Previous Page Table of Contents Next Page


A.I. Payne
Renewable Resources Assessment Group,
Imperial College Centre for Environmental Technology,
8 Prince's Gardens, London SW7 INA, U.K.
V. Cowan
Marine Resources Assessment Group Ltd.,
27, Campiden Street, London W8 7EP, U.K.


There must be a rationale for the enhancement of fisheries stocks. In Bangladesh there is a strong biological and social rationale for enhancements. This paper covers the results achieved and the lessons learned from large stock enhancement projects in Bangladesh including the Third Fisheries Project, ADB Second Aquaculture Project and the Oxbow Lake Project. The benefits of habitat restoration also are considered. Important conclusions are that stocking must be based on a cost-effective mass production, management and monitoring is necessary to establish benefits, more systematic work is required to devise management models and stocking is not the unique enhancement capable of providing benefits.


There must be a rationale for the enhancement of fisheries stocks. It is unrealistic and wasteful merely to introduce large numbers of eggs or young fish into a water body and expect an automatic increase in catches. There must be a vacant niche, a shortfall in recruitment or an increase in resources for enhancement to have any prospect of success.

Within Bangladesh there is a strong rationale. There is the perception that the major carps, that is rui (Labeo rohita), catla (Catla catla) and mrigal (Cirrhinus mrigala), had greatly declined in the catches. These species are all microphagous or detritivores of the kind normally expected to be abundant on floodplains (Bowen, 1983). Earlier catch records in part of the basin during the 1950s showed major carps comprising up to 47% of the catch at sites where now they only represent 3% (Payne and Temple, 1996). This has been due to a number of factors including over-exploitation, water regulation and pollution. A potential gap has, therefore, been opened up amongst the niches of the natural community. It is also true to say that the major carps are the most favoured fish across the sub-continent. With regard to resources, Bangladesh is well endowed from the viewpoint of inland fisheries. During the flood periods of the Padma (Ganges) and Jamuna (Brahmaputra) rivers, between 30–45% of the country is inundated floodplain. Increasingly, however, the floodplain has become empoldered and compartmentalised for flood protection and agriculture. Most of these compartments still flood but the regulatory mechanisms inhibit movement and recruitment by major carps from the main channel. In addition, natural recruitment of major carps has also been reduced by the widespread capture of eggs and fry along the banks of the main channels, in order to stock the extensive pond aquaculture developments in the country.

There are a number of clear criteria, therefore, why stock enhancement is of potential value in Bangladesh. To these can be added one more, and that is the social need. More than 13 million people live on the floodplains of Bangladesh who benefit in some way from fishing. The increasing number of landless people means that more people are coming to rely on fishing as their main livelihood (Temple and Payne, 1956). Considerable means may then be justified in helping to sustain these people in employment.

Not surprisingly, therefore, Bangladesh has been in the forefront of the development of stock enhancement procedures through a number of government and international projects. The Department of Fisheries has conducted its own revenue stocking programme for almost a decade. Other agencies have co-operated with the government in large-scale projects, such as the World Bank/UNDP/ODA Third Fisheries Project; ADB Second Aquaculture Project; and the World Bank and Danida on the Oxbow Lake Project, phases one and two respectively. All have been based upon introduction of young major carp, although with some admixture of Chinese and common carps.

There is, however, one critical feature which has enabled all of these large-scale stock enhancement projects to take place. Each has required many millions of carp juveniles. This only becomes feasible when a cost-effective method of mass production of fingerlings is possible. In Bangladesh this has been achieved by the spectacular spread of low technology mini-hatcheries throughout towns and villages across the country, albeit with a central focus around Jessore. The mini-hatcheries were originally developed under the Oxbow Lake Phase 1 Project, but spread rapidly as the people of Bangladesh appreciated their value. There is now a buoyant private sector hatchery and nursery industry which supplies not only the stocking projects, but also a highly successful aquaculture industry.

This underlines a very simple principle. Large-scale direct fish stock enhancement can only be a practical possibility if a cost-effective means of mass production exists. On this criterion, very few countries can realistically consider this option.

The lessons provided by the large stock enhancement projects in Bangladesh are outlined below but in the light of the above constraints alternative strategies, such as habitat restoration and rehabilitation, need also to be considered. Bangladesh can also provide useful insights on this.


