Charles R. Goldman
CHARLES R. GOLDMAN is Professor of Limnology, Division of Environmental Studies, University of California at Davis, California.
A considerable amount of research has been done on the effects of water impoundment in the tropics. However, despite such research a general lack of knowledge about hydroelectric projects and aquatic biology in the tropics persists.
Studies, by the author and Ecological Research Associates, of the El Cajon Dam site on the Sula river in Honduras (Goldman, 1972a) and the Purari River Project in Papua New Guinea (Goldman et al., 1975) are used throughout this article to illustrate many of the principles that should concern ecologists, engineers and governments alike.
Experience with siltation in previous impoundments indicates that it must be considered of major importance to both the longevity and ecology of the reservoir. A detailed study of reservoir longevity as influenced by siltation should, therefore, be a serious concern of any proposed dam construction.
The degree of erosion control in the drainage area of a water-storage project can greatly alter the sedimentation rate. In addition to increasing runoff, the lack of a dense canopy in the watershed surrounding the proposed reservoir or inflowing rivers heightens the impact of rain on soil, allowing soil particle aggregates to be broken down and surface runoff to transport finer particles. The concentration of soil in runoff water is thus increased.
Since certain logging and agricultural practices in tropical countries have greatly increased surface runoff, it is recommended that the watershed be managed with particular attention to soil conservation and forest regeneration. Revegetation and the cessation of human activities which foster erosion can reduce the rate at which the reservoir capacity is decreased by siltation and the reservoir waters are fertilized by materials washed into the system.
Heavy tropical rains, steep valley slopes, and the state of the watershed soils and vegetation all contribute to the rate of siltation. Extreme water turbidity often appears in the tropics during the rainy season.
The potential for slumping of the slopes surrounding a reservoir may also be an important factor as evidenced by the disastrous flood at Lungarone in northern Italy some years ago resulting from a massive landslide into a reservoir. Mass wasting following lumbering was responsible for the death of thousands during a hurricane several years ago in Honduras. But a previous hurricane, although of a greater force, was not as destructive because the root structure of the forest remained intact. Subsequent decay set the stage for the disaster of the next hurricane.
While a reservoir is filling and after filling is completed, it is possible to establish sampling procedures that can provide fairly accurate siltation estimates. In order to obtain these estimates major inflows near their point of entry into the future reservoir must be sampled for settleable materials during the entire year.
Determination of reservoir lifetime is essential in evaluating the true economic potential of the project and its multi-use aspects. A detailed sampling programme, therefore, must be initiated prior to dam construction in order to realistically predict the expected life of the impoundment.
Large-scale siltation can have pronounced effects on the reservoir biota and it, distribution. Turbidity caused by suspended silt can significantly reduce photosynthesis in rivers and reservoirs. This photosynthetic reduction has widespread effects on higher trophic levels of the food chain.
Scanning electron microscopy (SE M) dramatically illustrates the importance of organic and inorganic particles as substrata for aquatic bacteria (Paerl and Goldman, 1972a). The surface area present on particulate matter provides a site for the attachment of bacteria as well as organics and inorganics which adsorb to these surfaces. Thus, by providing surface area for attachment as well as nutrition, silts are able to stimulate aquatic bacterial growth (Paerl and Goldman, 1972b). In addition to providing food for filter feeders, there is good evidence that bacteria and adsorbed organic material may serve as energy for mud-feeding fish.
Although drawdown with exposure of shallow breeding and food-producing areas is probably the most serious threat of water impoundment to fish, sedimentation can also destroy potential spawning beds in the shallow littoral zone of the reservoir. Other aspects of silt deposition include: (a) entrapment of these sediments behind the dam, with resulting loss of nutrient sources to downstream agriculture and marine fisheries, (b) changes in coastal beach erosion rates, and (c) the reduction in deltas which formerly had been replenished by sediment recharge during annual floods. Reservoirs invariably affect the pattern of downstream sediment deposition, and this aspect of any water impoundment in the tropics should receive special consideration in the planning stages.
Water quality following the closing of a dam typically undergoes a transitional period of high organic and inorganic ion concentrations and low oxygen content. The length of the transitional period depends not only on river water quality but also on the presence of any residual submerged terrestrial vegetation, the thermal profile of the reservoir, nutrient concentrations of submerged soil, and the rate at which reservoir water is being replaced by inflow water. Following this transitional period is' a state characterized by lower concentrations of minerals and organic materials, although nutrient concentrations remain above levels found in the rivers. This lowering of reservoir fertility may prove very disappointing to a fishery that has been developing during the initial period of high fertility.
