C. THE MAHAWELI BASIN (SRI LANKA) (contd.)

4. INLAND FISHERIES IN THE MAHAWELI BASIN

In Sri Lanka the inland fisheries are characterized by two major factors: firstly it is almost entirely confined to the man-made lacustrine water bodies that are perennial, and secondly an exotic species tilapia Oreochromis mossambicus forms the backbone of this fishery (Fernando and Indrasena, 1969; Fernando, 1971; De Silva, in press; Fernando and De Silva, 1984). The reasons for the success of this exotic, the present status of the fishery as a whole, and certain management strategies of the fishery have been documented (Wijeyaratne and Costa, 1981; De Silva, 1985). The fishery of the Mahaweli Basin is no exception to this general pattern. However, the Mahaweli Basin being the largest, and having the largest floodplain as well as a large surface area of reservoirs both ancient and recent, has a major role to play in the inland fisheries.

4.1 Riverine fishery

The Mahaweli Ganga, like the other rivers in Sri Lanka does not support a commercial fishery. A subsistence fishery using sporadic cast netting, hook and line is carried out throughout the river length. Willey (1910) documented the paucity of a riverine fishery on the island and briefly dealt with the methods of existing subsistence fishery, which varied from hook and line to poisoning using extracts of poisonous creepers. The most intensive fishing takes place in the estuary where it is limited mostly to shrimps. Apart from shrimps, milkfish (Chanos Chanos) and grey mullet are captured by using castnets. The season of the shrimp fishery is between March and August. However, there are no statistics available.

4.2 Reservoirs

The distribution of reservoirs in the Mahaweli Basin is shown in Figure 11. The reservoirs in Sri Lanka are classified according to irrigation terminology into (i) the major irrigation reservoirs, (ii) the medium-scale irrigation reservoirs, (iii) the minor irrigation reservoirs, and (iv) the hill-country reservoirs.

Figure 10. Population density in 1978 in administrative districts falling within the Mahaweli Basin.

Figure 11. The major reservoirs, villus (= floodplain lakes) and mangrove swamps of the Mahaweli Basin (from Tams, 1980).

However, from the point of view of fisheries development, those in the Mahaweli Basin can be categorized into two types: (i) the perennial reservoirs, and (ii) the seasonal or village tanks. The latter tend to dry up almost completely for two to three months in the year, and generally are of a smaller size, often being less than 20 ha at the full supply level. These reservoirs are presently utilized only marginally for fish production, even though their potential use for fish production was highlighted by Mendis (1965). The major fishing activities of the Mahaweli Basin are confined to the major perennial reservoirs in the basin which approximate 10,000 ha of surface area.

4.2.1 The fishery

The total yield from the reservoir fishery of the basin presently stands at approximately 3,000 metric tons (1980 data) averaging a yield of 375 kg/ha/annum. Some morphometric, hydrological and physico-chemical features of relevance to fish production and the present status of the fisheries in the major reservoirs are summarized in Tables 12 and 13 respectively.

The fishery in all the reservoirs is dominated by the exotic species Oreochromis mossambicus (Peters). Fernando and Indrasena (1969) and De Silva and Fernando (1980) traced the history of the development of the fishery of the Parakrama Samudra reservoir and showed convincingly the importance of O. mossambicus to the fishery. Fernando and Indrasena (1969) and later Fernando (1971) suggested that the introduction of this species in 1952 has been instrumental in the establishment of the reservoir fishery in Sri Lanka (Fig.12). Mendis (1977) erroneously interpreted that Tilapia rendalli, introduced in 1962, dominated the fishery in this reservoir. There is increasing evidence that in almost all the reservoirs throughout the island, O. mossambicus accounts for more than 50% of the landings (De Silva, in press; Fernando, 1977).

4.2.2 Crafts, gear and efficiency

The reservoir fishery throughout the island remains somewhat uniform in that the type of crafts and the gear used remain the same. Two types of crafts were used until recently (Fig. 13), dug-out log canoe or a lograft, both being manoeuvered by two persons. These are gradually being replaced by fibre glass canoes of similar design and size under a recently introduced subsidy scheme (Sri Lanka, 1980). The fibre-glass craft has not increased the fishing efficiency in comparison to the traditional dug-out canoe. The biggest advantage of the former is its longer lifespan.

The gear is exclusively a gill-net of mesh size ranging from 8–12 cm. Each craft carries on average four net lengths, each length of approximately 20 m. The nets are laid at dusk and hauled during the early hours of the morning. Beach seining is prohibited in the reservoirs of Sri Lanka and the topography of the bottom and the numerous obstructions, primarily trees, prevent the use of any bottom dragging gear. The uniformity of the crafts and the gear used permits the computation and the comparison of catch per unit effort data from different reservoir fisheries. In most of the reservoirs fishing effort has been on the increase over the last few years, and particularly so in those reservoirs which have already had established fisheries and which are located in the vicinity of townships. The best examples are the Parakrama Samudra and the Minneriya reservoirs of the Mahaweli Basin.

a - NEDECO (1979);
b - Personal observations (1980–1983);
c - Oglesby (1981);
d - Mendis (1965);
e - Gunatilaka and Senaratne (1981);
* - Kantalai reservoir is located at the edge of the northernlimit of the Basin

* medium-scale reservoirs; Giritale - in the Mahaweli Basin and Pimburettewain the Maduru Oya Basin

Figure 12. Total catch from Parakrama Samudra (Fernandoand De Silva, 1984).

Figure 13. The crafts: the old type (A) and the new (b). Note the bundle of netting - gillnet of 8-12 cm stretched-mesh.

The changes in the effort, measured as the number of craft-days/ha year over the period for which reliable data are available for the Parakrama Samudra and Minneriya fishery are shown in Figure 14. The data available show that while the effort has been steadily rising (in the case of Parakrama Samudra since 1976 and in both reservoirs since 1980), in both reservoirs yield/craft/year showed a significant decline since 1980. the present yield per craft is 3.7 metric tons (mt) and 4.5 mt in Parakrama Samudra and Minneriya respectively - a yield which is barely profitable to the fisherman.

4.2.3 Seasonal changes

Seasonal changes in the total yield of the two major fisheries of the Mahaweli Basin reservoirs are shown in Figure 15. The seasonal changes in the total yields as well as the yield per unit of effort follow a similar trend in that the yield increase during the rainy periods, the highest yield being in September to December in both reservoirs. The biological significance of this phenomenon has been dealt earlier by De Silva (1983).

4.2.4 Species composition

As stated earlier, the reservoir fishery in Sri Lanka is dominated by the exotic cichlid, Oreochromis mossambicus (Peters), introduced in 1952 (Fernando and Indrasena, 1969; Fernando, 1977). The same holds true for the reservoirs of the Mahaweli Basin (De Silva and Fernando, 1980). Fernando and Indrasena (1969) listed the fish species that are found in Parakrama Samudra reservoir and this was recently updated by Schiemer (1983) (Table 14).

Apart from O. mossambicus only two other species (Labeo dussumieri and Puntius sarana) contribute significantly to the commercial catches, at any time of the year. In addition to these three species, others such as Wallago, Channa sp. etc., also occur in the catches in small numbers, and are important constituents because they command a relatively high price.

The percentage contribution of the two important indigenous species viz. Puntius sarana and Labeo dussumieri to the total yield is shown in Figure 16. Since 1976 there has been a decrease in catches of the indigenous species, and there are also seasonal changes (Fig. 17) with a significant increase during the rainy season (De Silva, 1983).

4.2.5 Recent trends

The data available on the mean landing size of O. mossambicus from the Parakrama Samudra fishery over the years indicate a gradual decrease in the mean landing size (Table 15). The decrease in the mean landing size was interpreted as an indication of stunting of the population (TAMS, 1980). TAMS (1980) however, failed to recognize the possible effects of increased fishing pressure on the mean landing size as well as the changes in the gear utilized in the early phases of the fishery.

More recently in all the major reservoirs of the Mahaweli Basin the half beak, Hyporamphus gaimardi has begun to appear. In the Kaudulla and Minneriya reservoirs, as well as in certain reservoirs of the Walawe Ganga basin a small fishery has developed for this species. The present estimated yield from Kaudulla reservoir approximates 20 kg/ha/year.

Figure 14. Fishing effort and yield per unit effort in the Parakrama Samudra and Minneriya reservoir fisheries.
• effort
o Yield

Figure 15. Seasonal change in the Parakrama Samudra and Minneriya reservoir fisheries.
•   Yield
o   Yield per crft

a - data from Schiemer, 1983;
b - personal observations; breeding - within reservoirsor not;
O - omnivore;
H - herbivore;
C - carnivore;
Z - zooplanktivore

Figure 16. Yield of indigenous species and their percentage contribution to the total yield in the Parakrama Samudra and Minneriya reservoir fisheries.

Figure 17. Seasonal changes in the yield of Labeo dussumieri and Puntius sarana (Parakrama Samudra and Minneriya reservoir fisheries) (from De Silva, 1983).

