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

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

VOLTA LAKE (GHANA)

by

C.J. Vanderpuye
Department of Zoology
University of Ghana
Legon, Ghana

1. INTRODUCTION

The greatest deficiency of the report has been in its lack of sufficient information on pre-impoundment activities. In fact there was very little organized study in any discipline before impoundment. In the field of fisheries and hydrology, the singular exception was in limnology. Biswas (1966) started his studies on phytoplankton and essential nutrients in September 1963. The diversion tunnel was finally plugged in May 1964 resulting in the formation of the Volta Lake. His monitoring programme which was extended in 1964 to cover other areas of limnology continued to 1972. Thus his studies covered (1) pre-impoundment; (2) period of filling; and (3) post impoundment. Consequently, the usefulness of his observations cannot be over-emphasized in the present excercise.

Studies on fish biology and species composition changes in the commercial catch started during the filling period.

In 1968 (the year in which the lake attained the maximum controllable level) the Volta Lake Research Project was formed and a multidisciplinary research programme was initiated. Studies covered from the inception of the project included: stock assessment, catch assessment surveys, fishery limnology, fishing gear technology, fishery technology, socio-economics and agronomy. Other fields of study were added on in the course of project operation.

In spite of careful planning, there is no gainsaying the fact that the sequence of studies and their contents in Volta Lake did not fully exploit the maximum potential benefit of the reservoir resources. The present exercise seeks to ensure that this deficiency is rectified in the future inasmuch as the potential for the building of more dams in Africa is far from exhausted.

2. INDICES FOR EVALUATION OF CHANGES IN BIOLOGICAL PRODUCTIVITY

2.1 Changes in Reservoir Surface Area

Average monthly water level, volume and surface area have been showing a decreasing trend since 1975 (Table 1). Poor rains in the catchment area and increasing demand for power generation have been the cause. The lowest water level since impoundment is being experienced this year (1982). The overall effects of water level fluctuations and currently, decreasing volume, on hydrobiology and fisheries have not been studied.

2.2 Water Chemistry, Phytoplankton and Zooplankton

Limnological investigations started in the original Volta River in 1963 (Biswas, 1966). Parameters studied included: temperature, percent oxygen saturation, phosphate, silicate, sodium, potassium, calcium, total iron, colour of water in Hazen Units, nitrate, ammonia, sulphate, phytoplankton, pH, total alkalinity, secchi disc transparency, depth, blue-green algae, diatoms and green algae. Data were collected on weekly basis at Ajena (3 km uplake of Akosombo) until 1972 when this continuous monitoring programme to all intents and purposes teminated. During 1966–1968 investigations were extended to the Afram confluence, Kpandu, Kete-Krachi and Yeji. These stations were selected to reflect changes along the longitudinal axis of the lake. His results are summarized in Table 2.

Proszynska (1966) quantitatively studied Cladocera and Copepoda at Ajena (deepest part of the lake). Samples for comparison were also collected from a shallow bay (Ajena lagoon) close to Ajena. A few samples were also taken from the Afram confluence and the Volta River. Later study by Biswas (Lawson et al., 1969) included samples from the entire length of the lake. Her sampling method which differed in approach from that of Proszynska (1966), showed that on average the zooplankton was composed of around 90 percent Rotifera by number and only 10 percent for Crustacea and Protozoa (mainly ciliates combined).

2.3 Perphyton, Aufwuch and Benthos

Because the basin was not cleared of timber and brush before filling, flooded trees have provided a suitable substrate for periphyton in the epilimnion of inshore and offshore areas. Petr (1969) reported that because of the deoxygenated layer which developed at the bottom of the lake during the early stages, the biomass of periphyton exceeded by many times the biomass of benthos. Living in the mats of periphyton are various invertebrate forms, the bulk of which is made up of the larvae of Povilla adusta (Ephemeroptera) and chironomids.

The major importance of periphyton lies in the fact that the majority of invertebrates depend on plankton, periphyton, benthic organisms and detritus as their major food source. Such aquatic invertebrates formed the second most important food item of the fish investigated by Petr (1969).

2.4 Morpho-edaphic Index

The potential yield for Volta Lake as predicted with the morpho-edaphic index (MEI) model in Henderson and Welcomme (1974) is about 12 kg/ha (TDS = 50 mg/l) while the empirical catch ranged between 42.5 and 47.6 kg/ha per year in the period 1970 to 1976 (Coppola and Agadzi, 1977). The low potential yield calculated with the MEI is due to the low TDS present in Volta Lake since the mean depth is only 18.8 m.

3. FISHERY PARAMETERS

3.1 Trend in Yield

3.1.1 Commercial catch

Unfortunately, no quantitative data were collected in the Volta River before impoundment to enable an assessment of the total commercial catch to be made in the reservoir.