Decay of submerged terrestrial vegetation will deoxygenate the reservoir and release organic compounds into the waters, both of which lower the oxidation-reduction (redox) potential of the water. This lower potential and leek of oxygen favour certain biological reactions, fin particular the reduction of sulphate to sulphide by the bacteria Desulfovibrio (Hutchinson, 1957). Combined with microbes that release sulphur from proteins as sulphide, these bacteria can be responsible for increasing the hydrogen sulphide concentration of the waters to a level toxic to fish and unpleasant for human use. The low solubility products of sulphide and heavy metal ions can result in large-scale precipitation of metallic sulphides that may interfere with turbine operation. Working turbine generator parts that contact reservoir water may corrode and not function properly if made of materials that can form metallic sulphides (e.g., copper). Substitutes for these ma. serials should be investigated.
With proper watershed control of erosion and sanitation practices, noxious algal blooms are not as likely to be a problem except during the: transitional period following dam closure. Eroded materials can serve as a direct source of nutrients if they originate from soils with leachable nitrogen, phosphorus, sulphur and trace elements. However, even minerals resistant to decomposition can promote the process of eutrophication by providing a substrate on which organic materials can be adsorbed and concentrated. Bacteria colonize these particles and recycle the nutrients for algae by decomposing adsorbed organic matter.
Reservoirs have a high water inflow relative to volume so that reservoir productivity is determined to a great extent by the quality of incoming rivers. Special attention should therefore be given to the water quality of inflowing rivers. A continuous water quality monitoring programme should be established with any impoundment project to monitor important ions in rivers flowing to the proposed reservoir. Bioassays performed to determine macro-and micronutrients limiting primary productivity would also be informative (Goldman, 1972b) and helpful for designing appropriate watershed management programmes.
The existing productivity of river water can be determined in a number of ways. One method is to measure its primary productivity directly. Another method is to add river water to algal culture, and measure the resulting changes in productivity. This latter technique is a bioassay of river water and is best performed on natural plankton communities found in the area.
Bioassays to determine the most likely limiting factor should also be performed using water from the inflowing river to the proposed reservoir with varying levels and combinations of N, P, Si, Fe and trace elements (Goldman, 1963 and 1967). The most likely limiting factor can then be utilized in developing strategies for lake management.
Stratification in tropical lakes is still an inadequately understood phenomenon but one, of very great importance. Deep Lakes stratify during the summer with a warmer surface layer (the epilimnion) at the top and a cooler layer (hypolimnion) below. Decaying organic matter in deeper waters consumes oxygen so that the hypolimnion is often depleted. Altitude, wind exposure, basin shape, and the temperature and turbidity of inflowing waters are all important variables in determining stratification.
Reservoirs in the tropics with steep sides and protection from winds are most likely to be permanently stratified. Mixing of waters may occur in the tropics, but often at such a slow rate that de oxygenation of the hypolimnion persists (Beadle, 1966). Cooling of shallow littoral water has been proposed as a slow mixing method in tropical lakes by Beadle (1966) and was shown by Talling (1963) to occur in Lake Albert.
Modelling of the thermal characteristics of reservoirs has been accomplished in temperate latitudes, and there are excellent prospects for achieving relatively accurate predictions in tropical areas.
The placement of dam outlets in relation to stratified layers of the reservoir has important implications for downstream fisheries and agriculture. Water from the hypolimnion may be much cooler than downstream waters and may shock the fishery below. Warmer epilimnetic waters, on the other hand, may be too warm for some species. It is essential that these potential effects be evaluated so that reservoir operation can be adjusted appropriately (e.g., variable level discharges).
One of the many lessons emerging from experiences with dams in Africa is the seriousness of the threat of infestation with aquatic macrophytes. The water fern (Salvinia auriculata) and water lettuce (Pistia stratiotes) invaded Lake Kariba on the Zambesi shortly after it began to fill in December 1958 (Harding, 1966). Eventually the water fern severely hampered multipurpose use of the lake and a buoy-barrier had to be constructed to keep higher aquatic plants away from turbine inlets (Coche, 1975).
Occasionally, for reasons unknown, a lake or reservoir may remain completely free from macrophytic infestation even through a source of infection is present and the water characteristics appear to resemble other areas which do suffer. For example, water hyacinth has not colonized certain areas apparently suited for it (Little, 1966).
