Fish Farmers Contribute To Conservation Of Endangered Lake Species

Harrison Charo-Karisa from Kenya's KMFRI Sagana Aquaculture Research Centre, and Jacob Maithya from Moi University report on fish farmers&rsquo' contribution to the conservation of endangered Lake Victoria Basin fish species.


Freshwater ecosystems have suffered greatly from interference from human activities over the past 100 years rendering many fish species extinct, rare or endangered. The need for conservation action is therefore paramount (Cowx and Collares-Pareira, 2002). Both the Mbiru (Oreochromis variabilis (Boulenger)) and Ngege (Oreochromis esculentus (Graham)) are presently listed in the IUCN Red List of Threatened Species as critically endangered (see IUCN, 2000). Both fish species are endemic to Lake Victoria and its satellite lakes. Their disappearance from commercial and subsistence landings of the Lake Victoria fishery has been explained by several authors as due to predation by Nile perch (Lates niloticus (L.)), an introduced predatory species, competition for food and nesting sites from Nile tilapia (Oreochromis niloticus,), another introduced species, environmental degradation; including pollution, over-fishing and destructive fishing methods (Lowe-McConnell, 1982; Ogutu-Ohwayo, 1990; Greboval and Mannini, 1992).

Before L. niloticus, O. niloticus and Tilapia zillii (Gervais) were introduced into Lake Victoria, the majority of the ichthyofauna of the lake was made up of haplochromines and tilapiines. From as early as 1908, an organized gillnet fishery introduced in Nyanza Gulf three years earlier had intensified, putting a lot of pressure on the native tilapiines, O. esculentus and O. variabilis (Graham, 1929), which were the only large species of commercial value. In less than 20 years, several hundred species had disappeared from Lake Victoria (Witte et al., 1992). The Nile perch upsurge coincided with the eventual total disappearance of these species from the main lake.

Until the mid-1960s, Lake Victoria contained large stocks of O. esculentus of relatively large sizes. In the earlier surveys, Graham (1929) recorded O.esculentus of 30 cm total length (TL) in Kavirondo Gulf and 31 cm TL in the open lake and 40–50 cm TL in the southern waters of the lake. Greenwood (1966) continued to record modal adult size 30–32 cm TL for O. esculentus in the lake. Due to increased fishing pressure the fish became smaller and the catches, especially of O. esculentus, which was the most targeted species, reduced (Fryer and Iles, 1972).

Tilapias are well known for their abilities to hybridize in the natural environment when native species are in contact with introduced species (Agnèse et al., 1998). Hybridization with O. niloticus is considered as one avenue in which populations of the two species suffered genetic modification and disappearance from the lake (Mwanja and Kaufman, 1995). Welcomme (1967) observed hybrids between O. niloticus and O. variabilis before the latter species disappeared from the lake. It also seems likely that O. esculentus hybridized with O. niloticus, which is presently the most dominant tilapiine species in Lake Victoria. Thus, the best way of conserving these species may be to produce them as pure species in an environment ex situ away from the predatory L. niloticus and the closely related O. niloticus.


Breeding for restocking and maintenance of genetic variation

Long-term success of any fish species in aquaculture depends on the existence of genetically diverse stocks, since long-term adaptability of populations is dependent upon a sufficiently wide base of genetic variation with which to respond to environmental or biotic change (Fisher, 1930). A series of studies was carried out on the assumption that under the prevailing conditions of food insecurity, food species such as Oreochromis variabilis and O. esculentus can most effectively be conserved through farming. Efforts to conserve the species will be more meaningful if they lead to restoration of the species to the market place. This will depend on:

  • The presence of stocks with sufficient genetic variation for aquaculture development
  • Ease of breeding in captivity
  • Acceptance and willingness by farmers to farm them

Today, O. esculentus is no longer present in Lake Victoria proper, and only a few pockets of O. variabilis can be found in the lake (Jembe et al., 2006). The two species can be found in a few refuge impoundments and satellite lakes of Lake Victoria (Ogutu-Ohwayo, 1990; Balirwa, 1992). Fish for the present conservation programmes were obtained from some of these refuge ecosystems. The conservation efforts included characterizing the refuge ecosystems of the two species, determining their growth performance under culture conditions, assessing their suitability for aquaculture, recruiting farmers to culture the species and testing the suitability of new dams and ponds for aquaculture. Both species breed easily under culture conditions. Therefore production of the fingerlings and their subsequent stocking in ponds, small water bodies and other larger water masses, including Lake Victoria, was a course of action likely to bring about their restoration.