After fining a correlation between the variables: (1) volume of the expanding water; (2) abundance of fish available for exploitation; (3) concentration of fishing effort by yearly numbers of fishing canoes; and (4) limnological conditions (with R. Ryder), Bazigos (1970) made a reasonable backward projection of the total catch between 1964 and 1970 (Table 3). Further he found that the empirical yield for the years 1969 and 1970, when grouped on a six-monthly basis, followed closely the Gompertz curve. He suggested after evaluating the trend of the curve that the established zone of fluctuation of the expected yield has an upper limit of 45.14 kg/ha (empirical yield of 1970) and a lower limit of 42.90 kg/ha (lower asymptote value of the curve). Vanderpuye (1972) cautioned predictions on the then existing inadequate data. Since then fish catch statistics from 1970 to 1976 have shown that only 2 out of the 7 empirical annual yields fell within the predicated interval. So far, the pattern (Fig.1) has followed, to a large degree, that found by Kimsey (1957) in some California impoundments. In the study, he found that the decline in the fishery was followed by recovery and stabilization at a new and lower level. As of now, one can say with a reasonable degree of confidence that fish yield in Volta Lake has largely stabilized around 40 000 metric tons/year and that it will continue to fluctuate around this value for some time to come.

3.1.2 Comparison of commercial and experimental catches

Like the commercial fishery, no quantitative data were collected on stock abundance before impoundment. Stock sampling started at about the same time (Dec. 1968) as the catch sampling.

The following are some of the findings which emerged from comparison of the experimental and commercial catch data (Vanderpuye, 1972) (sampling sites are shown on Fig.1):

(1) Catches of the small meshed (13–90 mm) experimental gillnets (Fig.2) which largely represent species hardly exploited by the commercial fishery, accounted for 73 percent by weight of the total catch. Of this, the 13 and 25 mm meshes contributed 35 percent. Two species of clupeids (Cynothrissa mento and Pellonula afzeliusi) alone accounted for 23 percent.

Catches of the large meshes (102–205 mm) showed Tilapia as the most abundant (50 percent) in the commercial catch whilethe same genus accounted for only about 15 percent of the experimental catch. On the other hand, Lates niloticus was more abundant in the latter than in the former (Fig.3). Other key commercial species, too, showed differences to varying degrees. Vanderpuye (1972) suggested that the major source of difference was the grounds fished in the two instances. While the experimental fishing attempted to sample all depths of water, both inshore and offshore, and as many habitats as could be identified, the commercial fishery mostly specialized in the capture of Tilapia. Thus pressure on Tilapia was out of proportion to its apparent abundance. Consequently, Vanderpuye (1972) postulated that if it could be assumed that the stock data more approximate the true population structure than the catch data, then the ratio: catch percentage/stock percentage, should reflect the pressure being applied on each of the commercial species (or species groups) (Table 4). Calculated values of ratios less than l indicate that the species (or species group) is being underutilized in relation to its apparent proportion in the population. A ratio of more the l implies that the species (or species group) is being overutilized in relation to its apparent proportion in the population. Examples of genera which fell in the former category were Hydrocyon and Distichodus, and in the latter, Tilapia and Citharinus. The index provides a rough guide to management as regards the proportional utilization of the species present in Volta Lake.

Vanderpuye (1972) found a significant correlation (r = 0.95 at the 0.05 level of propability) (Tables 5a and b) between the stock and commercial catch data. This high correlation indicated that for the period investigated the stock data could be used to predict the catch data. Further, if the stock programme is kept constant, changes in amount and types of gear in the commercial fishery should be accompanied by a change in the ratio. Hence the parallel methods are most valuable in confirming changes in effort characteristics of the fishery.

The stock data again proved a good indicator of trend in the commercial catch when each set of data reflected independently the upsurge in abundance of Lates and a concomitant decline in Tilapia stocks (Vanderpuye, 1976).

3.2 Input Items

Items of information on the commercial fishery in this section and in Section 2.3 have been summarized from Bazigos (1970) and Coppola and Agadzi (1976 and 1977). They were collected during frame surveys in 1970 and 1975 and catch assessment surveys from 1969 to 1973 and from 1976 to 1978. Estimates for 1974, 1975 and 1979 were made from mathematical projections (Agadzi, pers.comm.).

3.2.1 Total number of fishing sites

The total number of fishing sites recorded in 1970 was 1 275. By 1975 the number had increased to 1 470. Regional changes in number of fishing villages are presented in Table 6a.

3.2.2 Total number of canoes

The 1970 Frame Survey recorded 12 074 canoes. By 1975 the number had increased to 13 814. Thus there was an increase of 14.4 percent over the five year period. Regional changes in number are presented in Table 6b.

3.2.3 Number of fishermen

Number of fishermen recorded in 1970 was 18 358. In 1975 it was 20 615 representing an increase of 12 percent.

3.2.4 Principal fishing gears

Principal gears employed in Volta Lake are: gillnets, castnets, lines and traps. Table 7 shows the gears used per day from 1970 to 1972 and 1976.

3.3 Output Items

3.3.1 Fish landings

This section is covered under trend and yield (Section 2.1).

3.3.2 Annual value of catch

A summary of data for 1974, 1976 and 1978 is presented in Table 8.

Using 1974 as a base period, it can be observed that the first stage transaction of fresh fish more than quadrupled by 1978. The total landed fresh fish also dropped by 8 percent from 1976 to 1978. One factor contributing to this may be the insufficient supply of fishing inputs to the fishermen.