Important characteristics of potential aquatic weed problems include: (a) the presence of aquatic weeds in the watershed and within naturally transportable distances, (b) the presence of protected swamps or littoral areas suitable for colonization, and (c) the presence of submerged terrestrial vegetation to serve as protection for weed colonies.
Large stands of higher aquatics influence many aspects of reservoir use. One of their benefits to man is to provide additional food for desirable herbivorous fish or for fish feeding on organisms which are favoured by the protection and food offered by the higher aquatic plant habitat. Dense growth of these macrophytes may also serve to protect fish, particularly smaller fish, from excessive predation.
The negative aspects of the presence of aquatic macrophytes, however, overwhelmingly point to the need to control their growth:
1. Decomposition of the macrophytes by aerobic bacteria can contribute to an extensive deoxygenation of the reservoir waters, the chemical effects of which have been mentioned. In addition, the lower deoxygenated waters will not be able to support fish populations, and plant productivity in this region will thus go unharvested.2. Huge thick floating mats (sudds), sometimes sufficiently large to support growth of other higher plants (e.g., Scirpus cubensis on Salvinia) can present real hazards to navigation and plug turbine intakes.
3. Heavy infestations can interfere with operation of gill-nets and seines and seriously hamper commercial as well as sport fishing.
4. Certain disease-bearing organisms (e.g., the snail vector of schisto-somiasis and the mosquito vector of malaria) find desirable habitats among these macrophytic "forests."
5. The volume of these plants may displace significant amounts of water and thus reduce reservoir capacity during times of overflow.
6. The macrophytes transpire tremendous amounts of water, perhaps five times the amount evaporating from a water surface of equivalent area. Thus, the 20 percent cover of water fern found on Lake Kariba in 1962 probably doubled the evapotranspiration rate!
A detailed study should be made of the occurrence and relative abundance of the animal and plant life in the area to be inundated. Since dam construction results in the loss of the riparian and watershed habitats to be inundated, these areas should be studied both with respect to the losses in vegetation that would occur and the effects this submerged and decaying vegetation would have on the proposed reservoir. The submergence of vegetation is a particular problem in the tropics since tropical forests hold more nutrients relative to the soil than temperate areas (Freeman, 1974). Large amounts of nutrients will be released to the lake and deoxygenation will occur in the hypolimnion through decay of these rotting plant materials. The Siranumum Dam near Port Moresby, Papua New Guinea, has a hydrogen-sulphide-rich outflow because terrestrial vegetation was not removed and decay resulted in increased levels of hydrogen sulphide in the hypolimnion. Hydrogen sulphide is very corrosive to metal and can destroy turbine parts and electrical connections.
Rotting vegetation is best controlled by removal and it is often recommended that the entire area to be inundated should be clear-cut and the vegetation removed or burned. Burning terrestrial vegetation is less desirable, however, since it will permit nutrients released by the destroyed terrestrial vegetation to be quickly utilized by aquatic plants (Ewer, 1966). However, the cost of physically removing terrestrial vegetation from the watershed is often prohibitive.
The control of higher aquatic plants has been accomplished using mechanical harvesting, although the quipment, expense, and the steep sides of a reservoir make this an impractical solution. Chemical control, especially with 2,4-D which is harmless to fish, has been successfully utilized, but toxicity to downstream agriculture suggests that its use be as limited as possible. In addition, by returning the plant nutrients to the reservoir waters, chemical applications merely prolong the problem.
Biological control of higher aquatic plants is the most desirable method, but experiments with manatees, snails, insects and herbivorous fish have yet to show great promise.
Infestation must be prevented in its earliest stages, control methods being expensive or ineffective once exponential growth is under way. Prevention consists of both limiting infection sources and controlling the chemical environment of these plants. In Ghana, at the Volta Dam, it was realized that close guard was required to prevent infestation. Posters were printed in eight languages displaying papyrus, Eichornia and Salvinia, asking that persons seeing these plants report them immediately to the authorities. It was recommended (Goldman, 1 972a) that lake patrols be established if El Cajon Dam were to be constructed in Honduras, for the purpose of regularly surveying the reservoir shoreline and removing any observed macrophyte growth.
Although it would be disastrous to encourage high yields of aquatic macrophytes in the reservoir, it may be noted that the Chinese have found water hyacinth and water lettuce to be an excellent duck and pig fodder.