4.2.6 Untapped resources

The existing fisheries of the major reservoirs remain uniform in the gear and the crafts utilized. Catches are dominated by the exotic O. mos-sambicus, which is supplemented mostly by two indigenous species. As such, the smaller indigenous species, in particular the long snouted barb - Puntius dorsalis, the filamented barb - Puntius filamentosus, the green barb - Puntius chola, giant danio - Danio aequipinnatus, the attentive carp - Amblypharyngodon melettinus, the indigenous carnivores such as snake heads - ophiocephalids, cat fishes and eels remain underexploited or unexploited. Preliminary results of experimental fishing in Parakrama Samudra (Schiemer and Hofer, 1983) have indicated considerable resources of minor cyprinids. In a few reservoirs elsewhere on the island, notably in Mahawillachchiya and Nachchaduwa, a fishery for D. aequipinnatus and P. filamentosus exists in addition to the O. mossambicus fishery.

The occurrence of carnivores in present catches is accidental and negligible. It is suggested that initial surveys for long-lining and use of small-meshed gill-nets should be carried out to ascertain the extent of these resources. Based on such evaluations, management measures could be introduced to effectively tap these resources without harming the existing fishery. A tentative estimate based on the very preliminary fishing activities suggests that a minor carp yield of 20–50 kg/ha/year and a carnivore yield of 10–25 kg/ha/year could be obtained in addition to the existing yield from the major fishery.

4.2.7 Biological considerations

The reservoir capture fisheries on the major reservoirs depends almost entirely on natural recruitment. Apart from the lacustrine species, which are exotic, the reservoir fish populations contain a number of indigenous riverine as well as marsh-dwelling species, some of which presently contribute significantly to the constituent fishery. Evidence has been brought forward to show that the commercially important indigenous species do not breed in the reservoirs, and that they undertake a spawning migration, which is also indirectly responsible for the higher yields during the rainy period (De Silva, 1983). Apart from these species the same appears to be true for some other indigenous minor cyprinids such as Puntius dorsalis, P. filamentosus and a few other species.

A gradual decrease from 34.2 cm in 1957 to 21.8 cm in 1980 in the mean landing size of O. mossambicus in the Parakrama Samudra reservoir has been interpreted as an indication of stunting (TAMS, 1980). However, this could also be the result of the increased fishing pressure which has taken place over the same period of time. The fishing pressure has increased almost ten fold and under such conditions even if the gear and the mesh size have not changed, a considerable decrease in the mean landing size is to be expected. The present mean landing size of O. mossambicus is comparable to that in fisheries on much larger water bodies elsewhere in Asia or Africa (e.g. Gwahaba, 1973; Welcomme, 1967), and fish of this size have a high market acceptability.

Since 1952, the major reservoirs of the Mahaweli Basin have been stocked with at least four exotics: O. mossambicus, Tilapia rendalli (= melanopleura), Cyprinus carpio and O. niloticus and to a lesser extent with Ctenopharyngodon idella. However, apart from O. mossambicus which dominates the fishery, ony T. rendalli occurs in the fishery to any extent. The biological reasons for the dominance of O. mossambicus and the lack of success of the other species remain obscure. Apart from the two cichlids there is no evidence of natural recruitment of the other introduced species.

4.2.8 Marketing

Marketing of the reservoir fish is very traditional. Each fisherman or groups of fishermen sell their catches to specific vendors, most of whom transport the fish on bicycles (Fig. 18). The traders arrive at the landing sites - which for a major reservoir vary from 4 to 6 - in the early hours of the morning, and buy the catch at a pre-fixed price. The price paid to the producer by the vendor is determined by the latter, and has not increased over the years in relation to the increase in the cost of living index.

The vendors trade in specific townships which they have allocated amongst themselves. Trading is mostly from house to house. The retail price of fish increases when outside labour is brought in for harvesting. In addition, the prices also tend to be determined by availability.

In most areas the selling price is determined by the size of fish, with a bigger fish fetching a higher price than a smaller one. Fish are rarely sold by weight.

During the peak fishing periods a significant proportion of the catch is salted and sun-dried. Sun-drying is particularly popular amongst the migrant fishermen who move into the reservoir from the coast during the monsoonal periods.

4.3 Flood lakes - Villus fishery

Sri Lanka is estimated to have only 40,000 ha of flood lakes (here-after referred to as villus) of which 16,000 ha are reported to be in the Mahaweli Basin (Hunting Surveys, 1962). The more recent estimate (TAMS, 1980) of villu area is given as 12,800 ha, with 50 distinct villus. Of these the three largest villus (Fig. 11) are Handapan Villus, Karapola Villu, and Velankadu Villu, which range in area from 400 to 900 ha.

Villus are vegetated land areas saturated with water seasonally and retaining their connection to the main river most of the year. A villu generally consists of two portions; the ponding area which is permanently filled with water, and the floodplain which is inundated by water during the flood period.

Villus constitute a major habitat for a number of indigenous species of fish, amphibians and birds, and a grazing ground for livestock. The villus are also the breeding ground of some riverine fishes, such as the commercially important Labeo dussumieri (Dr. M. Smith, personal communication).

Fishery statistics are not available for most of the villus. Fernando (1977) gave a total mean yield of 84,490 kg (141 kg/ha/year) for the years 1964–67 for Karapola Villu, one of the largest villus of the Mahaweli Basin. It was also shown that O. mossambicus constituted 65% of the catch. More data are available since 1980 (Table 16), with a mean fish yield from the villus being 133 kg/ha/year. The available evidence suggests that the fish yield is, at least partially, determined by the existing fishing pressure and that in some of them the fishing pressure could be increased without harming the fish stock. The author has observed that the mean size of O. mossambicus landed was 18 cm (approx. 170 g), which is considerably smaller than that landed from reservoirs. Mahaseer (Tor khudree), each weighing 8–10 kg, were not uncommon in the villus. During the flood seasons an increased number of people take up to fishing with cast-nets, primarily in the littoral, and the main constituent species of the catch during this period are Labeo dussumieri and Puntius sarana (Dr. M. Smith, personal communication).

Figure 18. A - Vendors gathered at one of the landing sites.
B - A vendor selling his goods.

A fishery also exists for a shrimp (Macrobrachium sp.) in all the major villus. This fishery is seasonal, being confined to the dry months and is carried out by women wading into the littoral and seining through aquatic plants (Fig. 19). The average catch per fisherman per day ranges from 6 to 8 kg (wet weight). In Karapola Villu, for example, about 20 women are engaged in this activity each day and the season lasts for about 12 weeks. The approximate shrimp yield from this villu per annum is computed as 12 mt/year.

4.4 Aquaculture

The aquaculture practices in the Mahaweli Basin which are still in their formative stages are both extensive, e.g. stocking and harvesting of village seasonal tanks, and entensive, e.g. pond and cage culture.

The Mahaweli Basin is estimated to have nearly 4,000 ha of seasonal tanks, most of which are less than 10 ha in surface area. Stocking and harvesting of four village tanks was undertaken during the 1981/82 growing period. These tanks were stocked with equal proportion of bighead carp, grass carp and O. niloticus fingerlings ranging from 2,500 to 3,500 fingerlings per ha. Only two of the tanks were harvested while the others dried up before the end of the growth period (Thayaparan, 1982).

Intensive aquaculture is still carried out only on an experimental scale. Sri Lanka has no aquaculture tradition in contrast to other countries of the region, perhaps primarily due to its strict Buddhist traditions. The current intensive aquaculture involves pond and cage culture. Preliminary results of these operations are discussed by Hettiarachchi (1982) and Wannigama and Weerakoon (1982).

Figure 19. A - Women fishing for shrimp in the littoral area of the villu - Karapola villu, Manampitiya.
B - A catch of shrimp.

5. THE MAHAWELI GANGA DIVERSION AND DEVELOPMENT

The Mahaweli Ganga diversion and the associated river basin development remains the biggest project ever undertaken in Sri Lanka, in terms of the capital input, the extent of sociological, economical and ecological impact and in the time-scale for its completion. The present development scheme cannot, however, be divorced from the ancient works associated with the Mahaweli Ganga, and this is to be expected in the light of the irrigational supremacy that was prevalent in the Sri Lankan nation for over 2,000 years (Schnitter, 1967).

5.1 Historical aspects

“Let not even a drop of rain water go to the sea without benefitting man” (PARAKRAMA BAHU THE GREAT, 1153 – 1186 AD).

The old Sinhala civilizations were centered in the dry zone. The earliest kingdoms were based in Anuradhapura and Polonnaruwa in the heart of the dry zone. It was therefore, imperative for their survival that all efforts were made to maximize the utilization of the water that fell on the earth. It was due to the foresight of rulers such as King Parakrama Bahu, whose regime is recognized as one of the most important irrigational eras in the history of the island, that vast-scale irrigational works were created. There were several ancient major irrigation systems in the Mahaweli Basin (Fig.20). These went into disrepair over the years. Since the 19th century a number of channels have been reconstructed, such as the Elahera-Minneriya Yoda Ela, almost all the major tanks (reservoirs) reconstructed and rehabilitated and the downstream irrigation works developed and improved. However, there have been no major improvements in channel alignments.

