), but their numbers were always low. Fish distribution is also influenced by limnological conditions in the lake. The two western basins appear to be more riverine in character which favours the cyprinids Labeo altivelis and L. congoro (Fig.9). These fish were accordingly more abundant in these basins.

The only true pelagic fish in the lake is the introduced sardine Limnothrissa miodon. The tigerfish Hydrocynus vittatus became partially pelagic as a result (Marshall et al., 1982), but no other indigenous species has done so. It was suggested that Alestes lateralis would have, but for the introduction of the sardine (Balon, 1974); however, they were taken in water 20 m deep, and may not have moved into open water.

Change is perhaps the most characteristic feature of fish populations in a man-made lake. Although Kariba is one of the oldest and is now in its “stabilisation” phase its fish populations cannot be considered stable and further changes could still take place. Several factors may bring these about. Fish introductions are common in man-made lakes and may radically alter their fish fauna. Only two of any importance have been in Kariba. About 26 t of cichlid fry were stocked from ponds at Chilanga, Zambia, in 1959–61 as it was feared these would not be abundant in the new lake. The intention was to stock Tilapia rendalli and Oreochromis macrochir but other cichlids may also have been introduced (see 3.2.4). This was, in the event, unnecessary as the river species (T. rendalli and O. mortimeri) expanded rapidly after the lake formed (Jackson, 1960; Harding, 1966).

The introduction and spread of Limnothrissa miodon is well documented (Bell-Cross and Bell-Cross, 1971; Junor and Begg, 1971) and is the most striking success in the management of the Kariba fishery. Further changes occur as part of the lake's maturation process, or as a result of subsequent environmental changes, or because of commercial exploitation. These influences cannot always be separated.

Few long-term data are available except for a long-term gillnetting programme which has been carried out at “Lakeside” near Kariba town (Table 10). A standard fleet of 12 gillnets, ranging from 37 mm (1 ½ in) increments was used with each net being about 25 m long so that the total fleet is 300 m long. The fleet is laid almost every week and so the catch is expressed as number per 50 standard fleet settings.

The major environmental influence that occurred during this period was a change in the mean lake level in 1974, from about 483.5 m a.s.l. to nearly 486 m a.s.l. (Fig.10). Limited commercial fishing took place from 1977 to 1980, but this probably had little effect.

The data in Table 10 can be used to show changes in the major groups, even though one station cannot be typical of the whole lake.

(i) Mormyridae

Four momyrids were taken during this programme. Mormyrops deliciosus was the least abundant although it increased slightly after 1974. It is an opportunistic feeder, taking a high proportion of fish (Joubert, 1975) and it may have benefitted from the sardine introduction.

The other three species are of interest because their diets tend to overlap and their trophic relationships are obscure (Bowmaker, 1973; Joubert, 1975; Blake, 1977). Some competition between them could therefore be expected.

They first became abundant in 1969 and reached a peak in 1971–73 (Fig.11a). Marcusenius macrolepidotus was the most important species to begin with, but it declined as Hippopotamyrus discorhyncus increased. Mormyrus longirostris became increasingly important from 1973 onwards (Fig.11b).

(ii) Characidae

Two characids were taken in the gillnets. The population of Alestes imberi increased during the rich filling phase but could not maintain itself, declining steadily from 1960 to 1974; it was during this period that A. lateralis appeared to be replacing A. imberi (Balon, 1971a). However, in 1975 A. imberi became more abundant and has remained an important component of the catch. The reasons for this are not clear; one possibility is that the rise in the lake level in 1974 (Fig.11) created favourable conditions. Another is that it is unable to compete with H. vittatus and its numbers increased whilst the tigerfish was declining.

The tigerfish responded immediately to the lake's formation and large numbers were taken. There appeared to be a slight decline between 1964 and 1967 followed by an increase; data for this period are sparse however and the situation may have been relatively stable between 1960 and 1970. Limnothrissa reached the eastern end of Kariba in 1970 and became the tigerfish's most important food item (Kenmuir, 1971). This led to a population increase which continued until 1974 when commercial purse seining led to a rapid decline. From 1979 there was some evidence to suggest that the population was recovering to at least pre-1970 levels.

These trends were apparent in both the inshore fishery and the pelagic fishery. Tigerfish catches in the former reached a peak in 1974, then declined but in 1978 levelled off at about its 1973 level (Fig.13). Pelagic catches rose from 1973 to 1977, then declined rapidly. These were high during the period when gillnet catches were declining and it can be suggested that the removal of pelagic fish affected recruitment to the inshore ones by removing future breeding stock.

Another trend that can be shown clearly is the overall reduction in individual size, reflected by catches at the Kariba International Tigerfish Tournament (Fig.14). Only fish 1 kg are weighed in, and the mean size from 1962–1980 was 2.9 kg. There was relatively little deviation from this although fish taken from 1962 to 1965 were above average in size. The catch per angler, however, shows a decrease from nearly 15 kg in 1962 to about 3 kg in 1980. A fitted curve shows that the decline was initially rapid but later slowed down and is now relatively small.

The fish taken in 1962 were clearly those which had spawned and grown rapidly in the highly productive early post-impoundment years. Since then growth rates slowed and the number of large fish has not increased. The abundant tigerfish present from 1971 to 1974 were mostly small fish.

(iii) Cyprinidae

The cyprinids generally do not adapt to lacustrine conditions. Both Labeo altivelis and L. congoro were initially abundant, probably because they bred in the Zambezi just before the dam was closed and their survival and growth rates were good. However, they rapidly declined through lack of recruitment and none were taken in 1967. Only L. altivelis is still taken in small numbers, and L. congoro is restricted entirely to riverine areas.

The labeos are more abundant in the western end of the lake (Fig.9) and in estuaries; in the Mwenda, for example, L. altivelis made up 12.6 percent of the biomass (J.D. Langerman, unpubl.)

