Like many other freshwater systems, floodplains fulfil a range of functions other than that of supporting fish and fisheries. Most of these are associated with the dry phase (see Section 3) and do not greatly interfere with the fisheries productivity of the flood phase. Certain activities may indeed even be beneficial as in the case of dung dropped by grazing animals.
Major conflicts with fisheries occur in management schemes involving either the alteration of the hydrological cycle or the contamination of the environment.
Dams installed upstream of the floodplain: To date, dams installed upstream of the floodplain system are the main cause of damage to their fisheries. The Lower Volta floodplain has disappeared completely following the construction of the Akosombo dam (Hall and Pople, 1968). Considerable areas of the Niger River floodplain have been lost below the Kainji dam where the flood regime has been altered as far downstream as the Benue confluence and fish catches have diminished to between 39 and 75 percent of their original values (FAO, 1972). The closure of the J.F. Strijdom Dam resulted in the complete dessication of many of the floodplain pools of the Pongola River (Phelines, Coke and Nicol, 1973) although here controlled release of flood water refilled the pools and induced breeding in the fish in 1969 and plans are going forward for the fishery to be maintained by this means.
In addition to the general loss of production, changes in the composition of the fish fauna have occurred downstream of the Kainji dam (Lelek and El-Zarka, 1971 and 1973). Here a decline of the swamp dwelling and herbivorous fishes was noted and the predators increased in abundance. Such a disproportion of predators obviously cannot persist for any length of time and will eventually act to reduce the abundance of the stock as a whole.
The effects of two projected upstream dams have also been assessed for Senegal and Zambia. In Senegal, the closure of dams in the upper course of the river would result in the loss of some 7 000 t of floodplain fish production although nearly twice this might be gained from the dams themselves. The creation of the Itezhi-Tezhi Dam on the Kafue might have a beneficial effect on fish stocks by the regulation of water levels to create improved floods and longer retention time of the Kafue Gorge dam downstream (Muncy, 1973). Over a long term upstream dams are likely to effect the total productivity of the system by slowing down or preventing the deposition of silt on the plain.
Dams installed upstream of the plain may not necessarily disturb the hydrological regime but may interfere with the normal migratory pattern of the fish preventing their dispersal fully throughout the river. This has occurred with the Markala dam on the Niger (Daget, 1960a) which has interrupted passage of many species of fish from the Central Delta to the reaches upstream of the dam. From the limited experience available it would appear that the installation of fish ladders to bypass such dams is impractical in Africa in view of the numbers of fish involved and the differences in specific behaviour.
Dams installed downstream of the floodplain: These generally result in a loss of the floodplain by its complete submergence. The transformation of the area to permanent lake may generally produce similar or increased fish yields but it will do so at the sacrifice of alternative dry season forms of production. Dams which are designed to retain water in the floodplain for a longer period will theoretically improve yields by prolonging the flood phase (University of Idaho, 1971; University of Michigan, 1971) although there is, at present, no evidence of this happening in the case of the Kafue Gorge dam (Muncy, 1973). In fact Dudley (1974) proposes that prolonged flooding may cause a decrease in fishing effort by lessening the accessibility of the fish to highly efficient gears such as seine nets.
Irrigation: Little is known of the effects of irrigation projects on the fisheries component of floodplains, although it is to be anticipated that the shortening of the flood cycle, due to improved drainage, would affect the population of fish adversely. There is, furthermore, evidence that channelization of streams considerably reduces both ichthyomass and species diversity (Congdon, 1973). On the positive side, a larger dry season area of water in irrigation canals and reservoirs could improve survival and low water fish yields. Many of the Ouémé floodplain whedos were originally drainage canals.
Pollution: There is at present no evidence that pollution is a serious problem in Africa although the use of fertilizers and insecticides, associated with intensive agriculture, could conceivably be a source of trouble in future. Possible difficulties are likely to arise where migrating fish are prevented from passing through polluted reaches of river. Excessive nutrients might also cause overproduction of swamp vegetation, which, by its later decay and the resulting de-oxygenation, would cause extensive mortalities in floodplain lagoons and pools.