3.3.3 Species distribution and composition

The only known pre-impoundment studies were by Irvine (1947) and Roberts (1967). Their efforts resulted in the listing of 110 species for the Volta basin. From October 1964 (five months after the lake had started filling) to August 1966 the Fisheries Department fished a battery of experimental gillnets near the dam at Akosombo. Catches were dominated at this early stage by characids: Alestes nurse, A. dentex/baremoze, and Hydrocynus spp. (Denyoh, 1969).

At the end of April 1965, some changes in composition were already evident (Table 9). Hydrocynus spp. which comprised 10.68 by number of the January catches had dropped to 1.80 percent in February. In March and April, none of the members of this genus was captured. Synodontis spp., Bagrus spp., Citharinus spp. and tilapias also failed to appear in appreciable numbers after January.

The available data from the Fisheries Department show a break in collection from April to October 1965. Collection was continuous from November 1965 to August 1966 (Tables 10 and 11). By January 1966, Chrysichthys had become the most important at this station comprising as much as 30.68 percent of the total catch for that month. Tilapias placed a poor second with 9.3 percent. The former species reached their peak in dominance in June 1966 when they comprised as much as 80 percent of the total catch. Thereafter, they began to decline gradually. It is worth noting that at this stage many of the species had started to disappear in catches. Notable among these were the genera: Bagrus, Clarias, Citharinus, and Heterotis.

The above observations were to a large extent confirmed by the investigations of Wuddah (1967) who fished the same mesh range at different localities but always in the general dam area. His results are given in Table 12. It shows that in January Synodontis spp. comprised 53 percent of the catch followed by Chrysichthys spp. (21.7 percent). Tilapias ranked third with 10.80 percent. After this month, the average catch of Synodontis was 10.38 percent. Concomitant with this decline was the steady rise of Chrysichthys spp. which reached a peak of 79.66 percent in July 1966. The general order of abundance agrees with that of the Fisheries Department.

Petr (1967, 1968) and Petr and Reynolds (1969) who sampled commercial catches at stations located along the longitudinal axis of the lake showed that in the riverine condition, the bulk of the population consisted of mormyrids, characids, mochokids, and schilbeids and that during the first two years many species (especially of Mormyridae) completely disappeared while tilapias became common in some parts of the lake. Later, there was: (i) a decrease in number of Alestes baremose and (ii) an increase of Citharinus in Area 4 and (iii) an increase in number of tilapias in fish catches in Areas 6 and 7.

Loiselle's (1972) sampling of the fauna in inshore habitats revealed that Chrisithys velifer had the highest biomass, accounting for 18.3 percent of the littoral ichthyomass. Sarotherodon niloticus was second (13.9 percent) and C. auratus third (7.5 percent) (Table 13).

As shown by Vanderpuye (1973), the clupeids (Cynothrissa mento and Pellonula afzeliusi) accounted for 23 percent by weight of the experimental gillnet catches for the period 1969 to 1971 (Table 14).

From 1968 to 1973, species distribution as reflected by experimental gillnet catches (Vanderpuye, 1976) showed two generalized patterns: (a) widespread distribution: e.g., Pellonula afzeliusi, Alestes macrolepidotus and Lates niloticus; (b) mainly limited to estuarine or riverine environment: e.g., Hydrocynus spp., Labeo spp., Mormyrids, Schilbeids, Cynotrhissa mento, Alestes nurse, A. baremose, A. dentex, and Citharinus spp.

Sampling with experimental gillnets in the Afram in 1980, however, showed that as the lake matures some species which at the initial stages of impoundments showed preference for the northern riverine environment were spreading to the southern lacustrine area. Species specifically collected were: Hyperopisus bebe, Petrocephalus simus, Hydrocynus lineatus, H. brevis, and Mormyrops deliciosus (Tables 15a and 15b).

3.3.4 Commercial catch

Because of the selectivity of gillnet for species and size, mesh sizes employed in the commercial fishery determine to some extent the species composition. After survey of the nets in use at 60 villages scattered throughout all the eight areas of the lake, Taylor (1969) showed that the proportion used in each mesh size was as given in Table 16.

Catch assessment surveys which aim at producing reliable estimates on a current basis of fish catch, and fishing effort were initiated in Volta Lake in 1969. Results of the survey for the period 1969–70 (Bazigos, 1970) showed that gillnets accounted for about 80 percent of the total commercial catch and that seven genera (namely, Lates, tilapias, Hydrocynon, Alestes, Citharinus and Distichodus) comprised about 80 percent of the total catch.

Data for 1970–1972 (Coppola and Agadzi, 1977) (Table 17) showed Lates and tilapias as the most abundant contributing about 50 percent to the total catch. Seven major genera (Tilapia, Sarotherodon, Lates, Labeo, Alestes, Citharinus and Distichodus) contributed 80 percent. Further, Lates showed an upward trend (16.40, 25.44 and 25.75 percent) while tilapias showed a downward trend (37.80, 27.35 and 29.16 percent).

3.3.5 Experimental catch

The stock assessment programme samples 30 stations with gillnets within the mesh size range: 13–205 mm at 13 mm interval, monitors changes in distribution, and species composition and abundance of the stocks on a continuing basis. For computer analysis, however, a geometric mesh size range of 13, 25, 40, 50, 65, 90, 115, 150 and 250 mm is used. The area (about 230 m²) in each mesh size is the same.

Tables 18 and 19 give species composition of total catches with experimental gillnets for the period May 1969 – October 1973 and by rounds, respectively.