Analysis of all wildlife values, including habitats, ranges, migrations, and the presence of endemic or rare species that might be affected by the proposed dam, should be considered. It was found at the El Cajon site that the original biotic communities of the proposed catchment area had already been severely damaged by the activities of man, including swidden agriculture, poor forestry practices and burning. It was concluded that the biological interest of the area could be enhanced if swidden agriculture and burning were abandoned and a renewed stand of timber was encouraged. It was recommended in this study that immediate steps be taken to prevent the lumbering or clear-cutting of the few remaining stands of virgin timber in the watershed of the proposed reservoir. Not only would conservation of large. tracts of broad-leaf and coniferous forests help to maintain the relatively low silt levels observed in the Rio Yure, but the forests could also serve as source areas for faunal and floral repopulation of the watershed.
One of the trade-offs in converting a large area of terrestrial habitat to an aquatic one is the potential development of viable fisheries. This can be of great importance in areas where the average protein intake is low and where agriculture is at a marginal level. There is, however, a tendency to overestimate the production and overfish the stock of reservoirs. Fish production rose in Lake Kariba along with the total dissolved solids (TDS) during the filling stage only to decline drastically in 1963 (Baron and Coche, 1974). The decline coincided with a drop in TDS from 81 to 63 ppm.
The best means of achieving a high protein-yield from an inland fishery is to promote the growth of species that feed on plant rather than animal food. There is a great loss in available food energy at each transfer of energy from herbivore through successively larger carnivorous fish.
The importance and complexity of fisheries development merits the creation of some permanent body to research the various problems and develop stringent laws to protect the new fishery. It is always easier to relax old laws rather than institute new ones which conflict with people's newly acquired habits.
Consideration of the potential ecological effects of dam construction should not be limited to the catchment area but should be extended to downstream effects as well. The decreased silt in waters downstream from the impoundment has significant effects on the nutrient load, and hence productivity, of these waters.
Nutrients contained in or bound to river-borne particulate matter have a fertilizing effect on the land and on ocean waters near rivers. This may contribute significantly to the productivity of existing coastal fisheries. The removal of this silt can decrease phytoplankton photosynthesis which in turn decreases fish productivity. An example of this probably occurred in the eastern Mediterranean, where sardine catches decreased in volume following the closure of the Aswan High Dam.
A similar problem exists with the effects of reduced silt load on the fertilization of river valley agricultural lands. Flood plains are normally very rich areas of vegetative growth and agriculture, but the richness can slowly be lost by the lack of nutrient renewal during flooding. A decrease in traditional agriculture has been reported in the flood plains below the Volta. Dam (Hilton and Kown-tsei, 1972), and Lake Kariba apparently does not produce the protein that the area was capable of producing prior to the darn.
Besides changes in silt content certain other parameters of down-stream water quality will be altered by chemical and biological processes occurring in the reservoir. For example, water removed from the reservoir hypolimnion may be deoxygenated and laden with hydrogen sulphide which can alter its usefulness as a domestic and industrial water source.
Only actual surveys of downstream and coastal fisheries and wildlife resources enable protective measures to be planned and implemented before completion of the dam, and such surveys are strongly recommended. The Gulf of New Guinea, for example, could be greatly altered from its currently rich fishery potential by a complete hydroelectric development of the Purari river system.
The possibility that changes in the sediment load carried to the sea can alter patterns of beach and bay formation is of serious concern. Sandy beaches, by nature of their small extent, are a valued commodity in the world for their aesthetic, scientific and recreational interest. Water storage projects throughout the world have reduced the sand formation process and in many areas, including much of the eastern seaboard of the United States, beaches are being continuously reduced. In some cases tourist industries have been adversely affected.
The effect of impoundment on coastal wildlife resources also deserves special attention since these coastal areas are essential to the maintenance and reproduction of many valuable and endangered wildlife species.
Deltas and their associated estuaries arc especially rich and dynamic areas in which sediment loads are deposited, resuspended and redeposited (Tison, 1964; Mikhailov, 1964). Altered flow rates and sediment loads can have particularly dramatic effects in these areas, particularly in relation to the distance of salt water intrusion (Goldman, 1971). The delta environment remains one of the richest of all ecosystems. Any alteration of its fertility should be viewed with great concern and the long-term cost/benefits reviewed with care.