(iv) Catfish

Four catfish (various families) were taken in the experimental nets. Eutropius depressirostris was rare at first but steadily increased until 1971. Since then its numbers have fluctuated considerably.

Synodontis zambezensis was less abundant than E. depressirostris and was rarely caught until 1969. Since then, it has been taken in small, but fluctuating numbers.

Clarias gariepinus was initially abundant but its numbers declined although it showed another increase in 1974. Since then its numbers have fluctuated and generally declined. Evidence from smaller Zimbabwean lakes suggests that this species is not well adapted to lacustrine conditions (Marshall, 1977).

Heterobranchus longifilis would also appear to be poorly adapted to lake conditions and none were taken before 1970. Since then they have been caught in small numbers.

(v) Distichodontidae

Two Distichodus spp. occur in the lake and both show a similar trend to the Labeo spp. They were quite abundant at first but declined rapidly and are rarely caught at present. D. schenga still occurs in deep rocky areas which approximate to river conditions but D. mossambicus is now very rare in the lake.

(vi) Cichlidae

The cichlids are probably the most important family in the lake and the one which has adapted most successfully to lacustrine conditions. Five important species occur in the lake but two were infrequent at the Lakeside station.

Oreochromis macrochir was introduced soon after the lake filled. They first appeared in the nets only in 1974, and have been taken in small numbers ever since. Serranochromis macrocephalus is more abundant in the western part of the lake; in the Mwenda it made up 2.9 percent of the total biomass (J.D. Langerman, unpubl.). It appears to have been spreading eastwards and one specimen was taken at Lakeside in 1982, where it is expected to increase in numbers.

The remaining species, Oreochromis mortimeri, Serranochromis codringtoni and Tilapia rendalli are the most abundant cichlids. Their numbers were initially high but declined to a very low level in 1965 (Fig.15a). They apparently remained low until 1970 when they steadily began to increase, reaching a peak in 1974–75. Another decline then took place until 1980; 1982 data suggest that the population has begun to increase again. A striking feature is the steadily increasing proportion of S. codringtoni with O. mortimeri becoming less important (Fig.15b).

These fluctuations are difficult to explain. The increased lake level after 1974 may have improved survival and enhanced growth of submerged macrophytes, but the population had begun to increase before this happened. Similarly the later decline could be attributed to the commercial fishing that took place from 1978–1980, but it had already started before fishing began.

The Lakeside station on which most of this discussion is based is probably typical of lacustrine parts of the Kariba basin and so cannot be extrapolated to other areas. However, they do show the importance of long-term monitoring and the desirability of further sampling (although limitations of finance and manpower often prevent this). It also clearly illustrates that although a lake may be apparently stable, this may not be so and changing fish populations can be expected for some time.

3.2.6 Stock assessment, population dynamics and yield prediction

3.2.6.1 Biomass and stock assessment

A major study of the inshore species was made by Balon (1973, 1974) working on the northern shore. The last publication in particular presents a vast amount of data and is an attempt to summarize growth, biomass and production of the major fish species. Unfortunately, its value is reduced by the use of unusual and poorly-defined concepts, complex mathematical techniques and conflicting final calculations.

Other attempts to estimate the biomass of inshore species were made by Mitchell (1976) and Langerman (unpubl.).

Biomass estimates were made by poisoning blocked off coves (Balon, Mitchell) or by using an explosive grid (Langerman). The latter enabled estimates to be made in areas not readily poisoned. Standing crops obtained by these methods were faily similar (Table 11) although species composition varied considerably.

Balon's estimate of 580 kg ha^-1 was the highest (in Balon, 1973, Table 5, the figure can be recalculated as 533 kg ha^-1; this is also the estimate used in Balon, 1974a, p.40). This was probably because he sampled a wider range of habitats. Mitchell and Langerman's estimates were made in a smaller range of habitats and in restricted areas; they are almost identical.

Balon also recorded more species, again probably because his programme was more extensive. Mitchell examined areas with submerged macrophytes and Salvinia mats whilst Langerman was specifically studying H. vittatus populations and sampled areas where this species would be abundant. This work was carried out in the Mwenda estuary, which accounts for the high proportion of L. altivelis. Submerged macrophytes were abundant and the standing crop of T. rendalli was accordingly high. Salvinia mats were still present when Balon and Mitchell made their estimates and this may account for the abundance of mormyrids which appear to favour these mats (Joubert, 1975).

These data are of interest as they suggest that the lake can support a similar biomass in similar areas, even though species composition may vary.

A major drawback of this type of programme is that it tends to examine areas which are most suitable for fish in any event. This is particularly true of blocked off cove sampling and neither Balon nor Mitchell were able to examine steep, rocky exposed shores. The importance of this was shown by Langerman who found that the abundance of fish declined rapidly as the slope of the bottom increased and vegetation decreased (Table 9; Fig.16). The biomass in the most favourable areas was comparable to those obtained by Mitchell and Balon (about 500 kg ha^-1) but was much smaller in less favourable ones.

This is important if an attempt is to be made to estimate total standing crop by extrapolating samples over the whole lake. In Kariba a high proportion of the shore is rocky, “armoured” with boulders, steep and exposed and so unsuitable for fish (no data available, but at least 40 percent of the Zimbabwean shore may be of this type). Estimates made without taking this into account are thus likely to be too high.

Balon attempted to estimate total standing crop in this way; unfortunately his figures are contradictory. In Balon (1973) he stated that fish were limited to 25 percent of the 0–20 m depth zone, and estimated that 33 422 ha (6 percent of the total lake) was inhabited by fish. At 533 kg ha^-1 the total standing crop was 17 814 t. In Balon (1974) however he claims that fish inhabited the entire 0–20 m zone (37.7 percent of the lake area) which was 202 200 ha. At 533 kg ha^-1 the total standing crop will be 107 726 t although the figure of 18 000 t is retained (p.440).