Changes occuring in the whole floodplain throughout one annual cycle are summarized in Fig.18, which shows in cyclic form the nutrient/energy flow within the system. It is known that in stable floodplains the rate of alluvial deposition equals the removal of alluvial matter by erosion over a period of years (Leopold, Wolman and Miller, 1964). It may be assumed that, in the wild state, a similar dynamic equilibrium occurs with respect to the nutrient and biomass. If such an equilibrium does exist, then the productivity of one component can be affected by the bad management or overexploitation of one of the other components, and by other interventions which change the essential characteristics of the system. An example of this type of interaction is cited by Kapetsky (1974) in his examination of the evidence for a diminution in fish population in the Kafue Flats, resulting from the reduction in number of the primary users of floodplain vegetation i.e., Lechwe and hippopotami.
As a general principle floodplains should be managed for the full realization of the potential for both dry and wet season crops of which fisheries are but one component. Thus elements A, B, C and D in Fig. 18 should be so adjusted as to maximize their sum.
Figure 18 Diagram of nutrient and energy cycles on the floodplain
This operation in itself may vary depending on the management objectives of any particular country. The sum can be judged in several ways; for example, output by weight, financial benefit, nutritional value or energy use. In each of these the equation will change according to the different weightings placed upon the factors. In some cases, maximization of an individual component can only be pursued to the detriment of another. For maximum production of rice, the use of insecticides against stem borers, herbicides to clear irrigation channels and dykes to protect the rice from the ravages of rizophagous fish are necessary. Similarly, maximization of dry season crops could require rapid drainage of the floodplain to increase the length of the dry phase. Both these would result in a qualitative and quantitative diminution of the fish population leading to a fall in fish catch. Frequently one factor of the floodplain economy is judged of such importance that all other uses of the plain are suppressed and the environment is modified irreversibly in its favour. Unfortunately the factors leading to these decisions may change within the span of a few years to reverse the relative values of the original components without there being any possibility of regaining them.
One of the most thoroughly utilized of floodplains at a relatively low level of technology is the Ouémé valley (Fig.19) where, during the dry season artificial fish ponds alternate with maize fields, and vegetables are grown on the banks of the ponds. Cattle are grazed on the raised levées which also support stilt villages with an aggregate population of 70 000 inhabitants (equivalent to 70/km2). During the flood season, activities have until recently been confined to fishery although floating rice culture is at present spreading. Such intensive use depends on a complete sociological adaptation to the environment and is not advocated for transplantation elsewhere. The example of this valley does, however, indicate ways in which several uses of the floodplain can be developed harmoniously.
Figure 19 Portion of the Ouémé floodplain showing the distribution of different activities
A. Air photograph (by courtesy of I.G.N., Paris)
B. Interpretation for explanation see text
Natural waters: As detailed in section 6 the fishery usually concentrates on three points of the annual cycle and any policy for the rational management of fish stocks in floodplain systems should aim at defining and controlling these. At present there is little definite information on the effects of fishing practices on the floodplain but the following observations have been made.
Migration at beginning of flood: Fisheries exploiting migrating fish before reproduction generally seem to be harmful to the fish stock. Additional pressure is laid on the population at its weakest period and numbers of breeding fish can be seriously reduced. At least two important fisheries, that for Labeo victorianus in the Nzoia River (Cadwalladr, 1965) and that for Labeo altivelis in the Luapula (Soulsby, 1959) have collapsed due to overexploitation at this time. It therefore seems advisable to discourage this practice under normal circumstances.
Return migration of juvenile fish: Fisheries aimed at the capture of migrating fish as they leave the floodplain are widespread and much difference of opinion has arisen on the harmfulness of this practice. It has been widely supposed that most of the juveniles returning to the river are destined to die due to the restricted environment available to them in the dry season. This may well be true in the Niger River or the Barotse floodplain where neither Reed (FAO, 1969) nor Bell-Cross (1971) considered that the practice was particularly harmful.
Dry season fisheries in permanent water bodies: It is a general practice for most of the dry season pools of the floodplain to be fished exhaustively. As many of the pools dessicate toward the end of the dry season, and others become de-oxygenated, much of the population in such waters will in any case not survive to the next flood. Pools are usually restocked by immigration from the river and by drain-in during the floods and this fishery practice does not appear to be harmful.