Clupeidae (Pellonula afzeliusi and Cynothrissa mento) and Schilbeidae (Eutropius niloticus and Schilbe mystus) alone accounted for 57 percent of the total weight. Most of these species were caught with mesh sizes within the range of 13–50 mm. Eight species/families (comprising Clupeidae, Alestes baremose/dentex, A. macrolepidotus, Labeo, Lates, Chrysichthys spp. and Alestes nurse/leuciscus) contributed 90 percent.

To get an idea of the vertical distribution of the fishes, records were kept as to whether a fish was gilled in the top or the bottom half of a 3 m deep net and the top, middle or bottom one-third of a 9 m deep net. No such divisions were made in the catch of the 1.5 m deep net.

The distribution pattern for inshore and offshore catches of the important species is shown in Table 20. The pattern for many species was that of a truncated cone with the broad size facing upward. This indicates that many species prefer surface waters.

Three categories of nets: 1.5, 3 and 9 m deep nets, set in waters equivalent to their depth provided data for the determination of the lateral distribution of individual species by number and weight (Table 21).

The catch per net tended to increase from inshore to deeper water. The ratio of the catch in terms of numbers and weight in inshore waters by the three nets were: 1:2.22:3.61 and 1:2.26:4.23, respectively. In offshore waters where the nets fished only the surface waters, the ratio of catches by the 3 and 9 m deep nets dropped to 1.25:1.81 and 0.96:2.03 for numbers and weight, respectively. The difference between the two categories of nets was thus comparatively small.

Of more interest, is the catch per net area fished. Since the 3 and 9 m nets were 2 and 6 times, respectively, the size of the 1.5 m nets, the expected ratio of catches of the 3 nets should also be 1:2:6. In this respect, the highest catch per net area for both numbers and weight was realized in the 3 m deep net. The catch dropped to about half the magnitude in the 9 m deep net. In offshore areas, catches were still lower.

3.3.6 Diversity and ecogroups

The species diversity index which measures the number of different species present in the population and the relative proportion of their abundances (Pielou, 1969) was determined for each area of the lake and each round sampling (Table 22).

The species diversity increased fairly regularly from the southern lacustrine to the northern riverine. Except for two rounds, the highest index was obtained for Area 8. Sampling stations in this area were located in the Black Volta, White Volta and their confluence. Lowest values were recorded during four rounds for Area 1 (southern lacustrine).

Evans and Vanderpuye (1973) divided the fish species in Volta Lake into four categories according to principal feeding habits: aufwuch-insect-and-detritus-feeders, piscivores, semi-pelagic omnivores and benthic omnivores (Table 23).

This broad division of the fishes into ecological groups (ecogroups) based on feeding habits provides an insight into overall utilization of food resources of the lake and also ichthyomas in the various trophic levels.

Abundance of the groups measured by weight of the catch revealed that: semi-pelagic omnivores comprised 80 percent of the overall catch. The bulk of this group was made up of the clupeid, Pellonula afzeliusi which accounted for 32 percent by weight of the group. Other important elements were Eutropius niloticus (16 percent), Alestes baremose/dentex (10.6 percent) and Alestes macrolepidotus (8.0 percent (Fig.4).

The next important group was the aufwuch-insect-and-detritus-feeders which comprised 10 percent by weight. The group includes some of the large elements in the population and hence it is of considerable economic importance.

The piscivores ranked third in abundance accounting for 6 percent by weight of the total catch. Important elements comprised Lates niloticus and Hydrocynus spp. The group seemed to be increasing in catches up to the end of round 8. Their level in the total catch for round 9 however was lower.

3.3.7 Standing crop of fish

The result of Loiselle's (1972) ichthyomass survey of the littoral fauna at 23 stations located in all areas (except VIII) is given in Table 24. Highest ichthyomass was recorded for Area IV (interphase between lacustrine and riverine) (268.9 kg/ha). This was followed by the southern lacustrine (Areas I-III) (209.48 kg/ha). The poorest was the northern riverine part (Area V-VII) (189.38 kg/ha). The average for the whole lake was 200.86 kg/ha). Mean depth of the stations ranged from 15–180 cm with an overall average of 55.6 cm.

3.3.8 Frequency of occurrence of individual species

Since there were 30 stations and each station was sampled 8 times, each species had the chance of appearing in a maximum of 240 samples in experimental gillnetting. The frequency with which-species appeared in the samples and the calculated percentages are shown in Tables 25 and 26.

4. RESEARCH AND DEVELOPMENT HISTORY

4.1 Applied Fishery Research Activities

These have been outlined in Table 27. For the purpose of this paper, all research activities on Volta Lake could be considered to fall in this category.

With the exception of limnology, all studies commenced after the closure of the dam. Studies on some aspects of the biology of the principal species, changes in species composition of the commercial catch, microbiology, primary productivity, and secondary productivity were initiated during the filling period. Full scale multidisciplinary developmental research activities came into being with the formation of the Volta Lake Research Project in 1968. The overall objective of the project is to maximize the advantages and minimize the disadvantages (mainly public health problems) attendant to the formation of the lake.