It is thus unclear which are the correct figures. Since much of the shoreline is unsuitable for fish, and fish numbers decline with depth (Fig.7), it is unwise to extrapolate standing stock figures to the whole of the 0–20 m zone. Balon's estimate of 17 814 t may be realistic because it is based on the assumption that only 25 percent of the 0–20 m zone, is inhabited by fish.

From July 1981 to October 1982 biomass varied from 35–175 kg ha^-1 for ten stations in the Sanyati basin (Table 12). It was also highly variable, suggesting that fish were unevenly distributed. Peak biomass was reached in August 1981, followed by a sharp decline until December. This closely follows the pattern of the commercial catches and plankton abundance (Marshall et al., 1982). There was no peak in August 1982 which was probably caused by exceptionally poor river flows in the 1981–82 rainy season.

River flow has been shown to affect sardine catches, at least in the eastern Sanyati basin (Fig.17). This is supported by the comparison between the August 1981 biomass of 175 kg ha^-1 which followed exceptionally good river flows, and the August 1982 biomass of 35 kg ha^-1 which followed the worth river flow into Kariba since its creation (exact river flow data are not yet available).

Another factor which could influence the sardine biomass is the outflow: volume ratio of the lake. There is an indication that lakes with a long replacement time will be richer and support a higher biomass of sardines (Table 13). Thus Tanganyika supports a much higher biomass than Kariba, but although Kainji has a similar conductivity to Kariba its clupeid population is much lower. It is probably less suitable as it is shallow with a very short replacement time.

An attempt has also been made to relate commercial catches to biomass (Fig.18). At present the correlation is low, but it is hoped that it will improve with further sampling; this would simplify biomass estimations.

3.2.6.2 Population dynamics

Some knowledge of population dynamics is required if yield estimates are to be made with precision. The most important and useful is mortality (Z), made up of natural mortality (M) and fishing mortality (F).

Total mortality coefficient (Z) have been estimated for a number of inshore fish species (Table 14). Mortality rates were predictably high for small, short-live species (e.g., A. lateralis, M. acutidens, P. darlingi) and low for large, long-lived species (e.g., M. deliciosus, H. longifilis).

Mortality coefficients were also obtained by Langerman (unpubl.) for H. vittatus in the Mwenda estuary and the Sanyati basin (Table 15). Total mortality for fish from the Mwenda (an unexploited area) was 0.58, very close to Balon's estimate of 0.52. Total mortality for commercial samples from the Sanyati basin rose steadily from 0.65 in 1979 to 0.73 in 1981. If the Mwenda estimate is regarded as natural mortality (M) then fishing mortality (F) can be obtained by subtraction. Fishing pressure has increased from 1979 to 1981, but no accurate measure is available.

Similar preliminary results have been obtained for the sardine stocks, supported by detailed fishing effort data (Table 16). Limnothrissa is a small fish with a rapid life cycle and the monthly mortality rates are accordingly very high.

Some comparative mortality rates suggest that Kariba is not an optimal clupeid habitat (Table 17). The data may not be entirely valid as they compare Stolothrissa with Limnothrissa although both species are ecologically similar in their respective lakes. Mortality rates for unexploited sardines in Kariba are comparable to those exploited fish in Tanganyika which supports the biomass data in suggesting that Kariba is less suitable for clupeids. This may apply to other man-made lakes because of the hydrological features common to most.

3.2.6.3 Yield prediction

Early estimates of potential yield from Kariba were made in the absence of empirical data and ranged from 9 000 to 30 000 t yr^-1 (Hickling, 1956; Maar, 1959). At the time of its creation no-one could estimate with any certainty, what the fish production would be.

Balon's (1974) study was a major attempt to estimate production. Unfortunately, his terminology of “yield” did not correspond closely to that used by other workers (e.g., Beverton and Holt, 1957; Ricker, 1975; Gulland, 1969), however, his term “available yield” was similar to “surplus production” (p.411), and this should give an indication of potential fish catches.

An estimate of the total annual yield can then be made by multiplying these data by the total fish-inhabited area. As already discussed, these figures are inconsistent; using Balon's estimate of 33 422 ha yield ranges from 4 812 t to 5 448 t yr^-1. Balon's (1974) estimate of 202 200 ha of fish-inhabited areas would lead to yield estimates of 29 117 t to 32 958 yr^-1. He suggested that maximum sustained yield could be about 40 000 t (including non-commercial species).

Actual inshore catch from Kariba is about 3 000 t yr^-1 so Balon's 1973 data would appear to be the most accurate (the reasons whey the 1974 data may be overestimated have already been discussed).

Some other means of predicting yield can also be used. The morphoedaphic index (section 3.2.1) suggests that overall production would be about 23.3 kg ha^-1 or 12 445 t. This does not, however, differentiate between the inshore and pelagic fisheries.

Gulland's yield equation may be more useful. In this equation,

An attempt can be made to estimate yields for commercially important species using this equation with three biomass estimates (Table 19). The production data so obtained can probably be halved; the range from 30 to 57 kg ha^-1 is probably closer to the truth. Langerman's data were collected in an estuarine area (Mwenda) and may thus be atypically high.

This highlights the importance of a detailed study of the distribution and abundance of the inshore fish, such as carried out by Mitchell, S.A. (1976).

The potential yield from the clupeid fishery has not been estimated with any accuracy. Gulland's equation can be applied to the available data and very high potential yields are obtained (Table 20). Current catches are far below this and they are almost certainly overestimates. Most of the data on which this is based still require detailed analysis and a more realistic figure may be obtained in the future.