Although the possibility of rescuing threatened populations through translocations and assisted colonization appeals to many conservation biologists in the face of rapid climatic changes (Hunter, 2007; Hoegh-Guldberg et al., 2008), there are those who believe that assisted colonization is not a viable conservation strategy (Ricciardi and Simberloff, 2009). Several issues require strict consideration in attempting to utilize endangered species even when the intention is to conserve.

These include the need to maximize effective population size when carrying out hatchery operations as well as ensuring that the populations in the original refuge are left to thrive. Maintenance of genetic diversity in the population being cultured is important in aquaculture because of the need to maintain long-term success of the population, minimize inbreeding depression, increase selection intensity and assure high selection response. The maintenance of genetic diversity is, however, even more important in conservation efforts due both to the factors above, as well as avoiding random genetic drift.

In the case of O. esculentus and O. variabilis, present efforts at conservation take the form of mitigation measures against repetition of the ecological mistakes which rendered the original ecosystems completely uninhabitable to these fish species. As long as O. niloticus dominates in Lake Victoria and changes in water quality continue unabated, the Nile perch continues to thrive in the lake, and habitat degradation continues, restocking into the lake is not a viable means of conservation due to the possibility of hybridization. However, for sustainability, these interventions should assume that these species could be conserved if used as a useful resource.

Production of fry for stocking

To achieve maximum survival in the remaining populations in the wild and ensure long- term genetic variation under culture, environmentally friendly harvesting and mating procedures were adhered to. O. esculentus was obtained from Lake Kanyaboli and Damside Kisii, the only two places where the pure species was found. O. variabilis broodstock were found in several dams and impoundments such as Kalenyjuok, Mauna, Tinga Mwer, and some bays in Lake Victoria (Maithya, personal communication).

Deliberate efforts were made to ensure minimum mortalities during collection of broodfish. The fish were harvested from dams and impoundments using large mesh nets, ensuring that no young fish were caught and minimizing general fish mortality. Broodfish were then transported to the hatchery in 1 x 0.3-m polythene bags at a density of 50 fish per bag and half-filled with oxygen to ensure maximum survival of broodfish. This made it unnecessary for follow-on brood-stock collection from the same sites.

Maximising both the effective population size and effective breeding numbers is important for long-term continuity of the breeding population. Both the O. esculentus and O. variabilis therefore were produced from a large number of male and female brooders. The number of O. esculentus was 70 males and 150 females obtained from Damside and 49 males and 53 females were obtained from Lake Kanyaboli.

Mass mating has been shown to increase the variance of family size and thus reduce effective number of brooders considerably over what would be achievable with single-pair mating in Nile tilapia and several other fish species (Fessehaye et al., 2006; Gile and Ferguson, 1995; Withler and Beacham, 1994). To obtain the maximum effective number of brooders, the mating of O.esculentus was carried out in a half-sib design in which each male was mated to two to three females. A synthetic population, including all the genes, was produced; crosses were made between the two populations of O. esculentus. 100 males and 200 females were used as brooders.

Selection of farmers and small water bodies

A number of farmers with ponds and small water bodies were visited and selected for the exercise. A total of 250 small water bodies and several ponds were visited. The small water bodies included dams, excavation pits and impoundments. Farmers who had ponds and were willing to take part in the conservation efforts were asked to donate two ponds each for the project. This was because it was not clear what growth was expected from these relatively unknown species in aquaculture. At the same time a comparative growth experiment was carried out at Kenya Marine and Fisheries Research Institute (KMFRI), Sangoro, and some chosen farms.

Farmers with small water bodies have to meet some criteria for the conservation programme. All small water bodies had to be of a surface area of at least 5000 m² and should be free from any other fish, especially other tilapiines. The reason for not allowing smaller water bodies into the programme is to ensure that each founder population is large enough to prevent high rates of inbreeding. Since the aim of the fish stocking programme is to enable the creation of a fishery based on the indigenous tilapiines, either O. esculentus or O. variabilis were stocked in dams, under exceptional cases, where other indigenous fish such as Ningu, the African carp (Labeo victorianus), lungfish (Protopterus aethiopicus) or African catfish (Clarias gariepinus) already existed provided such a dam was over a hectare in surface area.