Originally this project was conceived as restocking 100,000 ha of floodplains with a mixture of major carp, silver carp, bighead and common carp of 7–10 cm length, at 20 kg/ha (World Bank, 1990). The target was to be approached progressively over a 5-year period. Each year, the floodplain areas selected were stocked by contractors to tender specifications in June, as the floods rose. The fingerlings were protected for three months through an awareness campaign, persuading the fishing communities not to catch them during this period pending greater incremental benefits, and then allowing harvesting from October on as the flood waters subsided. Detailed inspection, reception and cross-checking procedures had to be used to ensure fulfilment of contracts. Over the 5-year period more than 50 million fingerlings were stocked, thereby emphasising the need for cost-effective mass production. The key indicator of the effectiveness of the process is the multiple increase in weight harvested, compared to weight of fingerlings introduced (Table 1). This finally averaged at around 10-fold increase (Table 1). There was considerable variability within these compiled multiples, however. Generally smaller floodplains of around 2000 ha produced higher increments than larger floodplains, sometimes with multiple increments of more than 30. Of the original three large floodplains stocked, one was entirely empoldered for agriculture, one was entirely open and one was semi-enclosed. In fact, the empoldered beel performed quite well showing that compartmentalised areas, although essentially for rice culture, still provide the resources for a fish crop. The semi-enclosed floodplain gave very poor results. This was because the water was very turbid and the dense aquatic macrophyte cover typical of most floodplains in Bangladesh did not appear. There was, effectively, no detrital food chain.

Table 1. Results of fingerling stocking of floodplains. Third Fisheries Project, Bangladesh.

 Area stocked haStocking density kg/haTotal stocking quantity kgProduction mtWeight increase
PY 1 (1991)3,70019.773,0005969.5
PY 2 (1992)23,20017.5405,5907331.8
PY 3 (1993)24,70017.5432,4762,9116.7
PY 4 (1994)23,90018.0430,7414,64010.8
PY 5 (1995)33,90017.7601,5975,2228.7
PY 6 (1996)44,90014.7661,031**6,610**10.0

** estimated

Source: Implementation Completion Report, Floodplain Stocking Component
Department of Fisheries and MTA, December 1996

There was also considerable year-to-year variation in production. This is to be anticipated in any floodplain system since production is always dependent upon the flood, which itself varies from year to year (Welcomme and Hagborg, 1977). Indeed a significant positive relationship between catch, rainfall and flood has been demonstrated for the Bangladesh floodplains (Payne and Temple, 1996). This always renders the establishment of baseline production very difficult. For example, total annual yields from the open floodplain varied between 300–800 kg/ha, of which around 75% were wild fish and the remainder stocked carp. Production of wild and stocked species tended to respond in the same way to variations in the annual flood. The poor performance in 1992 (Table 1) was due to a 1 in 25 years drought in that year.

Considerable differences were also found from species to species. Initially, following the example of the Oxbow Lake Project, over half the tonnage put into the water was to have been silver carp. These, however, did very badly due to a combination of lack of phytoplankton development in the floodplains, which is their prime food source, and their highly migratory behavioural pattern which was more marked than in other carps. By contrast, common carp often did very well with increments in excess of 30 times being recorded and individuals of 2 kg appearing after 4 months in the floodplain. They were obviously able to exploit the conditions under the macrophyte cover to an exceptional extent.

One feature of the Third Fisheries Project was the insistence by the donors on an intensive monitoring system in order that a realistic cost-benefit analysis could be carried out. Over all floodplains over all years the average internal rate of return was 17%, well above that required by the World Bank for viable investment. Clearly, fingerling cost is a major element in this, but the procedure remains economically viable.

A rigorous analysis has shown there are benefits to be had for the enhancement procedures of this project. In fact, if the 20,000 mt incremental production (Table 1) is divided by the annual per capita fish consumption in Bangladesh, this amount of fish will provide 2.7 million people with their fish supply for a year. Nevertheless, within the hierarchical society of rural Bangladesh the question arose as to how fairly the benefits may have been distributed. It is true to say that the social indicators of the project were inadequately developed to judge this. It became clear that NGO intervention was generally required to ensure that the fishing communities were adequately organised to gain the full benefit. Increased organisation also led the way to improved financial management by the fishing groups, and presented the prospect of the community paying for and managing the stocking themselves. By the end of the project a small amount of cost recovery was being achieved. This experience clearly demonstrates that the social dimension of such a project is at least equally important as the production element.

Direct stock enhancement is essentially an expensive process. It is therefore important to optimise all the parameters in order to cut costs and maximise benefits. In reservoir stocking, for example, Lorenzen (1995) has shown that final production is a function of size at first stocking and stocking density, both of which are price sensitive. Similar relationships and models are required to render floodplain stocking more efficient. There is also the question of species composition. There is no one single recipe but a categorisation of types of floodplain could lead to a number of defined species mixtures for use by managers.