Three major ancient reservoirs are fully functioning today. The Parakrama Samudra in its present form is a result of a connection of three original reservoirs. The oldest, northern basin - the Thopawewa, the middle basin - the Eramudu Wewa and the southern basin - the Dumbutulla Wewa; the latter two reservoirs being constructed during the reign of King Parakrama Bahu. The southern basin receives water from a channel from the Amban Ganga, regulated by an anicut at Angamidella.

Minneriya tank is believed to have been constructed by King Mahasena (273–303 A.D.). The tank was restored to the present status in 1903 and subsequently improved and enlarged in 1953.

Kandulla tank was known to be in working order in 1680 A.D. The reservoir was probably constructed by King Mahasena (273–303 A.D.) and repaired and restored by Vijeyabahu (1055–1110 A.D.) and finally improved upon by Parakrama Bahu (Arumugam, 1969). The present restoration commenced in 1958. It impounds water from the Gal Oya and Alut Oya and also receives water from the Minneriya - Kantalai Yoda Ela.

5.2 Justification and rationale for modern development

At the time of initiation of the idea of the diversion of the Mahaweli Ganga and the development of the basin, together with the Maduru Oya Basin and that of others in the North and North Central Provinces of Sri Lanka, the population of the island was 11.8 million. The actual growth of the population was far exceeding the trends based on census in the latter part of the century (Fig.21). However, the present rate of growth which stands at 1.7 (1975–1982) is much lower than that of 3.1 (1965–1976) when the project was being planned.

Figure 20. Anicient irrigation works of the Mahaweli Basin (from Hunting Survey Corporation, 1962).

To meet the demand for food for the increasing population as well as for other subsidiary demands such as power, the agricultural resources of the land - in particular in the dry zone - would have to be developed. Most of the ancient civilizations were centred in the dry zone, which was agriculturally developed. For reasons which are still not well understood, the dry zone was gradually abandoned and the population centres shifted to the wet zone. This process was accelerated with the introduction of cash crops by the colonial powers. The abandoned dry zone area was soon overgrown with jungle and diseases such as typhoid and malaria plagued the region; the gradual breakdown of population centres in the dry zone could have been a result of these epidemics.

These epidemics were controlled since the Second World War with the introduction of pesticides. Since the population in the wet zone has been growing rapidly in this century, it has become imperative that attempts be made to redevelop the once prosperous dry zone. The post-independent governments therefore commenced the redevelopment of the dry zone, initially through the restoration of the ancient irrigation works.

Figure 21. Population changes in Sri Lanka (modified from UNDP/FAO Report, 1969).

Such developments and subsequent settling of these lands enabled the country to cut down on its food imports the value of which was nearing 50% of the total import expenditure of the country.

The dry zone agricultural area is almost entirely dependent upon irrigation and further development could be assured only with a reliable water supply. A reliable water supply would not only bring fresh areas under the plough, but would also enable cultivation throughout the year. Apart from food, the pressure for increased electric power was mounting, not only because of the increase in the population but also because of the industrialization which essentially commenced in 1960. The projected demand for electricity (Fig.22) necessitated harnessing of the hydroelectric potential of a number of rivers.

If the agricultural production was to keep pace with the growing population, the irrigation development was the only way of achieving it. This would be done in the first instance by ensuring an adequate supply of water to the existing systems, and secondly by expanding irrigational facilities to bring new areas under cultivation. It is in this context that the Mahaweli Ganga Development Scheme was proposed with the aim of attaining self-sufficiency in rice and highland crops and to meet the projected extensive demand for hydropower.

5.3 Initial planning

The earliest detailed survey of the Mahaweli Basin was carried out by Hunting Survey Corporation, as a result of an agreement made by the Canadian and Sri Lankan (then Ceylon) governments through the Colombo Plan Administration. The Mahaweli Basin study was one of the three river basin studies undertaken under this agreement. The report on the Mahaweli Basin was presented in 1962 to the Government. The survey provided an inventory of land and water resources of the basin, and provided a basis for evaluation of the potential for development, and preparation of tentative plans for the future use of land and water resources.

The availability of this detailed study may have prompted many political leaders who had interests in irrigational works, to give serious consideration to harnessing the resources of this river basin which covered one-sixth of the island. By the early 1960s it was generally accepted that the vast water resource of the Mahaweli Ganga had to be harnessed effectively. It is in this context that the then Ceylon Government requested assistance from the Special Fund of the United Nations to survey the Mahaweli Ganga Basin and the dry zone areas in the North and Central Provinces of the island in order to provide basic information for a plan to develop the land and water resources for irrigation and hydropower. The final agreements were signed on 12 October 1964 between the Government of Ceylon, the United Nations Development Programme and the Food and Agriculture Organization of the United Nations.

The initial survey was carried out in two stages between March 1965 and February 1967 and included the Mahaweli Basin, Maduru Oya Basin and the rivers of the West side of the North and Central Provinces.

The Master Plan envisaged to utilize 5,800 × 10⁶ m³ of water of the Mahaweli and its tributaries, supplemented by further 1,111 × 10⁶ m³ of water from Maduru Oya and the rivers of the North-Central Province. The potentially usable resource of 6,911 × 10⁶ m³ per year was considered sufficient to irrigate 285,000 ha of paddy throughout the year, or 530,000 ha of alternate crops such as cotton and groundnuts.

Figure 22. Hydropower demand (projected) and development from 1951 to 1982 in Sri Lanka (from UNDP/FAO Report, 1969).

Of the total grassland area of 2.53 × 10⁶ ha in the project, the total suitable land area for irrigation was estimated as 610,000 ha, of which 122,000 is already under cultivation. Nearly 90% of the latter land is irrigated, the remainder being used for rain-fed crops.

The proposed scheme was expected to bring under command 366,000 ha of which 266,000 ha would be underdeveloped land, of which about 150,000 ha or 56% would be in the Mahaweli Ganga and Maduru Oya Basins and the remainder would be in the north-central part of the country.

The proposal to meet the above demands contained the impoundment of 15 reservoirs, 11 of them with hydropower stations with a total installed capacity of 508 megawatts and with a total output of firm power of 2,037 million kwh per annum.

The overall project was to be completed in three phases, each phase with its own set of projects, spanning a thirty-year period. Each phase was to take into consideration the changes in demand for power, irrigation needs, etc., with time and was to be independent of each other in execution and subsequent evaluation.

The three phases suggested were:

It was envisaged that on completion of the scheme the agricultural production will: match the expected population increases; satisfy the expected rise in the demand for agricultural produce per caput; minimize the food imports and be a net exporter for certain agricultural produce; and increase the farm income. The major works of the Master Plan are also summarized in Table 17.

5.4 The accelerated Mahaweli development

The first project of Phase I was completed when the country was faced with steep inflation in 1973, primarily due to the oil price hike. The government changed in 1977, when 40% of the country's import bill was being spent on food imports and unemployment was in the region of 20% of the normal workforce (Anon., 1978). The rapidly rising inflation also indicated that the later the date of completion of the scheme, the higher would be the ultimate cost. According to one estimate the construction cost could double over a delay of eight years. It was in the light of this situation that the new government reviewed the project and made a policy decision to accelerate the Mahaweli Ganga Development Programme by telescoping the thirty-year planned programme to a six-year programme, commencing in 1979.

a - location as in Fig. 24;
b - estimates made in 1969, extracted from UNDP/FAO, 1969

The Accelerated Mahaweli Ganga Development Programme did not alter the basic infrastructure of the original Master Plan, but the area to be developed was brought down to 140,000 ha in the Mahaweli Ganga and Maduru Oya Basins and an additional 23,000 ha in the Kalawewa Basin (Figs.23 and 24). The reservoirs to be constructed are the Kotmale Reservoir, Victoria Reservoir, Ulhitiya-Rathkinda Reservoir and Maduru Oya Reservoir (Table 18). The interconnections of the Mahaweli Ganga with the adjacent waterways and the envisaged water distribution pattern of the Mahaweli Ganga Diversion and Development Programme is schematically shown in Figure 25. As a result of the proposed acceleration, instead of sequencing the programme of financing the construction, maintenance and operation, settlement and agriculture, hydropower generation and distribution, community development and infrastructural services, the government has decided that the major aspects be carried out simultaneously. Expectedly, many problems are bound to arise in such a telescoping of events, particularly with respect to the nature of colonization. Therefore, as an important aspect of the planning and implementation, considerable flexibility is permitted and continuous evaluations are made and necessary adjustments carried out.

5.4.1 Environmental considerations

Although some of the present trends of human endeavour bring man into direct conflict with wildlife, the extremely old traditions, associated with the spread of Buddhism in Sri Lanka have undoubtedly resulted in a great concern for preservation of wildlife and their habitats. Such concern is often expressed openly by the populace and endeavours are made in all quarters to preserve this heritage in the context of the modern age.

The Accelerated Mahaweli Ganga Development Programme undoubtedly will come to fruition only with the destruction and/or clearing of a vast area of land of its natural forest vegetation, and the subsequent introduction of monocultural practices on these lands. It is heartening, however, that considerable attention has been paid in the scheme to preservation of wildlife.