Production of seed for stocking

Fish seed for initial stocking of the ponds and small water bodies were delivered from hatchery ponds at KMFRI Sangoro Aquaculture Research station. The earliest stocking of experimental units started in 2000 for O. variabilis and 2007 for O. esculentus. The collection and breeding of O. esculentus was carried out with support from the International Foundation for Science (IFS). Fingerlings of O. variabilis used for stocking were 3.9±1.25 g average weight and 4.7±2.16 cm mean length. On the other hand, the average size of O. esculentus used for stocking was 8.8±2.93 g. The experimental design was aimed at determining production levels in ponds and small water bodies (SWB) in terms of growth rates, condition and maximum attainable size (g) of O. variabilis.

Over time, farmers were encouraged to start their own hatcheries for production of the species to augment KMFRI’s efforts. Currently, Mr. Benson Atega in Khwisero Division, Western Kenya, is a producer of O. esculentus fingerlings and more fingerling producers are being recruited. The production of O. variabilis for stocking has been entrusted to community groups managing small water bodies where the fish have been stocked. Research work for the O. variabilis was carried out under the auspices of Jacob Maithya’s PhD work at Moi University, while research work on O. esculentus was carried out through a grant from International Fund for Science (IFS) given to Harrison Charo-Karisa.

Small water body culture systems were selected on the basis of ecological zones that reflect a range of climatic conditions. The small water bodies were Oki, Maranda and Ochot (low- altitude), Lorida Fish Farms and Suka (mid-altitude) and Wanyonyi Fish Farm (high- altitude). Stocking ranged from 240 to 10,380 fingerlings to represent a rate of 1-2 fish per m², depending on size of the culture system. The stocked pond populations were fed on a starter diet for one month and then on growers pelleted feed from Sangoro Aquaculture Research station. Fish in small water bodies relied on natural productivity of the water mass from the onset until maturity. The small water bodies relied entirely on run-off from the catchment area for fertilization and renewal of plankton growth. At least one of these dams (Oki) was also stocked with O. esculentus in order to test the effect of polyculture on growth of the two indigenous tilapiines.

Growth studies for O. variabilis and O. esculentus

Comparative growth studies of Oreochromis esculentus and O. niloticus indicated that O. esculentus can perform well under pond culture. While the growth of O. esculentus was comparable to that of O. niloticus and in some cases superior, the growth of O. variabilis was lower. Generally, the relatively better growth performance of O. esculentus over O. variabilis in different ecological zones within the Lake Victoria basin indicates the viability of the species for culture.

Growth performance of O. esculentus and O. niloticus stocked and harvested after five months in ponds: means, standard deviation (SD), and coefficient of variation (CV) for body weight at stocking and harvest, and percentage survival.

Species n Stocking Initial weight (g) % Survival Harvest
    Mean SD CV   Mean SD CV
O. esculentus 705 8.6 3.0 34.4 36.2* 53.4 11.1 20.8
O. niloticus 701 8.1 2.7 32.8 54.9* 35.8 10.2 28.2

*In this experiment all fish were tagged, which may have resulted in lower than normal survival rates because most of the fish died within the first few days after stocking. Replacement was carried out twice and stopped after one week. No attempts were made to control predation as the experiments were designed to test the conditions in which aquaculture of the species should be carried out in ponds. In large water bodies, the growth of O. esculentus can be better than that of O. niloticus.

Comparison of growth performance between Nile tilapia (Oreochromis niloticus) and the two endemic Oreochromis species (O. variabilis and O. esculentus) in different farms

Farm/dam O. esculentus O. variabilis O. niloticus
  Stocking(g) Harvest (g) Stocking(g) Harvest (g) Stocking(g) Harvest (g)
Oki 8.2 72.5±30.2 6.9 42.7±13.1 - -
Yenga - 72.5±30.2 6.9 42.7±13.1 - -
Benson 7.9 64.5±13.3 - - 8.1 67.3±15
Okot 9.3 99.2±25.7 - - - -
Mukolwe - - 6.8 42.5±8.7 8.3 60.2±9.8
Suka - - 6.7 51.4±14.2 - -

Environmental parameters (pH, dissolved oxygen, temperature, water hardness and turbidity) were measured in situ before stocking and during monthly sampling. Temperature and dissolved oxygen can be limiting to fish growth. The differences in limnological parameters in the stocked small water bodies, grouped under low-, middle- and high-altitude areas, are shown in the table below. Although there are overlaps between the different altitudes, temperature and dissolved oxygen declined with increasing altitude but there was no clear trend for pH, water hardness and turbidity.