3.2.7 General biological data

Some general growth rates for major species may be useful for comparative purposes (Table 21). Perhaps the most striking difference in growth rates is shown by Limnothrissa which in Kariba is only about half the size reached in Tanganyika or Kivu (Fig.19). This is a reflection of hydrobiological conditions as Kariba is less rich (conductivity = 80 μS cm^-1) than the two natural lakes (Tanganyika, 500 μS cm^-1; Kivu, 1 200 μS cm^-1).

Very little is known about the age of first maturity for Kariba fish, although some information was given by Bowmaker (1973). This is summarized with some fecundity data in Table 22.

Fecundity of Limnothrissa was estimated by Begg (Lake Kariba Report 17) and a 70 mm fish could have up to 7 000 eggs. Fish matured at about 40 mm which is close to the mean size taken in commercial catches.

3.3 Fishery Development

3.3.1 Planning aspects

Kariba is unique among African man-made lakes in that it is shared almost equally between two countries, Zimbabwe and Zambia. The difficulties inherent in this were made worse by the fact that, until recently, these countries were virtually at war. This prevented the development of a lakewide management policy, and both countries have devised their own policies.

Research on Kariba began with the establishment of the Lake Kariba Fisheries Research Institute as a joint Rhodesia/Zambia/FAO project in 1963. It was taken over completely by the Rhodesian (now Zimbabwean) government in 1965 and has been operated by them since then. Research in Zimbabwe is also carried out by the University of Zimbabwe, although their Kariba station was forced to close because of military activity in 1976.

Work on the Zambian side was carried out by the Central Fisheries Research Organization which established a laboratory and field station at Sinazongwe. Increasing military activity forced this to close in about 1973, and no further work has been done since then. The Department of Fisheries in Zambia now has an officer with responsibility for Kariba, but based at Chilanga, near Lusaka.

Contacts between Zimbabwean and Zambian workers have now been re-established and it has been proposed that a formal bilateral liaison group be set up. Joint projects have been suggested, to be carried out under the auspices of the Southern African Development Coordination Conference (SADCC).

3.3.2 Development aspects

Proposals to develop the fishery were made before the lake was formed. The first step was to bush-clear certain areas (down to 20 m deep), in order to facilitate fishing operations. A total of 954 km² (about 18 percent) of the lake floor was cleared. It has been suggested that this was not entirely successful, and that a system of pattern-clearing would have been better (Bowmaker, 1973). On balance, however, this is probably not true, since easier navigation and longer net-life are important.

The Batonka tribes people were not primarily fishermen, and a variety of sociological factors worked to inhibit them (see Colson, 1971). Some enthusiasm arose in the early years when catches were very good, but this was lost as they declined. This highlights the difficulty of planning a fishery in the early years of a lake's existence.

On the Zimbabwean side fishing concessions were given to European-owned firms, but these areas have gradually been reduced. Only one firm still fishes actively; the others have been forced out because catches became so poor that their operations were unprofitable.

3.3.3 Management aspects

On the Zimbabwean side fishing areas have been allocated and some closed areas established (these are all National Parks areas). The numbers of fishermen are presently unregulated, but plans for this are being drawn up. Numbers will be based on the estimated yield, the catch per unit effort, and the minimum catch required for a fisherman to make his living. These have been collected over a long period (see 3.3.4).

The maximum size mesh for gillnets on the Zimbabwean side is 100 mm (4"). Very little is known about the minimum size of maturity and the reasons for setting this limit were intuitive rather empirical. An attempt was made to determine minimum harvestable size (Balon, 1974) but the rationale for this is not clear. In the example for Mormyrus longirostris (Fig.20) the size is given as 302 mm or 360 g, apparently determined solely from the intersect of the growth and increment curves. The danger of this approach is that it takes no account of the size of maturity. In M. longirostris the minimum size is about 295 mm (♀) and 410 mm (♂), with means being about 500 mm (♀) and 550 mm (♂) (Joubert, 1975). Fortuitously, the smallest M. longirostris taken in experimental 100 mm nets in 1982 was 387 mm long.

This suggests that whilst the minimum harvestable size may be close to reality it would be unwise to rely on this type of estimate without assessing breeding data. It also suggests that the 100 mm is reasonable (at least for M. longirostris) and should not be reduced.

Another fortunate case concerns the tigerfish Hydrocynus vittatus. Yield/recruit models suggest the optimum yield will be from 2–3 year old fish, which is the age they are taken in 100 mm nets (Langerman, unpubl.). This is again an instance of intuition rather than empiricism.

Efforts have been made to control the pelagic fishery through the issue of licences and it is unlikely that anymore will be issued on the Zimbabwean side. The policy in Zambia is not known and this is a clear case where joint planning is required.

3.3.4 Commercial catch statistics

Commercial catches have been monitored since the lake was built. Enumerators are sent to fishing villages on the Zimbabwean side where data are collected over a 7-day period. Monthly and annual totals are then estimated. Pelagic fishermen are required to submit a monthly return with daily catches of sardines and tigerfish.

Inshore commercial catches were high soon after the lake formed, but rapidly declined (Fig.21). Zimbabwean catches have levelled off at about 1 000 t yr^-1; Zambian catches are probably similar. The initially high catches pose a difficult management problem as fishermen were enthusiastic while catches were good, but lost interest as they declined.

Pelagic catches began in 1973 and rose rapidly to about 12 000 t in 1981. The catch in 1982 is projected to drop to about 8 000 t, almost certainly because of the very poor inflows in 1981/82. A recovery is expected if the 1982/83 river flows are normal but the catch data suggest that fish fishery is nearing its limit. No data are available from Zambian pelagic fishery; this is an important omission as the pelagic fishery must, clearly, be managed as a single unit.