In the original Master Plan an approximate area of 120,000 ha was set apart for forest and wildlife resources. This area has been extended by a further 52,000 ha through the creation of another national park. Altogether there will be four national parks in the Mahaweli Basin and adjacent areas (Table 19). These reserves undoubtedly would help in the preservation of the endemic species (8 fish species, 4 amphibians, 19 reptiles, 8 birds and 3 mammals) and the several endangered species inhabiting the plains of the Mahaweli Ganga, as well as 53 species of endemic plants (TAMS, 1980).

Figure 23. Scope of the development scheme as envisaged in the Master Plan of 1969 and that of the Accelerated Mahaweli Ganga Development Programme (Anon, 1978).

Figure 24. Major features of the Accelerated Mahaweli Ganga Development.

Figure 25. Schematic representation of the water distribution management in the Accelerated Mahaweli Ganga Development (Sri Lanka, 1983).

Consideration has also been given to the fuelwood requirements of the increasing population of the project area, and special areas have been reserved for fuelwood plantations. In the same context, measures have been taken to preserve the forest cover in the upper catchment of the Mahaweli Ganga and also to restrict activities such as the cultivation of seasonal crops like tobacco on steep, degraded slopes which results in soil erosion and alleviate sedimentation of rivers and reservoirs.

5.4.2 Sedimentation

Useful life of a reservoir is determined by the rate of sedimentation and the rate of sedimentation is dependent on the rate of sediment input by river or rivers entering the reservoir.

The ancient reservoirs of the Mahaweli Basin have existed until now without a significant loss in their storage capacities. Data available on sediment transport of the Mahaweli Ganga are scanty. NEDECO (1979) computed the annual transport of sediment at two points in the river and in Amban Ganga (Table 20).

Using the above figures, the yield of sediment in the catchment areas has been computed. It averages 300 m³/km²/annum in the main river as compared to 100 m³/km²/annum in the tributaries.

Steps are being taken to reduce this sediment yield further through cultural practices. For example, plantation of annual crops on the slopes of the upper catchment have been prohibited and reforestation of the reservoir catchment is in progress.

5.4.3 Settlements, agriculture and infrastructural development

The detailed projected land-use pattern of the Mahaweli Basin and Maduru Oya Basin are shown in Figure 26. The settlement plans envisaged under the Accelerated Mahaweli Development are different from those hitherto adopted for other schemes. The objective of the pattern of settlement is to provide a scope for the development of well-knit communities, and is to be carried out on the principle of 'cluster" settlements. Under the Accelerated Mahaweli Programme a steep increase in the production output per unit land area is also envisaged. The tentative estimated production of the major crops under the scheme is given in Table 21.

Figure 26. The existing and proposed land use pattern of systems A, B, C & D in the Accelerated Mahaweli Ganga Development (modified from Trams, 1980).

During the planning and execution of the project, the following major aspects have been considered in detail in the Master Plan (UNDP/FAO, 1969), as well as in the re-evaluation carried out in 1977–79 for the acceleration of the scheme (NEDECO, 1979): hydrology, hydropower generation, human settlements, agricultural production, other facilities essential for human life such as roads, health care and some others. However, little attention was paid to the fisheries in the future reservoirs. As far as the author is aware, the only mention of fisheries activities in official publications were made in 1979 (Sri Lanka, 1979) and in 1983 (Sri Lanka 1983), when it was stated that the Inland Fisheries of Victoria Reservoir would be developed to yield 1,500 tons per year. In order to achieve this target the entire reservoir bed would have to be cleared of all vegetation prior to impoundment. It was also mentioned that a fish yield of 2,950 tons per year was expected from the Mahaweli Development Programme. However, fisheries was not included in the overall cost-benefit evaluation of the project.

5.4.4 Development of fisheries - Planning considerations

On behalf of the United States Government, a group of consultants, “Tippetts - Abbett - McCarthy - Stratton (TAMS)”, was called upon to provide an environmental assessment for the project. This also included fisheries considerations.

Recommendations made by TAMS (1980) and by other fisheries scientists were not heeded. One of such recommendations concerned bush clearing prior to reservoir flooding. It is accepted that if there are no major irrigational and/or hydraulic requirements, it is advantageous not to clear 20 – 25% of the reservoir bed (Bhukaswan, 1980). Such areas, though hindering fishing activities, would provide shelter from predators, an additional food resource and breeding areas for many fish. The decaying vegetation would continue releasing nutrients into the water, thus enhancing the aquatic production (Oglesby, 1977).

However, what happened was that in the Accelerated Mahaweli Development approach, all vegetation was completely cleared and burnt before impounding the area. But Sri Lanka offers an excellent example of the advantage of having vegetation in the reservoir. In Parakrama Samudra, one of its three distinct basins is devoid of decaying vegetation with the result that nobody is fishing there. In the other two basins, which have large amounts of vegetation, fishermen are very active in spite of the difficulties with moving around and with setting and removing fishing gear.

It has also been felt that in the settlement pattern and land allocation, some land near reservoirs should have been reserved for families to be involved in the fisheries. As such, allotees who are likely to take up fishing will do so only as a complementary activity to farming and consequently it will take a few years before they develop fisheries expertise.

In addition to the yield of 2,950 t per year from capture fisheries on reservoirs, the government is now also encouraging development of aquaculture in the area with a view to not only meeting the protein requirements, but also as a means of providing additional employment (Sri Lanka, 1983).

a - Master Plan projection
* - green gram, black gram and cowpea

a - estimated

6. MAHAWELI GANGA BASIN DEVELOPMENT AND FISHERIES

The backbone of the existing fisheries of the Mahaweli Basin is the reservoir fishery, which contributes significantly to the total fish yield of Sri Lanka. The development of the reservoir fishery in the Mahaweli Basin and its contribution to inland fisheries is shown in Table 22.

The hydrological development of the basin is likely to affect fish yield and fisheries in a number of ways:

1. by manipulation of discharges, which will especially affect the existing reservoir fisheries and those on floodplains (villu);

2. by providing 24,000 ha of lacustrine water bodies (TAMS, 1980), 360 km of main channels and 4,080 km of distributory channels which will become available to fisheries (Anon., 1978);

3. by allowing the penetration of seawater further up-river during low discharge periods;

4. by the impact of modified nutrient discharges on sea life.

6.1 Reservoirs

6.1.1 Existing fisheries

Fish production in lacustrine waters is closely linked to the nutrient availability (Ryder, 1965; Oglesby, 1977; Welcomme and Henderson, 1976). It is predicted that as a result of the Mahaweli Ganga Development the retention time in the existing reservoirs in the basin will decrease appreciably, i.e. by nearly half in the case of Kaudulla and Kantalai reservoirs. Recent limnological studies on the Parakrama Samudra (Schiemer and Duncan, 1983) have indicated that even under the existing regime the retention time is insufficient for the completion of the production cycle of some planktonic organisms. A further shortening of retention time may worsen the situation, resulting in a lower phytoplankton and zooplankton production and therefore in a reduced food supply for fish.

On the other hand, TAMS (1980) has suggested that although the flushing rates would increase, the availability of nutrients in the system should be adequate to maintain a high aquatic productivity. But Gunatilaka (1983) has shown that in the Parakrama Samudra, the currently most productive reservoir of the basin, the phosphate availability is limiting. The data available for this reservoir may be indicative of the level of fish production which we may expect in the new reservoirs.

The importance of water level fluctuations on the abundance and yield of commercially valuable fish species in floodplains (Welcomme and Hagborg, 1977; Dudley, 1979) as well as in lacustrine (Marshall, 1982; Beam, 1983) fisheries is well documented. In the case of the Parakrama Samudra fishery it has been shown that the yield of the cichlid O. mossam-bicus was positively correlated to the degree of fluctuation of the water level two years earlier (De Silva, 1985a).

According to NEDECO (1979) the development of the Mahaweli Basin should not result in any significant changes in water level fluctuations as compared with those prior to the diversion of the river. The data available seem to suggest the contrary. Figure 27 suggests that the degree of fluctuation of the water levels in the Parakrama Samudra and Minneriya reservoirs for which data are available have changed significantly to that of the pre-diversion period.

Figure 27. The mean monthly water level at the main sluices of the Parakrama Samudra and Minneriya reservoirs prior to and after receipt of Mahaweli waters in 1976.

n.a. - not available

For the time being, the yield predictions can only be based on a comparison of yield figures from the existing reservoirs for the years prior to the receipt of Mahaweli waters with those for the post-diversion period (Table 23). Based on this the expected yield from these reservoirs from the existing fishery is 2,800 t/annum.

6.1.2 New reservoirs

The Mahaweli Ganga diversion and the consequent development of the Basin, which is carried out in conjunction with the Maduru Oya Basin development, will result in the creation of large areas of lacustrine bodies. Their hydrological characteristics are summarised in Table 24.