The studies which did not subscribe to the immediate objectives of the project were pure water chemistry and microbiology. These studies were sponsored by the University of Ghana which had other commitments aside from its contribution to the total research effort of the project. Nonetheless, some of the results of these results proved useful in elucidating some of the phenomena in the lake.

The first phase of the project (January 1968 – October 1973) was research oriented. The second phase (October 1973 – October 1977) emphasized the use of already accumulated data for comprehensive development of the lake area. Since then the content and objectives of the work programme have remained essentially the same.

Table 28 gives the chronological order in which FAO experts arrived on the project. The order also provides a rough guide to the commencement of research activities in the areas of the experts. Studies initiated by Ghanaian researchers in the fields of fishing gear technology. Fishery biology and limnology were in progress before arrival of the FAO experts.

Ideally, many of these lines of research should have started before impoundment. Late approval of Ghana Government's request for assistance from the UN Special Fund was the principal cause.

Now with the benefit of hindsight, and with the experience gathered in the course of participation in pre-impoundment studies on the stretch of the Volta River later to become the Kpong Electric Project Head Pond (about 24 km south of Akosombo dam), I would have set up my research programme in the priority order outlined in Table 29. It must be emphasized that the order of priority in the table is specifically for Volta Lake. A broader generalization might be possible after the syntheses from other reservoirs have been considered.

4.2 Fishery Infrastructure

The fishing villages along the Volta sprang up spontaneously when the Volta Lake formed. Immigrant fishermen from the lower Volta area moved upstream and established their own villages with no Governmental assistance.

The keynote to the development of the fishing industry on the Volta Lake was the transformation of all the major landing points into ‘fishery complexes’ where the following facilities would be provided: (a) permanent ramps for safe landing of boats carrying fish and other agricultural produce; (b) market area; (c) supply house for sale fishing inputs; (d) smoking houses for proper processing of fish; (e) training centre for fishermen, boat-builders and outboard motor mechanics; (f) boatyards for construction of fishing and transport boats; and (g) cold storage to encourage landing of fresh fish.

Kpandu Torkor, the first of such complexes to be built has been operational since 1975. It is hoped that others will be built at Yeji, Kwamekron, Ampem, Dzemeni and Buipe, and minor ones at Dambai, Ketekrachi, Abotoase and Yapei.

Under a fish processing demonstration programme, processing complexes would be established at points planned to be developed into fishery complexes. There are also plans to install some of the project's improved ovens in a number of the large villages. Currently, fish processing plants built with governmental assistance (through Volta Lake Research and Development Project) have been established at Kpandu Torkor and Ekye Amanfrom on the Afram.

An embryonic cold storage with a flake ice machine was to be established within the Kpandu Torkor complex. The idea has, however, been dropped because of insufficient fresh fish from the lake.

Boatbuilding and repair yards are available at Kpandu Torkor for the construction of fishing canoes and transport boats. It is intended to build another at Yeji with CIDA assistance when the scheme takes off.

Cost of equipping a fishing unit is shown in Table 30.

4.3 Fishery Development Activities

Fishery development activities have comprised the following (Table 31):

1. Kpandu Torkor Fishery Complex: This was commissioned in October 1976. Cost of construction was C936 457.50.

2. Fish Processing Demonstration Programme: This is an ongoing activity with an annual budget of C10 000. It started in 1973 with the Amankwakrom Processing Centre. The centre wound up in 1975 owing to management problems. New centres have, however, been opened at Kpandu Torkor in 1957 and at Ekye Amanfrom in 1980.

3. Kpandu Torkor Complex Fishery School: Started in February 1977, this school had trained 61 boys and 31 girls as of December 191.

4. Mobile Fishery School: This is a Christian Council of Ghana project sponsored by the Bread for the World Council of Churches. It has an annual budget of C100 000.00. The project trains resettled populations under the VRA resettlement programme in the techniques of fishing.

5. Floating Jetty Programme: Was initiated in 1973 with four Jetties constructed at a cost of C27 981. The programme was suspended in 1977 for unsuitable design of jetties. (They were being damaged by squally winds).

6. Yeji Fishery Complex: Projected to cost ¢3.8 m in 1978. The project is yet to take off after completion of feasibility studies. C100 000.00 have been earmarked for site studies during the fiscal year 1982/83.

7. Net Revolving Fund Programme: This was initiated in 1974. Under the programme, improved fishing canoes constructed by the VLR and DP and 5–10 Hp outboard engines are sold to the fishermen. To date, 215 outboard engines (cost: ¢53 085.33) and 58 canoes (cost: ¢184 262.64) have been sold to the fishermen.

8. Fishery Extension Service: This principally introduces nylon monofilament fishing nets to fishermen. Threethousandsixhundredthirtysix bundles of nets have been sold since 1975. The programme also extends better techniques of smoking fish to fishermen. In this connection ten improved ovens with temperature regulating mechanisms have been installed at Kpandu Torkor and Ekye Amanfrom.

9. Loans to fishermen: This is a credit scheme planned for former students of Kpandu Torkor Complex Fishery School. It has yet to become operational.

10. Lakeside Fishermen Cooperative Development Programme: Involves the organization of fishermen into cooperative groups. These cooperatives serve as conduits through which aids and services are extended to fishermen.