The intensive fishing of the main river channel, however, may result in a general lowering of the breeding stock. In areas where this practice is common many of the larger species have disappeared from the fish community and signs of severe overfishing are apparent (e.g., the Ouémé River and the Senegal River in the recent Sahelian drought). To avoid this situation consideration should be given to the creation of protected ‘reserve’ areas, either of minor river channel or the larger permanent lake and lagoons which may be set aside for the maintenance of breeding stocks. The effectiveness of such areas is perhaps testified by the rapid reappearance of large fish in the Central Delta of the Niger after the Sahelian drought was broken. For two years only diminishing quantities of small fish had been caught but large inaccessible lake areas had remained relatively unfished and it is possible that these had served as refuges for the larger individuals of many species.
Obviously considerable work on the dynamics of natural and exploited stocks on floodplains needs to be done before any definite judgment on this or on other management measures can be made.
Artificial waters: While intensive aquaculture is probably impractical at the present stage of floodplain management, a number of extensive practices may be of value to extend present production, or to maintain some production in areas managed for other purposes.
The technique of digging drain-in pools is already well established in the Ouémé valley (Section 6) and blind ponds of this type seem well adapted for permanently swampy areas where the water table is at or just below the surface. By digging whedo like trenches on ponds reclaimed earth bounded on either side becomes dry enough for vegetable growing and the ponds themselves retain water at all times of year. Further adaptations of the ‘whedo’ method have been used to regularize and extend naturally occurring water bodies of the floodplain.
The retention of water within floodplain pools by damming is also under investigation in several regions of Africa at the present time; the results of two such experiments being shown in Figs. 7, 9 and 10. While some time will have to elapse before results become known, it would seem that this methodology, which is useful for increasing the production of fish from floodplain systems where the natural cycle is preserved, will be particularly valuable for maintaining fish production in areas where floods will be curtailed or altered by upstream flood control dams.
Ponds might also be envisaged in irrigation schemes, where canals can be enlarged at intervals for the rearing of fish or the placing of fish culture cages. Similarly, cage culture might be developed in the main channel of the river where at present ‘akadja’ fish park methods are used. Fish park like installations are currently used in several West African rivers and may be seen as dotted areas in the main river channel in Fig.19.
The capture and culture of fish in rice fields is widely practised in many parts of the world (Coche, 1967) including some African countries. Capture methods rely on the fish which introduce themselves naturally into the rice fields whereas culture (rizipisciculture) implies a deliberate policy of stocking selected species. In general increased yields of both fish and rice are obtained by their joint culture. Some species of fish are however notorious for their rizophagous habits and culture fisheries attempt to exclude them from the waters. On floodplains where rice culture is being developed two main approaches to the problem of attacks on the rice seedlings by fish have been adopted. Firstly fields are surrounded with extensive dykes to control the water regime and to exclude fish. Obviously in these planned stockings with selected species will result on a double harvest although much experimentation needs to be done to define the correct species, stocking rates, etc. That this is a practical and economic proposition in Africa has been shown by several authors including Vincke (unpublished manuscript). A second solution, adopted particularly in the local floating rice culture in the Central Delta of the Niger is to construct low dykes which suffice to delay the floods until the rice seedlings are sufficiently robust to withstand attacks by fish. Here capture methods for immigrating individuals are relied upon for the fish portion of the harvest. It is evident from present experience that the growing of rice and the continued exploitation of the floodplain for fish are far from being mutually exclusive. In fact there is considerable evidence to indicate that given correct techniques the two activities can be practised together to the benefit of both.
The close correlation of flood regime with fish catch shows that the fishery in a river downstream of a flood control dam must be affected by it. Controlled release of water may simulate normal flood conditions and ensure the maintenance of a fish stock. It is certain however that the way in which the water is released both in terms of absolute amount, and the shape of the artificially produced flood curve, will influence the productivity of the system. Controlled release would seem easier in systems with large permanent lakes and pools than in systems with large dry plains. In the former the shape of the flood curve, and the flood regime is possibly relatively unimportant once the pools are filled. It has been demonstrated that fish can be induced to migrate into such areas where they breed and develop satisfactorily on comparatively limited floods (Phelines, Coke and Nicol, 1973). In systems with little residual water the duration and intensity and rate of change of the flood would appear to be the only factors controlling the reproduction, growth and survival of the fish. It is for this reason, among others, that the creation of artificial impoundments on the plains is seen as a means of preserving fish production in floodplains with controlled flow.