The total reservoir area to be created approximates nearly one-fifth of that of the existing major reservoirs (Mendis, 1977; Fernando and De Silva, 1984). Of that, 3,770 ha will be located in the hill-country and have primarily a hydroelectric power production function. These reservoirs will be deeper than most of the other reservoirs. In these reservoirs the temperature could fluctuate as much as 10°C per day, and because of their depth they are likely to develop a strong thermocline and oxygen deficiency in deep water. Similar phenomena have been reported, for example, for Indian reservoirs (Ganapati, 1973). On the other hand, detailed limnological studies of Parakrama Samudra have indicated that the shallow, low-country Sri Lanka reservoirs are unlikely to develop and retain a strong thermocline and/or chemical stratification for long (Bauer, 1983).

6.1.2.1 Stocking

There have been no firm plans yet regarding the stocking of these reservoirs, although some views have been expressed by individuals interested in the development of the reservoir fisheries industry.

- not yet impounded or no data

Indian experts of the Ministry of Fisheries (Dr. A. Sreenivasan and Mr. K. Chakrabarthy) are of the view that the downstream reservoirs should be stocked with Indian major carps, viz. rohu (Labeo rohita), mrigal (Cirrhinus mrigal) and catla (Catla catla). Some of the older reservoirs have been repeatedly stocked with Chinese major carps since 1976, and with common carp since before 1976. However, there is no evidence that these species have established themselves in the major reservoirs, and their returns from the stockings have been almost nil (Fernando and De Silva, 1984). The likely reasons for this lack of success have been discussed by Fernando and De Silva (1985) and De Silva (1985a) If a stocking and harvesting procedure is to be carried out, the input costs are going to be very high, and without reasonable returns. Logistically, it would be more advantageous to utilize the facilities available and those forthcoming for the stocking and harvesting programmes of the sesonal tanks (FAO Report, 1982). This programme is to be extended to the Mahaweli Basin.

The author is of the view that the downstream reservoirs will follow a pattern already in existence in the dry zone major reservoirs, in that Oreochromis mossambicus will begin to dominate the fishery. It would be advantageous to stock this species together with O. niloticus, to enable a more efficient utilization of the likely high levels of algal production which will be encountered in the early post-impoundment period of reservoirs (Bhukaswan, 1980).

6.1.2.2 Yield estimates

Fish production of the hill-country reservoirs is likely to be significantly different from that of the downstream area and hence has to be considered separately. Estimates of potential fish production for the different reservoir types were made by Mendis (1965) and Oglesby (1980) and for those in the Mahaweli Basin by TAMS (1980) (Table 25).

The TAMS (1980) estimates were specifically made for the Mahaweli Basin reservoirs (Table 26). They, however, based their estimates on an equation derived by Wijeyaratne and Costa (1981) for Sri Lankan reservoirs which was based on one year's yield data from 7 reservoirs of which 4 were not taken into consideration on the unauthenticated suggestion that the available statistics for the latter were erroneous. The basis of the derivation of the yield/MEI relationship is contrary to the basic concepts put forward by Ryder (1982) and did not meet the necessary requirements of the model. The present author (De Silva, 1985a) has shown in detail that the MEI concept is not applicable to Sri Lankan reservoirs, both qualitatively and quantitatively. Therefore, the extension of the findings of Wijeyaratne and Costa (1981) and derivation of potential yields from specific reservoirs based on electrical conductivity measurements as carried out by TAMS (1980) is questionable. A more realistic method of estimation of the potential yield, in the light of the lack of a suitable model, would be to use the existing yield values of those reservoirs and make an extrapolation. The same approach was adopted by Oglesby (1980) for yield prediction for Sri Lankan reservoirs. Reliable estimates of the existing production of the reservoirs are only available for 4 reservoirs in the Mahaweli Basin and adjacent areas (Table 27), and the mean yield amounts to 372 kg/ha/annum. However, the mean yield for reservoirs over 2,000 ha was 262 kg/ha/annum and for those less than 1,000 ha was 710 kg/ha/annum.

Using the above figures and estimates of Mendis (1977) for floodplain lakes and hill-country reservoirs the estimated total fish production from the newly impounded perennial reservoirs as a result of the Mahaweli Diversion Scheme would be 3,539 t/annum. The details of the breakdown are given in Table 28.

The present estimate is 62% higher than that of TAMS (1980), but is in agreement with those based on estimates of Mendis (1977) and Oglesby (1981).

6.1.2.3 Yield from untapped resources

All estimates of fish production from reservoirs have taken only the existing fisheries into consideration, and by and large refer to a monospecific fishery. The preliminary work of Hofer and Schiemer (personal communication) based on experimental fishing on the Parakrama Samudra has indicated the presence of considerable standing stocks of indigenous minor carp, and the author's own personal observations have indicated significant resources of indigenous carnivores, which at present occur in catches only rarely. In two reservoirs the fishery for minor cyprinids which has existed for over a decade has indicated that a fishery for minor carps could exist together with that for the cichlids without apparent detrimental effects on the latter.

The available data, though of a preliminary nature, suggest a yield of 20–50 kg/ha of minor cyprinids and 10–25 kg/ha of carnivores. With the lower level of production the potential yield of this untapped resource would be about 680 t/annum from all the major reservoirs of the lower plains of the Mahaweli Basin.

The importance of these resources has gone unnoticed by a number of missions which dealt with inland fish production in Sri Lanka. The utilization of this resource would involve only small capital expenditure. Although it is questionable whether minor cyprinids would be preferred fresh, they could be dried. There is a market for such dried fish. It could also be made into fish meal which could form the base for the future fish feed industry.

The carnivorous species are highly priced and are preferred fresh. There is no difficulty for marketing them.

6.2 Villus

The manipulation of the Mahaweli Ganga flood peaks and the dry season flow is bound to result in an overall decrease in the floodplain including the villus, which is estimated to become approximately halved to 6,400 ha. The direct effect on the fisheries will be due to the corresponding loss in the floodplain area. Anticipating a yield of 50 kg/ha/annum, TAMS (1980) estimated that the yield from the villus and floodplains would decrease by about 320 t/annum.

In the light of the availability of more reliable production statistics the TAMS (1980) estimates of the potential yield of 320 mt per annum from the villus need to be reconsidered. There is no reason to believe that the reduction of the villu area as a result of the Mahaweli Ganga Basin development would affect their fish yield. The productivity could increase due to the overflow of the excess fertilizer from the increased cultivation in the river basin. Based on the present mean production of 133.0 kg/ha/annum. However, bearing in mind (Table 16) that the mean yield is not a true reflection of the potential due to the present underexploitation of the resource, a realistic yield of 150 kg/ha/annum or 960 t/annum is suggested.

The loss of floodplains is likely to result in the loss of some spawning grounds for most of the indigenous species of commercial importance for the reservoir fishery, particularly Labeo dussumieri, Puntius (= Barbus) sarana, and in changes in the migratory pattern of a number of indigenous forms. The changes observed in the yield of Puntius sarana and Labeo dussumieri for the pre- and post-receipt of Mahaweli waters are shown in Figure 28. There is a drop in their contribution to the fishery and also a shift in the peak periods of their occurrence in catches (see also Table 23, ‘Other’). It cannot be ruled out that this decline is a result of the impact of the Mahaweli Basin development.

6.3 Estuarine and sea fishery

The existing fishing in the estuarine portion of the Mahaweli Ganga is represented primarily by the subsistence fishery for shrimp. The peak shrimp catches occur in March to May. Off the Trincomalee Bay about 150 fishermen operate and the average daily catches range from 20 to 180 kg/day depending on the season.

The tidal influence in Mahaweli Ganga extends about 6 km from the coast. The water regulation and diversion due to the Mahaweli Ganga development scheme is likely to reduce the river discharge by about 50%, both in the dry season as well as during the wet seasons. The reduced flow is likely to result in tidal influence to be felt higher up the estuary, thereby increasing the nursery area available for larval shrimp.

The reduced water discharge and intensity of flood peaks are likely to reduce the nutrient discharge into the coastal area, and to increase the pesticide levels discharged into the sea.

Recently, appreciable populations of whales have been recorded off Trincomalee and a programme is being prepared to commence scientific studies on their biology as well as to make them a tourist attraction. The lower nutrient discharge into the sea might have some effect on the whales through lowering the aquatic productivity of the coastal sea, although it is also possible that nutrient levels will stay relatively high due to the intensive use of fertilizers in the catchment.

Figure 28. Seasonal changes in the indigenous species fishery in Parakrama Samudra (1972-1975): Prior to Mahaweli water input; 1979-1982: Mahaweli water input).

6.4 Aquaculture

The development of the Mahaweli Basin will result in an increase in the number of human settlements and in a population of one million by 1990 in the basin. The recommended animal protein intake is 60 g of fresh fish per day (Anon. 1983). The total animal protein requirement for the population by 1990 would be approximately 21,900 t. Even at the present level of consumption of 15.1 kg/year the requirement would be around 15,000 t/year.

The potential total yield from the capture fisheries of the major reservoirs and villus of the Mahaweli and Maduru Oya basins is estimated to be 9,058 t (Table 29). This estimate has not included the other areas of development of the scheme, although the expected population growth is for all the areas coming under the Accelerated Mahaweli Ganga Development. The major population centres are expected to be located in the Mahaweli Ganga and Maduru Oya basins and will account for approximately 70% of the expected population growth. The capture fishery, therefore, is unlikely to meet the demand and a considerable proportion of the short-fall has to be met by aquaculture.