5. OTHER USES OF THE LAKE

The Volta River project was declared economically viable on the basis of the hydroelectric project alone. Other ancillary projects originally planned to be attached to the hdyroelectric project were:

1. Akosombo/Greater Accra water supply and irrigation scheme;

2. Volta Lake transport system;

3. Volta Lake fishery scheme;

4. Rural electrification scheme; and

5. Bauxite development.

To date, the irrigation, rural electrification, and bauxite development projects have still not been implemented because of financial constraints. In the interim, however, other uses of the lake which are less financially demanding have emerged. Below is a brief summary of projects for which information is readily available.

Electricity: Available Volta River Authority (VRA) Annual Reports (1979, 1980) state that income surpluses in 1979 and 1980, after deducting depreciation and loan interests, were ¢35 320 000 and ¢37 116 000 respectively. The reports add that this rate of return is estimated to be equivalent to about 3 percent return if assets are valued to reflect current replacement costs.

Forestry: A major proposal made in this field has been the afforestation of selected areas with quick maturing trees for the establishment of a charcoal industry and to arrest soil erosion. The project has yet to take off.

Wildlife: It is proposed to set up a National Park on the Dwija Arm of the lake. In connection with this, management studies are in progress and the project is yet to be fully implemented.

Agriculture: Two major types are practised: namely, (a) drawdown, and (b) irrigated agriculture.

1. Irrigated Agriculture: It has been established that about 10 000 ha of land around the lake is potentially suitable for irrigated agriculture. Data on a pilot trial at Ampem are presented in Table 32;

2. Drawdown Agriculture: The normal drop in the lake level ranges between 2–4 m. When there is a drop of about 3.4 m, it is estimated that at least 80 000 ha of land become available for agriculture in the drawdown areas. Presently only 7.5 percent is being cultivated. Pilot crop trials by the project at selected points seem to be paving the way to a more intensive cultivation of these areas.

Volta Lake Transport: Passenger and cargo traffic services are run by a government-owned company known as the Volta Lake Transport Company (VLTC). The VLTC operates one passenger vessel and one cargo barge from Akosombo in the south to Yapei in the north. From all accounts, both services are heavily patronised (VRA Annual Report, 1980) (Table 31). The only factor which has been hampering progress has been the current low water level which makes the northernmost ports inaccessible. As a result there has been a diminution in freight volume after a peak in 1973. Passenger traffic, however, continues to show an upward trend.

The total capital investment in the Company at the end of 1980 stood at ¢1 396 000. In 1980 loan agreements covering DM75.2 million (or ¢104.80 million) for the improvement and expansion of the Volta Lake transport system were signed by a West German company on one side the Government of Ghana on the other.

6. DISCUSSION

The nutrient status of Volta Lake on which the overall productivity could be assessed, has been discussed by Livingstone (1974). He concluded that: (1) Volta Lake is oligotrophic; and (2) the nutrient-poor status of the lake waters inhibits fish production. These observations did not wholly find favour with Obeng-Asamoa (1977). While he is prepared to accept oligotrophy on the basis of the nutrient content of the lake waters, he states that the standing crop of other biological indicators and primary production rates point to a state higher than oligotrophy. This paradox is brought about by the following: essential nutrients released by the breakdown and mineralization of the organic load are immediately picked up by the algae and Ceratophyllum; the activity is fueled by high water temperatures resulting in quick turnover and high net productivity. Consequently, measurement of the TDS alone gives a wrong picture of the nutrient status of Volta Lake.

Because of this situation, the morpho-edaphic index which utilizes the TDS-like factor (conductivity) in the formula for prediction, underestimates fish yield in Volta Lake by as much as two fold. The proponents of the MEI (Henderson and Welcomme, 1974), however, set the confidence limits at ½ x and 2x. This means that the true value could be 2-fold less or more. This is certainly too wide for many practical purposes. Administrators and fishery managers would usually like a more precise prediction to enable them make definitive plans for the future. It behoves us therefore to find an alternative method (or a modification of the MEI) for a more precise prediction of yield. In this respect we might like to consider Jenkin's (1967) multiple regression approach to the problem. Among his multiple regressions of greatest apparent utility were:

1. Standing crop on dissolved solids divided by mean depth (i.e., MEI); and

2. Standing crop on dissolved solids, shore development and storage ratio.

Where the former does not appear to be satisfactory to our situation the latter might be.

Limnological sampling which started in 1963 offers us the opportunity of observing the continuous change in the behaviour of some of the important parameters from pre-impoundment, through filling to post-impoundment. A close examination shows the usual differences attributable to riverine and lacustrine environments. The general pattern is:

1. Virtually no difference between surface and bottom reading in the river;

2. Trend in the difference in the two readings during the filling period;

3. Marked difference in the two readings after formation of the lake; and

4. Thereafter, marked difference in the two readings during period of stratification and little difference during the period of mixing.

Ryder (pers.comm.) considered that the most important parameter monitored by Biswas which could be related directly to fish yield was alkalinity. Working closely with Bazigos (1970), he discovered a positive relationship between yield and alkalinity. Upon this basis, he projected yield backwards for the years 1964 to 1969.

The lake basin was not cleared of timber and brush before impoundment. Consequently, during the initial stages of lake formation, the decomposing organic matter which consumed most of the oxygen rendered the bottom unsuitable for fish life, and resulted in extensive fish kills.