Characterization of major environmental gradients influencing growth of fish in small water bodies and Lake Victoria

Parameter Low altitude Medium altitude High altitude Lake Victoria
Temperature (0C) 26.0-30.6 20.1-26.8 18.5-24.9 27.5-30.7
Dissolved oxygen (mg/l) 6.5-9.8 5.1-7.6 4.2-7.2 5.3-7.2
pH 6.92-8.8 5.0-7.8 5.4-7.4 7.4-8.3
Water hardness (mg/l CaCO3) 50.1-130 40.0-86 75.0-256 386.0-92.3
Turbidity (NTU) 72.9-375.0 53.2-188.7 69.7-367 69.4-100.8

It has been demonstrated that the fish species feed on a great variety of planktonic food, Anabaena spp, diatoms and green algae being common food items in many ecological regions. In earlier studies in Lake Victoria, Fryer and Iles (1972) reported that O. esculentus fed on planktonic material, collected from suspension, that were dominated by diatoms. Gee and Gilbert (1967) reported that schools of O. esculentus followed concentrations of diatoms in the lake. This suggests that diatoms may be a preferred food. An ecological study of the small water bodies found many potential food items in the water column. In a study on population characteristics of O. esculentus, Nagayi-Yawe et al.(2006) found that fish caught in Lake Kayugi, Uganda, where diatoms dominated in the stomach contents, attained the largest size, were most fecund and had a high value for condition factor K.

These findings suggest that diatoms are important in determining growth performance of O.esculentus. Although, plankton abundance was not determined in the present study, the growth achieved in the dams was comparable to that in fed ponds, indicating that the propagation of the two species in small water bodies may not require supplemental feeding if low fish densities are maintained. Maintenance of low densities in dams can be achieved by consistently controlling fishing and stock management.

Commonly identified food items for Oreochromis esculentus and Oreochromis variabilis in small water bodies in different ecozones.

Ecozone Food items
Ecozone Food items
Low-altitude Ankistrodesmus, Euglena, Microcystis, Anabaena, Oscillatoria, Nitzschia, discoid diatoms
Mid-altitude diatoms, Oscillatoria, Chlorococcus, Euglena, Anabaena, Scenedesmus
High-altitude Anabaena, Oscillatoria, Selanastrum, euglenoid flagellates


This study has shown that Oreochromis esculentus can grow at the same rate as Nile tilapia. This indicates that it can be promoted as an aquaculture food fish. Although O. variabilis did not seem to grow as well as the other two species in ponds, the culture of the species in small water bodies may be encouraged to target the market segment that consumes smaller fish and as a conservation measure since it is critically endangered. Observation in Damside indicated that O.variabilis could grow up to 300 g; indicating that although O.variabilis is a slow grower, it has the potential to become marketable with time. Small water bodies are therefore to be preferred over ponds for the culture of O.variabilis. O.variabilis may also be used as an ornamental fish given that it has an attractive orange tinge to its dorsal fin. This may earn the farmers extra income. The utilization of the fish, however, calls for stricter control to ensure that pure stocks are maintained.

The culture of indigenous food fishes such as O. variabilis and O. esculentus in small water bodies and ponds can play an important role in improving the food and nutrition security of rural populations (Maithya et al., 2005). However, for a farmer to adopt the culture of indigenous species, especially those whose growth is slow and therefore with low returns, it is necessary to ensure that they are economically empowered. The governments and non-governmental organizations can promote the conservation of indigenous species as a tourist attraction by encouraging fish conservation parks. Such parks can be associated with fish restaurants where fresh fish is sold and can be linked into ecotourism circuits.

At the same time, the indigenous fish could be marketed as an environmental friendly resource where tourists are encouraged to taste indigenous fish that are no longer easily available, akin to the game meat industry, thus allowing farmers to reap more profits than in conventional fish farming. Because tourism is one of the main earners of foreign exchange to the country, the government can offer incentives in terms of subsidies to ensure success. '

Farming of indigenous threatened species should be considered as a way of minimizing the risks of total extinction of the species. Thus strict adherence to practices that ensure conservation of genetic diversity should be adhered to at all times. Hatcheries that produce endangered fish for stocking should be manned by professional managers with adequate training. Farmers should also be trained to ensure they can adhere to the proper procedures.

Major challenges and opportunities in conserving indigenous species

  • Cross-breeding of different strains necessary to ensure genetic variation in farmed stocks
  • Each species is stocked separately or with non-related species to prevent hybridization
  • Poor or slow growth makes the farming of indigenous species less profitable
  • Care must be taken to ensure conservation of donor populations


  • Numerous un-stocked small water bodies exist in the Lake Victoria region
  • Presence of refuge sites with extant populations of endangered species
  • A national focus on support to aquaculture due to dwindling capture fisheries
May 2011

the Fish Site Editor

Learn more