Long-term catch statistics make it possible to use catch/effort models to estimate potential yield. An example from the Zimbabwean fishing area C.2 suggests that the average yield from an intensive fishery will be around 20 kg ha^-1 (Fig.22). If it is assumed that 30 percent of the lake is fishable water, then a total yield of about 3 200 t is possible. This will probably be reduced since extensive areas support low fish populations (see 3.2.4) so the present yield of perhaps 2 000 t (1 000 from each side) could be the most that the inshore fishery can produce.

Although these fisheries are producing to their limit the pressure of the resource is increasing because of the explosive population growth of 3–4 percent in the area. This is shown clearly in two fishing areas with different numbers of fishermen (Table 22). The two areas are ecologically similar with about the same potential yield but the final catch per fisherman is very much less in the more intensively fished area. This emphasizes the importance of finding alternative employment or occupation for these people.

3.3.5 Processing, marketing and distribution

The bulk of the inshore catch is processed at an artisanal level, with fish being sold fresh to traders, then smoked and transported inland by bus or bicycle. Two commercial organizations buy fresh fish for canning or freezing. This is a valuable service in remote areas as they also sell basic foodstuffs and nets, which are otherwise difficult to obtain.

Most sardines are sun-dried, packaged in bulk and sent to cities and towns. Some are canned or frozen.

3.4 General Impact of Lake Kariba

The greatest impact of Kariba has obviously been on the inhabitants of the are now flooded. About 57 000 people were displaced and the effects of this resettlement have been described, notably by Colson (1975). The lives of those people are generally thought to have been adversely affected through the loss of fertile land, overcrowding and the failure of the fishing industry. Movements of game and people led to an increase in Tsetse fly and a consequent reduction of cattle herds (Scudder, 1972). However, it has been suggested that, on the south bank, people were initially more prosperous as they were able to occupy areas recently freed from tsetse fly and so increase the size of their herds (Weinrich, 1977).

Throughout its history the Zambezi valley has been a recurrent famine area (Scudder, 1972). The agricultural potential is low and poor agricultural techniques have led to soil erosion and degradation. The human population has increased at a rate of about 3 percent yr^-1, leading to increased pressure on agricultural resources. On the Zimbabwean side, attempts are being made to utilize wildlife for the benefit of the local inhabitants (Taylor, 1982).

The lake has brought prosperity to Kariba town itself which is by far the largest settlement on the lake with over 12 000 people (1982 Zimbabwean census, unpubl.). Tourism is an important activity and there are six hotels, camping and caravan sites as well as boat marinas, tour operations and ancillary activities. The sardine fishery has had a major impact with a capital investment of over Z$5 million. About 1 500 people are employed in this fishery which is now probably the major employer in the town.

4. REFERENCES

Allison, G.F., 1969 The hydrological management of Lake Kariba. In Some problems of aquatic ecosystems in Southern Africa. Newsl.Limnol.Soc.South.Afr., 13 Suppl.

Attwell, R.I.G., 1970 Some effects of Lake Kariba on the ecology of a flood plain of the mid-Zambezi Valley of Rhodesia. Biol.Conserv., 2:189–96

Balinsky, B.I. and G.V. James, 1960 Explosive reproduction of organisms in the Kariba Lake. S.Afr.J.Sci., 56(4):101–4

Balon, E.K., 1971 Age and growth of Hydrocynus vittatus Castelnau, 1861, in Lake Kariba, Sinazongwe area. Fish.Res.Bull.Zambia, 5:89–118

Balon, E.K., 1971a Replacement of Alestes imberi Peters, 1852, by A. lateralis Boulenger, 1900 in Lake Kariba, with ecological notes. Fish.Res.Bull.Zambia, 5:119–62

Balon, E.K., 1973 Results of fish population size assessments in Lake Kariba coves (Zambia), a decade after their creation. Geophys.Monogr., 17:149–58

Balon, E.K., 1973a The eels of Siengwazi Falls (Kalomo River, Zambia) and their significance. J.Natl. Mus.Zambia, 2:65–82

Balon, E.K., 1974 Fish production of a tropical ecosystem. Monogr.Biol., 24:249–573

Balon, E.K., 1974a Fishes from the edge of the Victoria Falls: demise of a physical barrier for downstream migrations. Copeia, 1974(3):643–60

Balon, E.K., 1975 The eels of Lake Kariba: distribution, taxonomic status, age, growth and density. J.Fish.Biol., 7:797–815

Balon, E.K., 1978 Kariba, the dubious benefits of large dams. Ambio, 7(2):40–8

Balon, E.K. and A.G. Coche (eds), 1974 Lake Kariba: a man-made tropical ecosystem in Central Africa. Monogr.Biol., 24:767 p.

Bazigos, G.P., C. Grant and C. Williams, 1975 Revised fisheries time series of the natural waters of Zambia (1966–1974). Fisheries Training Development Project Statistical Studies. Rome, FAO, ZAM/73/009/St.S./3:66 p.

Begg, G.W., 1969 Observations on the water quality and nature of the affluent rivers of Lake Kariba, with reference to their biological significance. Newsl.Limnol.Soc.S.Afr., 13(Suppl.):26–33

Begg, G.W., 1970 Limnological observations on Lake Kariba during 1967, with emphasis on some special features. Limnol.Oceanogr., 15:776–88

Begg, G.W., 1974 The influence of thermal and oxygen stratification on the vertical distribution of zooplankton at the mouth of the Sanyati Gorge, Lake Kariba. Kariba Stud.Pap., (4):60–7

Begg, G.W., 1974a The distribution of fish of riverine origin in relation to the limnological characteristics of the five basins of Lake Kariba. Hydrobiologia, 44:277–86

Begg, G.W., 1976 The relationship between the diurnal movements of some of the zooplankton and the sardine Limnothrissa miodon in Lake Kariba. Limnol.Oceanogr., 21:529–39

Bell-Cross, G., 1972 The fish fauna of the Zambezi River system. Arnoldia, Rhodesia, 5(29):1–19

Bell-Cross, G., 1976 The fishes of Rhodesia. Salisbury, National Museums and Monuments of Rhodesia

Bell-Cross, G. and B. Bell-Cross, 1971 Introduction of Limnothrissa miodon and Limnocaridina tanganicae from Lake Tanganyika into Lake Kariba. Fish.Res.Bull.Zambia, 5:207–14

Bernacsek, G.M. and S. Lopes,(in press) Investigations into the fisheries and limnology of Cahora Bassa reservoir seven years after dam closure. FAO:GCP/MOZ/006/SWE (FI-1), Field Document No.