This leads directly to the question: Should the basin of a future reservoir be cleared before flooding? Biologically, the presence of flooded vegetation is advantageous to the fishery for the following reasons:

1. It serves as a substratum for periphyton. The deoxygenated layer which develops at the bottom at the early stages of impoundment may result in the biomass of periphyton exceeding the biomass of benthos. Periphyton also shelters various aquatic invertebrate forms which formed the second most important food item of the fishes investigated by Petr (1969);

2. It makes fishing inefficient at the peripheral areas where the commercial fishery is mostly located. The vegetation serves as a check against overfishing;

3. It serves as substratum for attachment of the eggs of some fishes;

4. It decreases the number of niches available which can be occupied by fishes and thus adds to species diversity. The more diverse the species the more effectively different food resources are utilized in the ecosystem;

5. During early stages of impoundment some fish species feed on leaves and soft parts of the drowned vegetation;

6. As the flooded vegetation rots, it adds to the organic content of the lake thus contributing to the general increase in lake productivity.

The list of disadvantages is short but some of the factors are as potent as to offset most of the advantages. These are as follows:

1. The deep deoxygenated layer which results as the vegetation decomposes may cause extensive fish kills during the early stages of impoundment. In reservoirs located in deep sheltered valleys, the deoxygenated layer may persist throughout the life of the reservoir thus limiting the volume of water available for fish life;

2. Flooded vegetation does considerable damage to fishing nets. In Volta Lake it has prevented the use of moving gear for more effective harvest of the clupeids which are presently underutilized;

3. It obstructs boating and other sporting activities.

A solution to the problem appears to be: cleared and uncleared areas. The degree of clearance should depend upon the location of the reservoir (sheltered or unsheltered), morphometry of the basin, density of the vegetation and available funds.

So far, the trend in yield in Volta Lake has closely followed the pattern demonstrated for some Californian impoundments by Kimsey (1975). Four phases are discernible: (i) a rise; (ii) a decline; (iii) a recovery; and (iv) stabilization at a new and lower level.

A well-designed gillnet sampling programme (Evans and Vanderpuye, 1973) which incorporates among other things (a) judicious selection of mesh sizes and (b) location of stations in all identifiable habitats, should be able to effectively monitor changes in the fish stocks under exploitation. If samples could be considered sufficiently representative then a reasonable fishing pressure index could be developed by calculating the ratio: catch per percentage/stock per percentage for each species (Vanderpuye, 1972). This fishing pressure index provides a rough guide to management in its goal towards rational harvest of the species in a multi-species fishery.

Fish population studies in Volta Lake seem to suggest the following order of dominance in species succession in the southern sector which has become the lost lake-like:

1. Alestes nurse, A. dentex/baremose and Hydrocyon spp.: These were mainly species which were dominant in the river before impoundment. The were favoured during the first year of impoundment when riverine condition mostly prevailed;

2. Synodontis spp.: These are benthic feeders which prefer deep offshore bottom areas. They were probably flushed out when the bottom vegetation began to decay;

3. Chrysichthys spp.: Are benthic omnivores which are also quite abundant in shallow inshore areas. Their predominance may be related to opportunistic exploitation of myriad food items which became available as new areas were flooded;

4. Inshore - tilapias: Are predominantly found in inshore waters of depth less than 3 m (Vanderpuye, 1976b). They feed on higher plants, aufwuch, detritus and plankton. Their predominance may be related to abundance of these food items in shallow waters.

   Offshore - Pellonula afzeliusi and Lates niloticus: The planktivorous Pellonula afzeliusi is the most abundant species (according to experimental gillnet catches) in the lake. Studies (Vanderpuye and Ocansey, 1972) have shown that it constitutes the main food source of piscivorous Lates niloticus.

In the northern shallow basin where riverine conditions still prevail, tilapias are most abundant. Here, they occur in both inshore and offshore areas (Vanderpuye, 1976). Other important species are: Lates niloticus, Alestes spp., Hydrocynus spp., Eutropius niloticus, Schilbe mystus, Citharinus spp. and Mormyrids. Many of these species appear to be riverine-specific and are consequently confined to this area of the lake. The tiger fish (Hydrocyon) was thought to fall in this category. Sampling in the Afram Arm (1980) and study of the recent commercial catches, however, have shown that they have spread to the southern lacustrine sector. Their upsurge in abundance coupled with that of Lates niloticus seems to lend credence to the theory that the piscivores tend to increase in abundance (with a slight time lag) following the explosion of the semi-pelagic fish and other organisms (Evans and Vanderpuye, 1973). If this is true, then the future should witness further expansion of these pescivores.

Species diversity was highest in the riverine sector of the lake. It appears to be a function of degree of flow in Volta Lake. If this is true then it should be negatively correlated with the degree of lacustrinization and for that matter the length of the retention period when the reservoir is considered as a unit.

7. REFERENCES

Bazigos, G.P., 1970 Yield indices in inland fisheries with special reference to Volta Lake. Rome, FAO, Volta Lake Research and Development Project, Statistical studies, FAO/SF/GHA/10/St.S./3:25 p.