* Including Pimburettewa and Kantalai reservoirs

6.4.1 Extensive culture systems - Seasonal tanks

The Mahaweli Basin has approximately 4,000 ha of seasonal tanks which dry up for 3–4 months every year. The potential use of these water bodies, most less than 10 ha in surface area, was first identified by Mendis (1965), and reconfirmed by Oglesby (1979). Trials were carried out at the district level to evaluate the potential production of these water bodies during the growing season in 1979/80, 1980/81 and 1981/82. Results of trials in four seasonal tanks of the Mahaweli Basin are summarised in Table 30.

The maximum production in all trials was 1,414.6 kg/ha. The potential of the seasonal tanks to enhance the fish production has been acknowledged (FAO, 1982).

The programme of stocking and harvesting, mostly utilizing the major Chinese and Indian carps, is to be stepped up further throughout Sri Lanka under the Asian Development Bank Programme (FAO, 1982) which commenced in 1984. This project envisages a yield of 15,000 t during the first five years through stocking and harvesting of 20,000 ha of seasonal tanks, a production of 750 kg/ha/annum.

In this programme it is envisaged to maximize fish production by introducing fast growing, late maturing species to fill all the ecological niches available in the water body. The list of species to be stocked includes the Indian carps Catla catla, Labeo rohita, Cirrhina mrigala, and the Chinese carps Aristichthys nobilis, Hypophthalmichthys molitrix and Ctenopharyngodon idella, supplemented by the indigenous Labeo dussumieri and the exotic cichild O. niloticus, if and when needed. All the Indian and Chinese carps and L. dussumieri have been spawned artificially in the breeding stations, some since 1976 (Weerakoon, 1979). The exact quantities of each species to be stocked are to be initially based on the results of already completed trials, on the survival and the growth of individual species, as well as on the interspecific compatibility. These observations will be supported by limnological and biological observations to assist in determining the overall stocking rate as well as the ratio of species.

Under the country programme two of the fisheries stations in the Mahaweli Basin, i.e. the Dambulla and Polonnaruwa Fish Breeding Stations, will be improved. There is also a likelihood of a fish breeding centre being located at Madura Oya Reservoir under the Canadian International Development Agency (Director of Inland Fisheries, personal communication). Under such circumstances it is to be expected that a considerable proportion of the seasonal tanks in the Mahaweli Basin is likely to be utilized for fish production, and a reasonable assumption will be that at least 75% of the 4,000 ha of the available seasonal tank area will be commissioned for fish production by 1988. The expected yield from this resource is around 2,250 t.

6.4.2 Intensive culture

The Buddhist traditions remain one of the main constraints for the development of intensive aquaculture. These traditions are very much preserved in the Mahaweli Basin which has a number of important places of worship for Buddhists and centres of culture for the Sinhalese people.

The likely location of three major centres of fish breeding and the training centre in the Mahaweli Basin are bound to minimize the hindrance to fisheries development brought about by the traditional attitudes of ‘do not rear to kill and eat’. It is expected that a reasonable number of settlers will take up intensive pond culture in the area (Hettiarachchi, 1982), with a possible emphasis on polyculture systems with incorporated poultry rearing.

Rice cum fish culture is unlikely to catch on in Sri Lanka as the varieties of paddy used in Sri Lanka need to retain water only for 2–3 weeks and thereafter the rice paddies need only a moist ground. Hence, paddy cum fish culture will have to evolve with changes in varieties of paddy utilized, which at present seem a very unlikely proposition.

6.4.2.1 Cage culture

The availability for large areas of lacustrine water bodies in Sri Lanka has prompted many people to advocate introduction of cage culture, particularly in the large, perennial reservoirs. The first commercial-scale cage culture operations were undertaken in 1980/81 in two major reservoirs, viz. Parakrama Samudra and Udawalawe. The trials were conducted with high stocking densities of O. niloticus, fed intensively on fish meal based on protein rich diets. The results have shown the feasibility of such operations in reservoirs (Wannigama and Weerakoon, 1982). In spite of the good rate of growth, high survival and good food conversion ratio, the present low market price for O. niloticus and for freshwater fish in general does not appear to make cage culture economically viable. However, a 10% to 15% reduction in feed cost would make it viable, and the present research of the Ministry of Fisheries is aimed in this direction.

Amongst the intensive culture operations, the cage culture offers the best prospects for potential small-scale entrepreneurs. With the increasing emphasis and introduction of incentives for aquaculture, prospects for cage culture look bright.

6.4.2.2 Irrigation channels

It is surprising that hitherto no attempt has been made to utilize for fishery purposes the vast number of irrigation channels already available or which will come into being as a result of the Mahaweli Ganga Basin development. The typical nature of irrigation channels is shown in Figure 29. These channels, mostly 0.5 to 1 m in depth, with their mud banks overgrown with vegetation and with changing water level and current speed provide ideal conditions for crayfish species of the family Astacidae (Bardach et al., 1972). The Sri Lankan freshwater fauna does not have any astacid species (Mendis, 1962) and contains only one freshwater species suitable for culture viz. Macrobrachium rosenbergii. This species, however, needs estuarine conditions to complete its life cycle. Sri Lanka has hitherto not introduced any crustacean species for culture and the possible ecological effects of any suggested introduction will have to be evaluated prior to such introduction.

Figure 29. Irrigation channel.

6.5 Species introductions

Sri Lanka has had one of the highest levels of food fish species introductions in the world relative to its indigenous finfish fauna (Table 3). Nevertheless, there is no evidence to believe that any of these introductions have been harmful either ecologically or biologically, except perhaps the accidental introduction of the cestode Bothriocephalus gowkonensis (Fernando and Furtado, 1963). The introduction of O. mossam-bicus has led to the development of viable freshwater fisheries, with this species forming the backbone of it. It is interesting to note that most of the introductions which were haphazardly carried out have been unsuccessful and remain as mere book records. Others, notably the major Indian and Chinese carps, are destined to play a major role in the seasonal tank and pond culture programmes, and their potential in the former in particular is already evident from the trials carried out throughout Sri Lanka (Thayaparan, 1982).

6.5.1 Reservoir fishery

Fernando (1965, 1971) advocated the introduction of deep-water herbivorous and carnivorous fish, particularly into the mid- and hill-country reservoirs. More recently, views have been expressed to introduce a zooplankton feeder, to fill in the vacant niche, in low country reservoirs. In this context suggestions have been made that the African clupeids Limnothrissa sp. and/or Stolothrissa sp. could be suitable candidates. The omnivorous Limnothissa was successfully introduced into the African natural lake Kivu as well as into the large reservoir Kariba, from where it has spread downstream into Cahora Bassa reservoir (Petr and Kapetsky, 1983). Stolothrissa, which feeds predominantly on zooplankton, was not successful. Such introductions into Sri Lankan reservoirs could be made only on the basic assumption that there is a sufficient food supply available. Unlike African lakes and reservoirs, which are considerably larger than those in Sri Lanka, the Sri Lankan reservoirs are also characterised by a short retention time and poor zooplankton which represents part of the diet of Limnothrissa, but much of the diet of Stolothrissa. The latter species would probably also fail in most of the shallow Sri Lankan reservoirs, as it prefers the off-shore deep waters of its natural habitat of Lake Tanganyika.

The recent appearance of the halfbeak (Hyporhamphus gaimardi) in some low country reservoirs, including two in the Mahaweli Basin, in sufficient quantity to support 2 to 3 full-time fishermen, may well be the first indication of utilization of the vacant ‘zooplankton niche’ by an indigenous, but a non-specialized zooplankton feeder in that H. gajmardi feeds on zooplankton, supplemented by a considerable proportion of terrestrial insects. There is also evidence that this species, although typically coastal and estuarine, is breeding in certain reservoirs. If this is true, there is no reason for the introduction of an exotic zooplankton feeder such as an African clupeid.

Amongst the candidate fish suggested for introduction is O. aureus, possibly based on the success of its counterpart, O. mossambicus. More recently suggested candidates are T. sparrmanni, an omnivore and S. galilaeus, a phytoplankton feeder (TAMS, 1980).

In the author's view it will need further biological and ecological investigations before any more introductions of finfish species are made to Sri Lanka. There is no reason to believe that even the introduction of suitable species into the deep hill-country reservoirs in the Mahaweli Basin will result in a viable fishery.

6.6 Research opportunities

The Mahaweli Development is destined to result in the impoundment of 9 major reservoirs, whose characteristics vary in terms of topography, climate and geological formations. The development scheme is also destined to bring about changes in the hydrological regimes of older reservoirs which range in age from nearly 1,000 years to 25 years.

This development scheme therefore offers a system of reservoirs which could be studied from the first day of their impoundment from limnological and fisheries point of view. They offer an ideal opportunity to follow the development of fish populations, both introduced as well as naturally recruited. The close proximity of the reservoirs makes such a comparative study logistically feasible.