An increase in health problems often accompanies the creation of large reservoirs in the tropics. The high concentration of people and the low level of sanitation common to work camps of the dam labour force create an ideal environment for the spread of malaria, dysentery, tuberculosis, venereal disease, etc. The labour force is both a source of new diseases and its people are highly susceptible to endemic diseases. The large population entering a previously sparsely inhabited region provides high host densities, thereby increasing the likelihood of the spread of many diseases. At different stages of dam construction in the tropics, increases and decreases in malaria, dysentery, schistosomiasis and onchocerciasis have been reported (Kershaw, 1966; Waddy, 1966; Warmann, 1969; Obeng, 1969; Paperna, 1969).
Factors relevant to vector ecology should receive special attention. Disease-causing agents, vectors, and the nutritional status of the people are all important variables in physical health and should be considered in evaluating the impact of hydroelectric dam projects. Methods of preventing health hazads associated with the project must be developed at an early stage in the planning process.
Health records in the El Cajon area reveal that aside from basic malnutrition, the most prevalent diseases in the area are tuberculosis, typhoid, dysentery and malaria. In the Purari delta area and coast, gastro-intestinal problems are common as well as a high incidence of malaria and tuberculosis.
It has been found in Africa that the stress associated with resettlement of the people inhabiting the water project site significantly increases disease and mortality, particularly among the very young and the very old. Therefore, it can be expected that the incidence of any diseases occurring at the site is likely to increase, and steps should be taken to offset this probability.
Aside from the necessity of an intensive health survey of the area and an inoculation programme, it should be pointed out that in the period of transition, normal cultivation will be interrupted, and the people's diet will be more tenuous than ever. Therefore, it is essential that the people receive sufficient nourishment during this transitional period until new communities are established. This could include milk programmes for children and food supplements for the families involved.
Malaria is endemic in most tropical areas of the world. Submerged trees and swampy higher aquatic plant habitats encourage the development of Anopheles mosquitoes, certain species of which carry the various malaria-causing protozoa. Mosquito habitats obviously must be avoided in the reservoir, and it is probable that the steep reservoir sides and fish population will provide adequate control over much of the areas. In view of the relatively high human population densities that the building of the dam will introduce into localized sections of the catchment area, anti-malarial prophylactic measures must be stringently followed.
Schistosomiasis bears certain ecological similarities to malaria since snail vectors (such as Bulinus and Biomphalaria) of the parasitic blood fluke (Schistosoma spp.) is favoured by submerged trees and higher aquatic plants. Transmission takes place, however, not through biting, but by penetration of humans by the free-swimming cercaria stage of the fluke. Although no cases were reported in Honduras, Biomphalaria-bearing Schistosoma mansoni does occur in South America and the Caribbean. Schistosomiasis has increased in association with water development in Rhodesia (Schiff, 1972) and Ghana (Paperna, 1969). Preventive measures include the destruction of terrestrial vegetation and control of higher aquatic vegetation as in the case of malaria. If schistosomiasis is not endemic to the area under consideration, detailed medical examinations should be performed on all personnel associated with the dam project in order to prevent the introduction of the disease. In addition, a careful survey should be made of aquatic habitats in the vicinity for possible snail hosts.
Onchocerciasis, or "river blindness", is a disease caused by the nematode Onchocerca which is transmitted to humans by the bite of the black fly, Simulium spp. The common name of the disease originates from the fact that blindness is produced in some of the victims. These black flies can breed effectively only in swiftly flowing and well-oxygenated waters, an environment which characterizes the catchment area rivers and which will decrease in extent as the reservoir basin fills (Waddy, 1966). In other words, the likelihood of onchocerciasis will diminish as a new dam slows the movement of water into the area and decaying vegetation produces deoxygenation in the reservoir. During construction on the Volta river DDT was used to control Simulium spp. larvae. Attention below the darn is, of course, warranted as this area. with lowered sediment transport may improve as a Simulium habitat. Further, there is the possibility that migrant labour could bring the disease into the area.
Gastro-intestinal problems (such as dysentery) attributable to water contact continue to be a common health problem in the tropics. Although the inevitable increase in human contact with the surface waters of the catchment area following completion of the dam will introduce protozoa and bacteria, the removal from the watershed of people not connected with construction of the dam would reduce human sources of infestation. This tactic should be combined with adequate treatment of drinking water supplies and wastewater produced by people remaining in the area.