Biswas, S., 1966 Ecological studies of phytoplankton in the newly forming Volta Lake of Ghana. J.W.Afr.Sci.Assoc., 11:14–9

Coppola, S.R. and K. Agadzi, 1976 Frame surveys at Volta Lake (Ghana). Rome, FAO, Volta Lake Research and Development Project, Statistical studies, FAO/GHA/71/533/St.S./5:148 p.

Denyoh, F.M.K., 1969 Changes in fish population and gear selectivity in the Volta Lake. In Man-made lakes: the Accra Symposium, edited by L.E. Obeng. Accra, Ghana Universities Press, for Ghana Academy of Sciences. pp.206–19

Evans, W.A. and J. Vanderpuye, 1973 Early development of the fish populations and fisheries of Volta Lake. Geophys.Monogr., 17:114–20

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 Morphoedaphic et le nombre de pecheurs des pecheries des eaux continentales d'Afrique. CIFA Occas.Pap./Doc.Occas.CPCA, (1):19 p.

Irvine, F.R., 1947 The fish and fisheries of the Gold Coast. London, Crown Agents for the Colonies, 352 p.

Jenkins, R.M., 1967 The influence of some environmental factors on standing crop and harvest of fishes in U.S. reservoirs. In Reservoir fishery resources symposium. Washington, D.C., American Fisheries Society, pp.198–321

Kimsey, J.B., 1957 Fisheries problems in impounded waters of California and the Lower Colorado River. Trans.Am.Fish.Soc., 87:319–32

Lawson, G.W., 1969 Hydrobiological work of the Volta Basin Research Project. Bull.Inst.Fondam.Afr. Noire (A Sci.Nat.), 31:965–1003

Livingstone, D.A., 1974 Changes in mineral cycling associated with the formation of the Volta Lake. In Environmental aspects of a large tropical reservoir. Preliminary report on a case study of the Volta Lake, Ghana. Washington, D.C., Office of International and Environmental Programs, Smithsonian Institution.

Loiselle, P.V., 1972 Preliminary survey of inshore habitats in the Volta Lake. A report prepared for the Volta Lake Fisheries Research Project. Rome, FAO, FI/DP/GHA/67/510/2:122 p.

Obeng-Asamoa, E.K., 1977 A limnological study of the Afram arm of Volta Lake. Hydrobiologia, 55:257–64

Petr, T., 1967 Fish population changes in the Volta Lake in Ghana during its first sixteen months. Hydrobiologia, 30:193–220

Petr, T., 1968 Fish population changes in the Volta Lake over the period September 1966 –December 1967. Tech.Rep.Volta Basin Res.Proj.Univ.Ghana, (X23):3 p.

Petr, T., 1969 Development of the bottom fauna in the man-made Volta Lake in Ghana. Verh.Int. Theor.Angew.Ver.Limnol., 17:273–82

Petr, T. and J.D.Reynolds, 1969 Fish population changes in the Volta Lake in 1968. Tech.Rep.Volta Basin Res.Proj.Univ.Ghana, (X32):3 p.

Pielou, E.C., 1969 An introduction to mathematical ecology. London, Wiley - Interscience.

Proszynska, M., 1966 A quantitative study of the Cladocera and Copepoda in the Volta Lake. Tech.Rep. Volta Basin Res.Proj.Univ.Ghana, (X10)

Roberts, T., 1967 A provisional checklist of the freshwater fishes of the Volta Basin with note on species of possible economic importance. J.W.Afr.Sci.Assoc., 12:10–8

Taylor, G.T. and F. Denyoh, 1969 Final report. Akosombo, Volta Lake Research Project Working report, Ref.No.122/52:10 p.

Vanderpuye, C.J., 1972 Fishery resource assessment and monitoring in the development and control of fisheries in Lake Volta. Afr.J.Trop.Hydrobiol.Fish., Spec.Issue 2:115–34

Vanderpuye, C.J., 1973 Population of Clupeids in Volta Lake. Ghana J.Sci., 13(2):179–93

Vanderpuye, C.J., 1976 Studies on the fish population in Volta Lake five to nine years after impoundment. Paper presented at the Volta Lake Conference held at the University of Ghana, Legon, 39 p. (mimeo)

Vanderpuye, C.J., 1976a Investigation into causes of increase in abundance of Lates niloticus and concomitant decline of Tilapia spp. stocks in Volta Lake. Paper presented at the Volta Lake Conference held at the University of Ghana, Legon, 18 p. (mimeo)

Vanderpuye, C.J. and P.Ocansey, 1972 A preliminary study on the biology of Lates niloticus (Linnaeus 1762) in Volta Lake. Akosombo, Volta Lake Research Project. Technical Report, 33 p. (mimeo)

Volta River Authority, 1979 Eighteenth annual report and accounts. Akosombo, Volta Lake Authority

Volta River Authority, 1980 Nineteenth annual report and accounts. Akosombo, Volta Lake Authority

Wuddah, A.A., 1967 Studies on the biology of the species of Tilapia in the Volta Lake. M.Sc.Thesis. University of Ghana, Legon, 121 p.

Table 1. Elevation, surface area and volume of Volta Lake from 1971–72

Year:1971.

Year: 1972.

Year:1973.

Year: 1974

Year:1975.

Year: 1976

Year: 1977

Year: 1978

Year: 1979

Year: 1980

Year: 1981

Year: 1982