So far there has been no initiative, either by the agencies of the government of Sri Lanka or international organizations to launch such a limnological and fishery biological study of the reservoir system. Undoubtedly, isolated pieces of research on these aspects will be carried out by individuals and/or institutions on a small scale, which is unlikely to give the same coherence and understanding of the system as a whole. This ideal opportunity, if lost, will be hard to come by again and will be a loss to the advancement of tropical limnology and fisheries biology and management.

7. CONCLUSIONS

Although the primary purpose of the Mahaweli river basin development and associated construction is not fisheries development, the incorporation of fisheries development in such strategic schemes would be most beneficial as fisheries may provide considerable returns. A case in point is the Ubolratana Reservoir in Thailand, where the return from the fishery surpasses that from electricity generation, although the reservoir was primarily constructed for the latter purpose (see Bhukaswan, in this Volume). The original Mahaweli Ganga Diversion and Development plans did not incorporate the fisheries sector, and the expected returns from this sector were not taken into consideration in the cost-benefit analysis. Nor were they taken into account in the settlement planning which is crucial and pivotal for the proper functioning of the scheme. However, under the Accelerated Mahaweli Development Programme the developmental strategies of the fisheries sector have been taken into consideration under an Environmental Assessment Programme. Some recommendations such as to retain the vegetation of a proportion of the reservoir bed to alleviate the fish production were not heeded. Although the yield from the various water bodies was computed, the results appeared to be not accurate enough for reliable predictions. The Environmental Assessment Programme stressed the need to increase the aquacultural practices.

The present report provides new yield estimates for individual water bodies and suggests possible ways of increasing fish production, in addition to the traditional - mostly monospecific - reservoir fishery which is bound to develop in the new reservoirs. Exploitation of non-traditional varieties, particularly the minor indigenous cyprinids and indigenous carnivorous species, should produce a yield of at least 650 t/annum. The success of such a fishery in the Mahaweli Basin would undoubtedly serve as an example which would be followed in reservoirs elsewhere in Sri Lanka.

While the emphasis, on aquaculture development has hitherto concentrated on pond and cage culture, the vast system of irrigation channels which will come into being as a result of the Mahaweli Ganga Development, as well as other river basin development schemes, is still waiting to be assessed. This water resource could well provide a base for an extensive culture of highly priced freshwater crayfish, aimed for the export market.

Introductions of new finfish species are not recommended as the existing species, both exotic and indigenous, should provide a sufficient species diversity for the development and maximization of the inland fishery.

It is also suggested that in river basin development due consideration needs to be given to the settlement pattern, in order to maximize the return from the fisheries sector, especially from reservoirs. A delay in acquisition of the expertise in the art of fishing by settlers who are mostly farmers, and are likely to take up fishing initially at a subsistance level, may result in under-fishing of the peak fish production to be expected to be achieved during the first few years after damming the river.

With the expected development of intensive fish culture practices and reservoir fishery in the Mahaweli Basin the available water resources are likely to meet nearly 70% of the overall recommended animal protein requirements of 21.2 kg/annum/person of the population of the Mahaweli Basin by 1990.

ACKNOWLEDGEMENTS

I am most grateful to the Director of Inland Fisheries, and the Officer-in-Charge of the Polonnaruwa Station, Mr. A. Hettiarachchi, for providing me with data on the existing fishery and to many others of this Division for giving their time to discuss the problems of the fishery. My thanks are also due to Mr. M.A. Jansen, Environmental Officer, Mahaweli Development Authority for providing me with some of the most relevant literature on the Mahaweli Ganga Development Scheme; to Mr. P. Maitipe of the Department of Zoology for making the drawings; to Ms. P.A. Wickremaratne for typing the manuscript; to Mr. B. Abeysundera for his valuable comments on some of the chapters, and finally to my wife, Dr. Celine De Silva for her invaluable criticism.

8. REFERENCES

Arumugam, S., 1969 Water resources of Ceylon. Colombo, Water Resources Board Publication, 407 p.

Bardach, J.E., J.H. Ryther and W.O. McLarney, 1972 Aquaculture: the farming and husbandry of freshwater and marine organisms. London, Wiley-Interscience, 868 p.

Bauer, K., 1983 Thermal stratification, mixis and addective currents in the Parakrama Samudra Reservoir, Sri Lanka. In Limnology of Parakrama Samudra, edited by F. Schiemer. The Hague, The Netherlands, W. Junk Publishers, pp. 27–34

Beam, J., 1983 The effect of annual water level management on population trends of white crappie in Elk City reservoir. Fish.Manage., 3:34–40

Bhukaswan, T., 1980 Management of Asian reservoir fisheries. FAO Fish.Tech. Pap., (207):69 p.

Brohier, R.L., 1934 Ancient irrigation works in Ceylon. Vol. I. Colombo, Ceylon Government Press, 37 p.

Brohier, R.L., 1935. Ancient irrigation works in Ceylon. Vol 2. Colombo, Ceylon Government Press, 43 p.

Central Bank of Ceylon, 1979 Statistical abstracts. Colombo. 316 p.

Ceylon Association of Science, n.d. A survey of potable and industrial waters in Ceylon. Presidential address. Colombo, 28 p.

Ceylon, 1969 Department of Census and Statistics, The Ceylon economic atlas. Colombo, Department of Census and Statistics, 69 p.

Chandrasoma, D. and S.S. De Silva, 1981 Reproduction biology of Puntius sarana an indigenous species and Tilapia rendalli, in an ancient, man-made lake in Sri Lanka. Fish Manage., 12:17–28

Cooray, P.G., 1967 An introduction to the geology of Ceylon. Spolia Zeylan., 31:324 p.

Darlington, P.J., 1957 The geographical distribution of animals. New York, John Wiley, 675 p.

De Silva, S.S., 1983 Reproductive strategies of some major fish species in Parakrama Samudra Reservoir and their possible impact on the ecosystem - a theoretical consideration. In Limnology of Parakrama Samudra, edited by F. Schiemer. The Hague, The Netherlands, Dr. W. Junk Publishers, pp. 183–91

De Silva, S.S., Impact of exotics on the inland fishery resources of Sri Lanka. In Proceedings of the Workshop on Tropical Limnology (Asia/Pacific) and Water Quality, University of Malaya, Kuala Lumpur, Malaysia, Nov. 1982 (in press)

De Silva, S.S., Status of the introduced cichlid Sarotherodon mossambicus (Peters) in the reservoir fishery of Sri Lanka: a management strategy and ecological implications. Aquacult.Fish.Manage., 1:91–102 (in press a)

De Silva, S.S., Observations on the abundance of the exotic cichlid, Sarotherodon mossambicus (Peters) in relation to water level fluctuation in a man-made lake in Sri Lanka. Aquacult.Fish.Manage., (in press b)

De Silva, S.S. and C.D. De Silva, Coastal lagoons. In Ecology and bio-geography of Sri Lanka, edited by C.H. Fernando. The Hague, The Netherlands, Dr. W. Junk Publications (in press)

De Silva, S.S. and C.H. Fernando, 1980 Recent trends in the fishery of Parakrama Samudra, an ancient man-made lake in,Sri Lanka. In Ecology and development, edited by J.I. Furtado. Kuala Lumpur, Malaysia, Univeristy of Malaya Press, pp. 927–37

De Silva, S.S., P. Maitipe and P.R.T. Cumaranatunga, Aspects of the biology of the euryhaline Asian cichlid, Etroplus suratensis. Environ.Biol.Fish, (in press)

Domros, M., 1976 Sri Lanka. Die Tropeninsel Ceylon. Darmstadt, Wissenschaft-liche Buchgesellschaft, 298 pp.

Dudley, R.C., 1979 Changes in the growth and size distribution of Sarotherodon macrochir and Sarotherodon andersoni from the Kafue flood plains. J.Fish Biol., 14:205–23

Ellepola, W.B. and C.H. Fernando, 1968 Hatching and rearing of gouramy (osphro nemus goramy, Lacepede), in the Polonnaruwa nursery. Bull. Fish.Res.Stn.,Colombo, (19):1–3

Evans, D., 1981 Threatened freshwater fish of Sri Lanka. Cambridge, U.K., IUCN Conservation Monitoring Centre, 59 p.

FAO, 1981 Technical Cooperation Programme, Sri Lanka. A synthesis of the reservoir fisheries in Sri Lanka. A report for the Project TCP/SRL/8804 Development of Fisheries in the man-made lakes and reservoirs. Based on the work of R.T. Oglesby, Fisheries Biologist. Rome, FAO, FI:TCP/SRL/8804, Field document, 2:30 p.