The past history of watershed management provides global evidence of the need to anticipate sewage-treatment requirements. Lack of such foresight has resulted repeatedly in undesirable eutrophication and loss of sanitary water supplies. Therefore, a waste-water treatment study should be initiated that would provide future watershed populations with satisfactory treatment facilities. A series of strategically located small secondary treatment plants and septic tanks near population centres on the watershed is indicated by past experience to be an absolutely necessary investment.
Impoundment projects in the tropics often necessitate the relocation or resettlement of people inhabiting the area. Whenever possible, large-scale resettlement schemes should be avoided. However, if resettlement is found to be mandatory, several factors must be considered. First, it is essential that population relocation be recognized as an integral and important part of the hydroelectric project, and that local people be actively involved in the planning of resettlement.
The proposed Wabo Project in Papua New Guinea provides a good example of a small resettlement project; approximately 300-500 people and three villages would be displaced by inundation. A. thorough understanding of their social structure and careful consideration of their opinions about resettlement were found to be essential in formulating relocation plans. Assessment of their spiritual and physical attachments to the land would help prevent resentment over evacuation. The Government's handling of these people will be a major factor in establishing their relationship to the Government and in the social success or failure of the entire project.
Careful consideration should be given to the. touristic value of the impoundment area as compared to the touristic value of the proposed reservoir. The recreational use of an area is enhanced (in most cases but certainly not all) by water, and care must be exercised in promoting a hydroelectric project on the basis of its recreational benefits.
In view of the changing shoreline of the reservoir and the primary Junction of the watershed as erosion control combined with the need to protect developing fisheries, it is recommended that no tourist or recreational development be considered until the reservoir has been filled for several years. Only when the water quality has somewhat stabilized, the characteristics of the fish population are known, and the biologically important areas of the terrestrial part of the watershed have been studied should recreational facilities be considered.
The possibility of valuable archae-logical sites existing in the area should be investigated carefully to ensure proper excavation by archaeologist well before inundation is due to begin. Of considerable importance are the secondary impacts of impoundment projects caused by increased development made possible by the availability of electrical power and the potential for industrial development.
FILAMENTOUS BACTERIA ON A DETRITAL PARTICLE MAGNIFIED 10 900 TIMES a garbage collector
An evaluation of the anticipated long-range effects of any project is at best speculative. Nonetheless, attention must be focused on the spectrum of most likely environmental impacts in order to facilitate their future management and help the project become a rational component of future development for the region and nation involved.
Natural resources such as timber and mineral or petroleum reserves are usually much more effectively exploited by the development of electrical power. If an industrial complex is planned as part of a new dam project, it should be described and brief scenarios of the potential impacts of alternative plans for utilization of these resources made. Primary impacts of the industrial complex associated with the water project will include the construction of facilities on natural areas, pollution, and increased utilization of renewable and non-renewable resources.
In addition, during construction of any water development project impacts particular to construction will arise. For example, road development in previously unopened areas causes a series of ecological impacts including increased levels of erosion both during and after road construction. Further, the road serves as an avenue of travel for diseases and pests. The transportation of construction crews and equipment by barges will also have ecological and sociological impacts.
Secondary impacts on environment and social life result from any irrigation or water diversion and subsequent agricultural changes. The decision to embark on comprehensive development of the area should be based on a careful study of the problems and advantages of replacing naturally balanced ecosystems with man-made ones. Effects on the nation's health, nutrition, style and standard of living, land-use patterns and population growth rate should receive particular attention. In every case it is essential that local leadership become involved at an early stage and be provided with a broad view of both primary and secondary impacts of the proposed project.
BALON, E.K. & COCHE, A.G. 1974 (eds). Lake Kariba: a man-made tropical ecosystem in Central Africa. The Hague, Dr. W. Junk Publishers 767 p.
BEADLE, L.C. 1966 Prolonged stratification and deoxygenation in tropical lakes. Part 1: Crater Lake, Nkugute, Uganda, compared with Lakes Bunyoni and Edward. Limnol. Oceanogr. 11(2): 152-163.
COCHE, A. G. 1974 Limnological study of a tropical reservoir. In Lake Kariba: a man-made tropical ecosystem in Central Africa. p. 1-246. The Hague, Dr. W. Junk Publishers.
EWER, D.W. 1966 Biological investigations on the Volta Lake, May 1964 to May 1965. In Man-made lakes, (ed. R.H. Lowe-McConnell), p. 21-31. New York, Academic Press.
FREEMAN, P.H. 1974 The environmental impact of tropical dams. Guidelines for impact assessment based upon a case study of Volta Lake. Washington, D.C., Smithsonian Institute.