Beverton, R.J.H. and S.J. Holt, 1957 On the dynamics of exploited fish populations. Fish.Invest.Minist. Agric.Fish.Food G.B. (2 Sea Fish.), 19:533 p.

Blake, B.F., 1977 Food and feeding of the mormyrid fishes of Lake Kainji, Nigeria, with special reference to seasonal variations and interspecific differences. J.Fish Biol., 11:315–28

Bowmaker, A.P., 1973 An hydrobiological study of the Mwenda river and its mouth. Ph.D. Thesis, University of the Witwatersrand, Johannesburg

Bowmaker, A.P., 1973a Hydrophyte dynamics in Mwenda Bay, Lake Kariba. Kariba Stud.Pap., (3):42–59

Bowmaker, A.P., 1976 The physico-chemical limnology of the Mwenda river mouth, Lake Kariba. Arch. Hydrobiol., 77:66–108

Bowmaker, A.P., P.B.N. Jackson and R.A. Jubb, 1978 Freshwater fishes. Monogr.Biol., 31:1181–230

Caulton, M.S., 1970 A quantitative analysis of the mode of replacement of inorganic phosphates and nitrates in the waters of Lake Kariba during the period December 1969 – 1970. Newsl.Limnol.Soc.South.Afr., 15:52–61

Chapman, D.W. and P. van Well, 1978 Observations on the biology of Luciolates stappersi in Lake Tanganyika (Tanzania). Trans.Am.Fish.Soc., 107:567–73

Coche, A.G., 1969 Aspects of physical and chemical limnology of Lake Kariba, Africa: a general outline. In Man-made lakes; the Accra Symposium, edited by L.E. Obeng. Accra, Ghana Universities Press for Ghana Academy of Science, pp.116–22

Coche, A.G., 1971 Description of physico-chemical aspects of Lake Kariba, an impoundment in Zambia-Rhodesia. Fish.Res.Bull.Zambia, 5:200–67

Coche, A.G., 1971a Lake Kariba Basin: a multi-disciplinary bibliography, annotated and indexed, 1954–1968. Fish.Res.Bull.Zambia, 5:11–87

Coche, A.G., 1974 Limnological study of a tropical reservoir. In Lake Kariba, a man-made tropical ecosystem in Central Africa. Monogr.Biol., (24):1–247

Cochrane, K.L., 1976 Catches of Hydrocynus vittatus Castelnau during sardine fishing operations in Lake Kariba. Kariba Stud.Pap., (7):98–108

Cochrane, K.L., 1978 Seasonal fluctuations in the catches of Limnothrissa miodon (Boulenger) in Lake Kariba. Proj.Rep. Lake Kariba Fish.Res.Inst., (29):1–163

Coke, M., 1968 Depth distribution of fish on a bush-cleared area of Lake Kariba, Central Africa. Trans.Am.Fish.Soc., 97:460–5

Colson, E., 1960 The social organization of the Gwembe Tonga. Kariba Stud., 1

Colson, E., 1971 The social consequences of resettlement. Kariba Stud., 4

Coulter, G.W., 1977 Approaches to estimating fish biomass and potential yield in Lake Tanganyika. J.Fish Biol., 11:393–408

Davies, B.R., 1975 A. Hall and P.B.N. Jackson, Some ecological aspects of the Cabora Bassa dam. Biol.Conserv., 8:129–201

Donnelly, B.G., 1969 A preliminary survey of Tilapia nurseries on Lake Kariba during 1967–68. Hydrobiologia, 34:195–206

Gulland, J.A., 1969 Manual of methods for fish stock assessment. Part 1: Fish population analysis. FAO Man.Fish.Sci., (4):154 p. Issued also in French. Spanish edition published by Editorial Acribia, Zaragoza

Guy, P.R., 1980 River bank erosion in the mid-Zambezi valley, downstream of Lake Kariba. Biol. Conserv., 19:199–212

Hall, A., 1976 B.R. Davies and I. Valente, Cabora Bassa: some preliminary physico-chemical and zooplankton pre-impoundment survey results. Hydrobiologia, 50:17–25

Hall, A., I.M.C.B.S. Valente and B.R. Davies, 1977 The Zambezi river in the Mozambique: the physicochemical status of the middle and lower Zambezi prior to the closure of the Cabora Bassa Dam. Freshwat.Biol., 7:187–206

Harding, D., 1961 Limnological trends in Lake Kariba. Nature, Lond., 191:119–21

Harding, D., 1966 Lake Kariba. The hydrology and development of fisheries. Symp.Inst.Biol.Lond., 15:7–20

Henderson, H.F. and R.L. Welcomme, 1974 The relationship of yield to Morpho-Edaphic index and numbers of fishermen in African inland fisheries. Relation entre la production, l'indice Morpho-Edaphique et le nombre de pecheurs des pecheries des eaux continentales d'Afrique. CIFA Occas.Pap./Doc.Occas.CPCA, (1):19 p.