FAO, 1982 Report of the Sri Lanka aquaculture development project preparation mission (FAO/ASDB Programme). Report 2/82AS-SRL 24, 34 p.(mimeo)

Fernando, A.D.N., 1973 The ground water resources of Sri Lanka. Colombo, Ministry of Irrigation, Power and Highways

Fernando, C.H., 1965 The development of Ceylon's fisheries. 11. The role of inland waters in relation to the development of Ceylon's fisheries; and a note on the pearl oyster fishery. Bull.Fish. Res.Stn., Colombo, 17:291–7

Fernando, C.H., 1967 The beach seine fishery in Ceylon's freshwaters. IPFC Curr. Aff.Bull., (50):1–21

Fernando, C.H., 1971 The role of introduced fish in fish production in Ceylon's freshwaters. In The scientific management of animal and plant communities for conservation, edited by E. Duffey and A.S. Watt. Oxford, Blackwell Scientific Publications, pp. 295–310

Fernando, C.H., 1977 Reservoir fisheries in South East Asia: past, present and future. Proc.IPFC, 17(3):475–80

Fernando, C.H. and S.S. De Silva, 1984 Man-made lakes; ancient heritage and modern biological resource. In Ecology and biogeography of Sri Lanka, edited by C.H. Fernando. The Hague, The Netherlands, Dr. W. Junk Publishers

Fernando, C.H. and J.I. Furtado, 1963 A study of some helminth parasites of freshwater fishes of Ceylon. Z.Parasitenkd., 23:141–63

Fernando, C.H. and H.H.I. Indrasena, 1969 The freshwater fisheries of Ceylon. Bull.Fish.Res.Stn., Colombo, (20):101–34

Ganapati, S.V., 1973 Man-made lakes in South India. In Man-made lakes: their problems and environmental affects, edited by N.C. Ackermann, G.B. White and E.B. Worthington. Geophys.Monogr.Ser., (17): 63–73

Gunatilaka, A., 1983 Phosphorous and phosphatase dynamics in Parakrama Samudra based on diurnal observations. In Limnology of Parakrama Samudra, edited by F. Schiemer. The Hague, The Netherlands, Dr. W. Junk Publishers, pp. 35–47

Gunatilaka, W.D. and C. Senaratne, 1981 Parakrama Samudra (Sri Lanka) Project a study of a tropical lake ecosystem. 2. Chemical environment with special reference to nutrients. Verh.Int.Ver.Theor.Angew. Limnol., 21:1000–6

Gwahaba, J.J., 1973 Effects of fishing on the Tilapia nilotica (Linne 1957) population in Lake George, Uganda over the past 20 years. E.Afr.Wildl.J., 11:317–28

Hettiarachchi, A., 1982 Efforts at raising fish in private ponds in the Polonnaruwa District. J.Inland Fish.,Sri Lanka, 1:115–8

Hunting Survey Corporation, 1962 A report on a survey of the resources of the Mahaweli Ganga Basin, Ceylon. Part 1. Colombo, Ceylon Government Press, 252 p.

Indrasena, H.H.A., 1965 Fresh water fishery resources. Bull.Fish.Res.Stn., Colombo, (18):87–90

Jhingran, V.G. and V. Gopalakrishnan, 1974 Catalogue of cultivated organisms. FAO Fish.Tech.Pap., (130):83 p.

Jones, G.W. and S. Selvaratnam, 1972 Population growth and economic development in Ceylon. Colombo, Ceylon, Hansa Publishers Ltd., 249 p.

Marshall, B.E., 1982 The influence of river flow on pelagic sardine catches in Lake Kariba. J.Fish Biol., 20:465–9

Mendis, A.S., 1962 A guide to the freshwater fauna of Ceylon. Bull.Fish.Res. Stn., Colombo, (12):160 p.

Mendis, A.S., 1965 A preliminary study of twentyone Ceylon lakes. 2. Limnology and fish production potential. Bull.Fish.Res.Stn., Colombo, (18):7–16

Mendis, A.S., 1977 The role of man-made lakes in the development of fisheries in Sri Lanka. Proc.IPFC, 17(3):247–54

Moorman, F.R. and C.R. Panabokke, 1961 Soils of Ceylon: a new approach to the identification and classification of the soils of Ceylon. Trop.Agric., 117(1)

NEDECO, 1979 Mahaweli Ganga development programme implementation strategy study. Annexes A to G. Colombo, Mahaweli Development Board and NEDECO

Oglesby, R.T., 1977 Relationship of fish yield to lake phytoplankton standing crop, production and morphoedaphic factors. J.Fish.Res.Board Can., 34(12):2271–9

Oglesby, R.T., 1979 Inland fisheries of Sri Lanka and their management. Interim report submitted to FAO, Rome, September 1979, 39p. and Appendices (mimeo)

Panabokke, C.R., 1967 Soil science, the soils of Ceylon and use of fertilizers. Colombo

Pattiarachchi, D.B. and N.R. De Silva, 1972 Some aspects of water quality in Sri Lanka. In Proceedings of the International Hydrological Decade, National Committee of Sri Lanka. Colombo, Geological Survey Department

Petr,T. and J.M. Kapetsky, 1983 Pelagic fish and fisheries of tropical and subtropical natural lakes and reservoirs. ICLARM Newsl., (6): 9–13

Ryder, R.A., 1965 A method for estimating the potential fish production of north temperate lakes. Trans.Am.Fish.Soc., 94:214–8

Ryder, R.A., 1982 The morphoedaphic index - use, abuse and fundamental concepts. Trans.Am.Fish.Soc., 111:154–64

Schiemer, F., 1983 The Parakrama Samudra project - scope and objectives. In Limnology of Parakrama Samudra, edited by F. Schiemer. The Hague, The Netherlands, Dr. W. Junk Publishers, pp. 1–18

Schiemer, F. and A. Duncan, 1983 Parakrama Samudra project - a summary of main results. In Limnology of Parakrama Samudra, edited by F. Schiemer. The Hague, The Netherlands, Dr. W. Junk Publishers, pp. 210–8

Schiemer, F. and R. Hofer, 1983 A contribution to the ecology of the fish fauna of the Parakrama Samudra. In Limnology of Parakrama Samudra, edited by F. Schiemer. The Hague, The Netherlands, Dr. W. Junk Publishers, pp. 134–54

Schnitter, N.J., 1967 A short history of dam engineering. Water Power, 19: 142–8

Senanayake, F.R. and P.B. Moyle, 1981 Conservation of the freshwater fishes of Sri Lanka. Biol.Conserv., 22:181–95

Sri Lanka, 1960 Department of Fisheries. Training course for freshwater fishery officers. Manuscr.Rep.Dep.Fish.Colombo, (20):69 p. (mimeo)

Sri Lanka, 1979 Ministry of Land and Land Development and Mahaweli Development, Mahaweli projects and programmes. Sri Lanka, Colombo, Department of Government Printing, 124 p.

Sri Lanka, 1980 Ministry of Fisheries, Master plan for the development of fisheries in Sri Lanka. Colombo, Sri Lanka, Ministry of Fisheries, 38 p.

Sri Lanka, 1983 Ministry of Land and Land Development and Mahaweli Development. Mahaweli projects and programmes. Colombo. 104 p.

TAMS (Tippetts-Abbett-McCarthy-Stratton), 1980 Environmental assessment. Accelerated Mahaweli Development Programme. Vols. 1,2 and 3. New York, TAMS

Thambaiahpillay, G., 1959 Tropical cyclones and the climate of Ceylon. Univ. Ceylon Rev., 17(3–4)

Thayaparan, K., 1982 The role of seasonal tanks in the development of fresh-water fisheries in Sri Lanka. J.Inland Fish., Sri Lanka, 1: 133–67

UNDP/FAO, 1969 Master Plan (1965–1969). Mahaweli Ganga Irrigation and Hydropower Survey, Ceylon.

U.S. Operation Mission (USOM), 1961 Mahaweli Ganga basin diversion scheme. Volume 3. Reconnaissance land classification and soil survey report. Colombo, Sri Lanka, Irrigation Department, 80 p.

Wannigama, N.D. and D.E.M. Weerakoon, 1982 Studies on cage culture of Sarotherodon niloticus in man-made lakes in Sri Lanka. J.Inland Fish.,Sri Lanka, 1:67–85

Weerakoon, D.E.M., 1979 Induced spawning of two major species of Chinese carps, Ctenopharyngodon idellus and Aristichthys nobilis in Sri Lanka. Bull.Fish.Res.Stn.Colombo, (29):55–62

Weeraratne, C.S., 1983, Pesticides - an overview with particular reference to Sri Lanka. Econ.Rev., 8(10):4–12

Welcomme, R.L., 1967 Observations on the biology of the introduced species of Tilapia in Lake Victoria. Rev.Zool.Bot.Afr., 76:249–79

Welcomme, R.L. and D. Hagborg, 1977 Towards a model of a floodplain fish population and its fishery. Environ.Biol.Fish, 2:7–24

Welcomme, R.L. and H.F. Henderson, 1976 Aspects of the management of inland waters for fisheries. FAO Fish.Tech.Pap., (161):36 p. Issued also in French and Spanish.

Wijeyaratne, M.S.J. and H.H. Costa, 1981 Stocking rate estimations of Tilapia mossambica fingerlings for some inland reservoirs of Sri Lanka. Int.Rev.Gesamt.Hydrobiol., 66:327–33

Willey, A., 1910 Notes on the freshwater fisheries of Ceylon. Spolia Zeylan., 7:88–105

Anon, 1978 The Mahaweli project. Econ.Rev., 4(8–9):4–22

Anon, 1983 Fisheries. Econ.Rev., 9(3):3–16