GOLDMAN, C.R. 1963 The measurement of primary productivity and limiting factors in freshwater with Carbon 14. In Proceedings, Conference on primary productivity measurement, marine and freshwater (ed. M.S. Doty), p. 103-113. US Atomic Energy Commision, TID-7633.
GOLDMAN, C. R. 1967 Integration of field and laboratory experiments in productivity studies. In Estuaries (ed. G.H. Lauff), p. 346-352. Publication No. 83, American Association for the Advancement of Sciences, Washington, D.C..
GOLDMAN, C.R. 1971 Reduced flows and the future ecology of the San Francisco-Bay Delta System. Miami, Florida. Proc. Symp. Hydrobiol., 8: 174-190.
GOLDMAN, C.R. 1972a El Cajon Project. Feasibility Study 5. Ecology. Prepared for Motor Colombus Consulting Engineers, Inc. and the World Bank with the support of Empresa National de Energía Electrica, Tegucigulpa, Honduras, 59 p.
GOLDMAN, C.R. 1972b The role of minor nutrients in limiting the productivity of aquatic ecosystems. In Nutrients and eutrophication (ed. G.E. Likens), p. 21-33. Amer. Soc. Limnol. Oceanog. Spec. Symp. No. 1.
GOLDMAN, C.R., HOFFMAN, R.W. & ALLISON, A. 1973 Environmental studies design, Purari River Development, Papua New Guinea. E.R.A. for United Nations Development Program, New York.
HARDING, D. Lake Kariba 1966- the hydrology and development of fisheries. In Man-made lakes (ea. R.H. Lowe-McConnell), p. 7-20. London, Academic Press.
HILTON, T.E. & KOWN-TSEI, J.Y. 1972 The impact of the Volta scheme on the lower Volta flood plains. J. Trop. Geogr., 1972: 29-37.
HUTCHINSON, G.E. 1957 A treatise on limnology. Vol. I. New York John Wiley and Sons, Inc. 1015 p
KERSHOW, W.E. 1966 The simulium problem and fishery development in the proposed Niger lake. In Man-made lakes (ed. R.H. Lowe MacConnell), p. 95-98. New York, London, Academic Press.
LITTLE, E.C.S. 1966 The invasion of man-made lakes by plants. In Man-made lakes (ed. R.H. Lowe McConnell), p. 75-86. New York, London, Academic Press.
MIKHAILOV, V.M. 1964 Hydrology and formation of river-mouth bass. Unesco Humid Tr. Res., 24(2): 59-64.
OBENG, L.E. 1969 The invertebrate fauna of acquatic plants of the Volta lake in relation to the spread of helminth parasites. In Man-made lakes (ed. L.E. Obeng), p. 320-325. Accra, Ghana University Press.
PAERL, H.W.W. & GOLDMAN, C.R. 1972a Heterotrophic assays in the detection of water masses at Lake Tahoe, California. Limnol. Oceanog., 17(1): 145-148.
PAERL, H.W. & GOLDMAN, C.R. 1972b Stimulation of heterotrophic and autotrophic activities of a plank-tonic microbial community by siltation at Lake Tahoe, California. Mem. Inst. Idrobiol., 29 Suppl.: 129-147.
PAPERNA, I. 1969 Snail vectors of human schistosomiasis in the newly formed Volta lake. In Man-made lakes (ed. L.E. Obeng), p. 326-330. Accra, Ghana University Press.
SCHIFF, C.J. The impact of agricultural 1972 development on aquatic systems and its effects on the epidemiology of schistosomiasis in Rhodesia. In The careless technology (eds M.T. Farvar and J.P. Milton), p. 102-108. Garden City, New York, Natural History Press.
TALLING, J.F. 1963 Origin of stratification 1963 in an African rift lake. Limnol. Oceanogr., 8(1): 68-78.
TISON, L. J. 1964 Problèmes de sedimentation 1964 dans les deltas. Humid Tropics Res., Unesco 24(2): 57 (Summary).
WADDY, B.B. 1966 Medical problems arising 1966 from the making of lakes (ed. R. H. Lowe- McConnell), p. 87-94. London, New York, Academic Press.
WARMANN, J. ST. G. 1969 Onchocerciasis 1969 and the Volta Dam construction. In Man-made lakes (ed. L.E. Obeng), p. 352. Accra, Ghana University Press.