Herman, C., 1978 La peche dans le lac Tanganyika: Situation actuelle et perspectives de developpement. CIFA Tech.Pap./Doc.Tech.CPCA, (5):16–41

Hickling, C.F., 1956 Unpublished report on the potential of Lake Kariba. (mimeo)

Jackson, P.B.N., 1959 Summary of the Zambezi Valley surveys. Annu.Rep.Jr.Fish.Res.Organ., Lusaka, (8):24–8

Jackson, P.B.N., 1960 The ecological effects of flooding by the Kariba Dam upon Middle Zambezi fishes. In Proceedings of the First Federal Science Congress, Salisbury, Rhodesia, pp.277–84

Jackson, P.B.N., 1975 Common and scientific names of the fishes of Southern Africa. Part 2: Freshwater fishes. Spec.Publ.J.L^B. Smith Inst.Ichthyol., (14)

Jackson, P.B.N. and K.H. Rogers, 1976 Cabora Bassa fish populations before and during the first filling phase. Zool.Afr., 11:373–97

Joint Fisheries Research Organization, 1959 Annual report. Annu.Rep.Jr.Fish.Res.Organ., Lusaka, (8): 58 p.

Joubert, C.S.W., 1975 The food and feeding habits of Mormyrops deliciosus (Leach) 1818 and Mormyrus longirostris Peters 1852 (Pisces: Mormyridae) in Lake Kariba, Rhodesia. Kariba Stud.Pap., 5:68–85

Jubb, R.A., 1961 An illustrated guide to the freshwater fishes of the Zambezi River, Lake Kariba, Pungwe, Sabi, Lundi and Limpopo rivers. Bulawayo, Stuart Manning (PVT) Ltd., 171 p.

Jubb, R.A., 1967 Freshwater fishes of Southern Africa. Cape Town, Balkema

Jubb, R.A., 1976 Letter to the editor. J.Limnol.Soc.South.Afr., 2:35–6

Jubb, R.A., 1976a Unintentional introduction of fishes via hydroelectric power stations and centrifugal pumps. J.Limnol.Soc.South Afr., 2:29–30

Junor, F.J.R., 1981 Development of fisheries of the Zimbabwean waters of Lake Kariba. CIFA Tech.Pap./Doc.Tech.CPCA, (8):130–42

Junor, F.J.R. and G.W. Begg, 1971 A note on the successful introduction of Limnothrissa miodon (Boulenger) “The Lake Tanganyika sardine” to Lake Kariba. Newsl.Limnol.Soc.South. Afr., 16:8–14

Kenmuir, D.H.S., 1971 An analysis of data on the tigerfish Hydrocynus vittatus Castelnau obtained from Kariba International Tigerfish Tournaments. Newsl.Limnol.Soc.South.Afr., 17:13–21

Kenmuir, D.H.S., 1973 The ecology of the tigerfish Hydrocynus vittatus Castelnau in Lake Kariba. Occas. Pap.Natl.Mus.Rhod., B.5:115–70

Kenmuir, D.H.S., 1973a The commercial exploitation of stocks of tigerfish Hydrocynus vittatus Castelnau in Lake Kariba, Rhodesia by means of small-meshed gillnets. Rhod.J.Agric.Res., 11:171–9

Kenmuir, D.H.S., 1975 Sardines in Cabora Bassa. New Sci., 65:379–80

Kenmuir, D.H.S., 1975a The diet of fingerling tigerfish Hydrocynus vittatus Castelnau in Lake Kariba. Arnoldia,Rhod., 9:1–8

Kenmuir, D.H.S., 1980 The mussel resources of Lake Kariba. Trans.Zimbabwe Sci.Assoc., 60:7–10

Maar, A., 1959 The fish potential of Lake Kariba. In Proceedings of the First fisheries day in Southern Rhodesia. Salisbury, Government Printer, pp.50–1

Magadza, C.H.D., 1980 The distribution of zooplankton in the Sanyati Bay, Lake Kariba. Hydrobiologia, 70:57–67

Mahon, R. and E.K. Balon, 1977 Fish production in Lake Kariba, reconsidered. Environ.Biol.Fishes, 1(2):215–8

Marshall, B.E., 1977 On the status of Clarias gariepinus (Burchell) in large man-made lakes in Rhodesia. J.Limnol.Soc.South.Afr., 3:67–8

Marshall, B.E., 1979 Fish populations and the fisheries potential of Lake Kariba. S.A.J.Sci., 75:485–8

Marshall, B.E., 1980 The mussel resources of Lake Kariba. Trans.Zimbabwe Sci.Assoc., 60:7–10

Marshall, B.E., 1982 The influence of river flow on pelagic sardine catches in Lake Kariba. J.Fish Biol., 20:465–70

Marshall, B.E., 1982a Fish production potential of Lake Kariba. Zimbabwe Agric.J., 79:173–7

Marshall, B.E. and F.J.R. Junor, 1981 The decline of Salvinia molesta on Lake Kariba. Hydrobiologia, 83:477–84

Marshall, B.E., F.J.R. Junor and J.D. Langerman, 1982 Fisheries and fish production on the Zimbabwean side of Lake Kariba. Kariba Stud., (10):175–231

Matthes, H., 1968 Preliminary investigations into the biology of Lake Tanganyika Clupeidae. Fish.Res. Bull.Zambia, 4:39–45

McLachlan, A.J., 1969 Some effects of water level fluctuation on the benthic fauna of two Central African Lakes. Newsl.Limnol.Soc.South.Afr., 13:58–63

McLachlan, A.J., 1969a The effect of aquatic macrophytes on the variety and abundance of benthic fauna in a newly created lake in the tropics (Lake Kariba). Arch.Hydrobiol., 66:212–31

McLachlan, A.J., 1970 Submerged trees as a substrate for benthic fauna in the recently created Lake Kariba (Central Africa). J.Appl.Ecol., 7:253–66

McLachlan, A.J., 1970a Some effects of annual fluctuations in water level on the larval chironomid communities of Lake Kariba. J.Anim.Ecol., 39:79–90