The essential difference between this project and the Third Fisheries Project was in the way the young fish were put into the water. Rather than being put in directly as well-grown fingerlings, within the Second Aquaculture Project nursery ponds were prepared directly on the floodplain, in areas of residual water, and fry were placed in these ponds at a few days old (DOF, 1995). The fry are thus conditioned and start to grow, then as the flood waters rise and overtop the ponds, the young are released naturally. Unfortunately this procedure had to be abandoned in the main, because the chemical treatments necessary to keep down predators and parasites within the ponds in the open floodplain were considered to be environmentally damaging on a wide scale. Ultimately the project adopted a procedure similar to that of Third Fisheries with similar results.


The oxbow lakes, or boars as they are known locally, exist on the perimeter of the Bangladesh floodplain in the south-west of the country. They are remnants of old river channels which now are rarely, if ever, flooded. They are essentially lakes. As such, silver carp have proved to be very successful, unlike the experience on the floodplains. Some major carp and grass carp are also used. Stocking has been going on for over a decade now. Areas are generally smaller than in the floodplains, with lakes of several hundred hectares, rather than several thousand, as in Third Fisheries. Stocking is by direct introduction at rates of 2,500–4,000/ha. Fish yields and phytoplankton density have been associated through a simple direct correlation and those lakes with clear water have proved to be poor. The oxbow lakes are clearly a different ecological system to the floodplains. Overall carp yields have ranged from 137–565 kg/ha with a mean of 520 kg/ha.


There is one documented example of the benefits of habitat restoration in Bangladesh. The community on Singharagi Beel in the Dhaleswari River in central Bangladesh perceived that fishing had declined due to siltation of the canal linking the floodplain or beel to the main river channel. The rationale here is that siltation was inhibiting recruitment into the floodplain from the main river. Here, NGO intervention (CNRS, 1995; Rahman et al 1996) not only helped to organise the community for the excavation of the canal but in addition, assisted with agreements on fishing patterns within the community following excavation. They were also able to organise effective monitoring, a key element in demonstrating benefit.

The outcome was that yields increased from 1863 kg/ha to 11,384 kg/ha, which partly resulted from improved access to fishing with the number of fishing days having been increased from 396 to 810 over a year (CNRS, 1995). Most significantly, the percentage of the catch due to major carps and large catfishes, that is, those migratory main channel species, increased from 2% to 24%, indicating that immigration and recruitment had increased substantially following the intervention.

It is this type of perceptive action in habitat improvement which may be more widely applicable than direct stock enhancement in many cases.


There are a number of fundamental conclusions which can be drawn from the Bangladesh experience.

  1. There must be a biological rationale for the process to be effective.

  2. Direct stocking must be based on a cost-effective means of mass production of young.

  3. Stocking is expensive and therefore needs careful management and monitoring to establish benefits.

  4. On floodplains more systematic work is required to devise management models to optimise parameters and reduce costs.

  5. There must be a social rationale and much depends on the organisation of the people to take full share in the benefits and also to take ownership of the process.

  6. Direct stocking is not the only answer; referring back to the biological rationale, habitat restoration may offer a more widely applicable means of making more resources available or improving recruitment of floodplain fish stocks.


Bowen, S.H. 1983. Detritivory in neo-tropical fish communities. Env. Biol Fish 9: 137–144.

CNRS. 1995. Community-Based Fishing Management and Habitat Restoration Project: Annual Report (July 1994-June 1995).

DOF, Bangladesh. 1995. An evaluation of 1994 floodplain stocking. Internal Project Report, Second Aquaculture Development Project. Department of Fisheries, Dhaka.

DOF, Bangladesh/MTI. 1996. Implementation Completion Report, Floodplain Stocking Component. Internal Project Report, Third Fisheries Project. Department of Fisheries, Dhaka.

Lorenzen, K. 1995. Population dynamics and management of culture-based fisheries. Fisheries Management and Ecology 2: 61–73.

Payne, A. I. and S.A. Temple. 1996. River and Floodplain Fisheries in the Ganges Basin: Final Report. Overseas Development Administration, Fisheries Management Science Programme.

Rahman, M., S. Halder and D. Capistrano. 1996. Community-based wetland habitat restoration and management: experiences and insights from Bangladesh. In: Sixth Annual Conference of the International Association for the Study of Common Property, 5–8 June 1996, Berkeley, California, USA.

Temple, S.A. and A.I. Payne. 1995. The Ganges Basin, an overview for fisheries. Overseas Development Administration, Fisheries Management Science Programme.

Welcomme, R. L. and D. Hagborg 1977. Towards a model of floodplain fish populations and its fishery. Env. Biol. Fish 2: 7–22.

World Bank. 1990. Staff Appraisal Report for Bangladesh Third Fisheries Project. World Bank, Washington, USA.

Previous Page Top of Page Next Page