McLachlan, A.J., 1974 Development of some lake ecosystems in tropical Africa, with special reference to the invertebrates. Biol.Rev.Cam.Philos.Soc., 49:365–97

McLachlan, A.J. and S.M. McLachlan, 1971 Benthic fauna and sediments in the newly created Lake Kariba (Central Africa). Ecology, 52:800–9

McLachlan, S.M., 1970 The influence of lake level fluctuation and the thermocline on water chemistry in two gradually shelving areas in Lake Kariba, Central Africa. Arch.Hydrobiol., 66:499–510

Mills, M.L., 1973 The “explosive” occurrence of the freshwater medusa, Limnocnida tanganyicae Gunther, in Lake Kariba during 1972. Arnoldia, Rhod., 6(14):1–8

Minshull, J.L., 1973 The establishment of a fishery among the resettled Batonka people at Lake Kariba, Rhodesia. In Eleventh Congress, International Commission for Large Dams. pp.413–9

Mitchell, D.S., 1969 The ecology of vascular hydrophytes on Lake Kariba. Hydrobiologia, 34:448–64

Mitchell, D.S., 1970 Antecological studies of Salvinia auriculata Aubl.. Ph.D. Thesis, University of London.

Mitchell, D.S., 1973 Supply of plant nutrient chemicals in Lake Kariba. Geophys.Monogr., 17:165–9

Mitchell, D.S. and B.E. Marshall, 1974 Hydrobiological observations on three Rhodesian reservoirs. Freshwat.Biol., 4:61–72

Mitchell, D.S. and D.J.W. Rose, 1979 Factors affecting fluctuations in extent of Salvinia molesta on Lake Kariba. PANS, 25:171–7

Mitchell, S.A., 1975 The biology of Central African chaoborids and their distribution over selected areas of Lake Kariba. M.Phil.Thesis, University of London

Mitchell, S.A., 1978 The diurnal activity patterns and depth zonation of marginal fish as shown by gillnetting in Lake Kariba. Kariba Stud.Pap., (9):163–73

Mitchell, S.A., 1976 The marginal fish fauna of Lake Kariba. Kariba Stud., (8):109–62

Otobo, F.O., 1979 The population of clupeids in Lake Kainji, Nigeria. Arch.Hydrobiol., 86:152–60

Ricker, W.E., 1975 Computation and interpretation of biological statistics of fish populations. Bull.Fish. Res.Board Can., (191):382 p.

Robarts, R.D. and G.C. Southall, 1977 Nutrient limitation of phytoplankton growth in seven tropical man-made lakes, with special reference to Lake McIlwaine, Rhodesia. Arch.Hydrobiol., 79:1–35

Roest, F.C., 1978 Stolothrissa tanganicae: population dynamics, biomass evolution and life history in the Burundi waters of Lake Tanganyika. CIFA Tech.Pap./Doc.Tech.CPCA, 5:42–62

Scudder, T., 1962 The ecology of the Gwembe Tonga. Kariba Stud., 2

Scudder, T., 1972 Ecological bottlenecks and the development of the Kariba Lake basin. In The careless technology, ecology and international development, edited by M.T. Farvar and J.P. Milton. New York, Natural History Press, pp.206–35

Scudder, T., 1975 Resettlement. In Man-made lakes and human health, edited by N.F. Stanley and M.P. Alpers. London, Academic Press, pp.206–35

Scudder, T., 1980 River-basin development and local initiative in African savanna environments. In Human ecology in savanna environments, edited by D.R. Harris. London, Academic Press, pp.383–405

Skehel, J.P., 1969 The African fishermen on Lake Kariba. Newsl.Limnol.Soc.South.Afr., 13:45–52

Skelton, P.H., R.A. Jubb and M.N. Bruton, 1980 Additions to the checklist and recent changes to the scientific names of southern African freshwater fishes. J.Limnol.Soc.South.Afr., 6:109–12

Taylor, R.D., 1982 Buffer zones; resolving the conflict between human and wildlife interests in the Sebungwe region. Zimbabwe Agric.J., 79:179–84

Thomas, P.A., 1974 Investigations into the biology of Paulinia acuminata (de Geer) (Orthoptera: Acrididae) with particular reference to the biological control of Salvinia auriculata. M.Phil.Thesis, University of London

Thornton, J.A., 1980 A review of limnology in Zimbabwe, 1959–1979. Salisbury, Ministry of Water Development.

Thornton, J.A., 1980a A comparison of the summer phosphorus loadings to three Zimbabwean water supply reservoirs of varying trophic states. Water S.Afr., 6:163–70

Toots, H. and A.P. Bowmaker, 1976 Serranochromis robustus jallae (Boulenger, 1896) (Pisces; Cichlidae) in a Rhodesian highveld dam. Arnoldia, Rhod., 7(39):1–16

Trewavas, E., 1981 Nomenclature of the tilapias of Southern Africa. J.Limnol.Soc.South.Afr., 7:42

Viner, A.B., 1970 Hydrology of Lake Volta, Ghana. 1. Stratification and circulation of water. Hydrobiologia, 35:209–29

Ward, P.R.B., 1978 Water surface fluctuations at Lake Kariba. Rhod.Eng., 16:133–42

Ward, P.R.B., 1979 Seiches, tides, and wind set-up on Lake Kariba. Limnol.Oceanogr., 24:151–7

Webster, M.H., 1975 Medical aspects of the Kariba hydroelectric scheme. In Man-made lakes and human health, edited by N.F. Stanley and M.P. Alpers. London, Academic Press, pp.69–88

Welcomme, R.L., 1972 The inland waters of Africa. CIFA Tech.Pap. (1):117 p. Issued also in French.

Weinrich, A.K.H., 1977 The Tonga people of the southern shore of Lake Kariba. Salisbury, Mambo Press.