This chapter is necessarily a very brief account on the use of fish for biological control of vectors and hosts of tropical diseases. For more information the reader should consult other sources.
Use of fish in mosquito control has been well known for more than 100 years and there are numerous reports of the effectiveness of various species, and some reports on limitations. The most widely know are Gambusia affinis (the “mosquito fish”), and Poecilia reticulata (the “guppy”). A great number of other fish species, including the young of various sport or food fish, prey on mosquito larvae. For example, for the countries of the Eastern Mediterranean Region, including Somalia and Sudan, and extending eastward to Pakistan, WHO (1981) lists 34 species. Of these, eight species are used directly, and 25 species are promising in mosquito control. Those used directly are: Tilapia zillii (Somalia), Oreochromis mossambicus (Pakistan), Carassius auratus (Iran), Aphanius dispar (Oman, Somalia), Nothobranchius palmquisti and N. guentheri (Somalia), Gambusia affinis (Jordan, Lebanon, Syria, Iran (Islamic Republic of), Iraq, Afghanistan, Sudan), Poecilia reticulata (Pakistan). It is recognised that the use of larvivorous fish for vector control is a simple, inexpensive and reasonably effective measure, and should be considered as a component of integrated strategies. The use of native fish should be given preference to avoid possible undesirable implications of introduction of new fish species. It was recommended to use more than one biological control agent, as, for example, a larvivorous fish together with a phytophagous fish, or two species of larvivorous fish, such as Oryzias latipes and Pseudorasbora parva, which are surface and middle-level feeders respectively (WHO, 1982). Predation on mosquito larvae and on molluscs by fish may be less efficient in dense underwater meadows of submersed plants, and in such situations a joint use of herbivorous and larvivorous and/or molluscivorous fish has been recommended. In Sudan, the introduced grass carp was reported to ingest vast numbers of Biomphalaria snails along with Potamogeton, on which it feeds (WHO, 1981). In subtropical countries, the use of grass carp in irrigation and drainage canals, and in fish ponds, is known to reduce maintenance costs by clearing the weeds, and to reduce mosquito breeding through increased flow of water or predation.
As an example from Africa, in Volta Lake in Ghana, West Africa, Pistia was found to harbour many mosquito larvae, among which Aedomyia africana dominated the fauna (49.5% of the total), followed by Ficalbia splendens (37.4%) and Mansonia africana (8.7%). Anopheles funestus, a malaria vector, and Aedomyia africana, a vector of yellow fever, encephalitis and filariasis, were found commonly associated with Pistia in Volta Lake (Obeng, 1969). The aquatic vegetation in Volta Lake has greatly enhanced the conditions for mosquito vectors of malaria and filariasis, and for snail hosts of schistosomiasis. Numerous laboratory and field experiments have shown that aquatic plant cover provides anopheline larvae with refuge from fish predation. In California, Orr and Resh (1987) have shown that Potamogeton pectinatus reduced the effectiveness of the larvivorous fish Gambusia affinis in controlling anopheline mosquito larvae, as the plant provided cover as a refuge from predation.
Water hyacinth in Malaysia, where this plant is widespread, has been observed to harbour insect vectors of filariasis, e.g. Mansonia uniformis, M. indiana and M. annulifera (Bakar et al., 1984).
An estimated 200 million people are infected worldwide by human schistosomes, worm parasites which are transmitted by various species of freshwater snails, which are frequently associated with aquatic macrophytes. The extent of infestation of human population can be seen on the example of Volta Lake, a man-made reservoir in West Africa.
In 1965, in the one-year-old Volta Lake, Pistia stratiotes supported populations of the gastropod Bulinus forskali, which is not a host of Schistosoma haematobium (Petr, 1968). However, in the following year, Bulinus truncatus rohlfsi was found both on Pistia and Ceratophyllum, while B. forskali was diminishing in numbers. While pre-impoundment studies showed the prevalence of infection in villages and townships situated on the Volta River ranging between 1% and 3%, with one locality on a tributary to the Volta River having 8%, in the Volta delta the disease was endemic, affecting in many villages over 80% or even 90% of the school children (Paperna, 1970). By 1968 the infection in some human shore communities on Volta Lake was as high as 99%. The disease was apparently introduced to the lake by the Ewe fishermen from the lower Volta, who became attracted to the rich fishing grounds in the forming reservoir and settled in small and large camps along the shores, occasionally close to the old villages or the areas of new settlements (Paperna, 1969).
In the 1970s Ceratophyllum was the major habitat for B. t. rohlfsi (Odei, 1979), with the snails also found on many other substrata, including the lake bottom, stones, twigs and particularly palm leaves, used to make fish traps. The snails are dispersed by a number of human activities, particularly fishing. Nets often become entangled and fouled with weeds, and cleaning of nets near villages brings in the snails. Odei (1979) observed a rapid snail build-up on reflooded drawdown areas of the lake, but where there was no offshore Ceratophyllum, there was usually a marked decrease during the dry season, as already observed by Paperna (1969, 1970). Nothing is known about the efficiency of the local fish in controlling the snails. Flow regulation downstream of Volta Lake has caused an explosive growth of Potamogeton octandrus and Vallisneria aethiopica, and the associated B. t. rohlfsi. This has naturally maintained the high level of schistosomiasis, which was widespread already prior to regulation, as mentioned above.
It is not known what proportion of fish food the mosquitoe larvae and bilharzia snail vectors represented in Volta Lake. Predation on mosquito larvae and on molluscs by fish may be less efficient in dense underwater meadows of submersed plants, and in such situations a joint use of herbivorous and larvivorous and/or molluscivorous fish has been recommended. In Sudan, the introduced grass carp was reported to ingest vast numbers of Biomphalaria snails along with Potamogeton, on which it feeds (WHO, 1981). In subtropical countries, the use of grass carp in irrigation and drainage canals, and in fish ponds, is known to reduce maintenance costs by clearing the weeds and reducing mosquito breeding through increased flow of water or predation. (See also below).
Biological control of snails using fish has been practised on a limited scale in fish ponds, small dams, irrigation canals, and in combination with rice-fish culture. In Africa, several species of fish are malacophagous (Serranochromis sp., Astatoreochromis alluaudi, Sarotherodon melanotheron, Clarias sp.) but rigorous observations on their efficiency in field situations are still lacking (McCullough, 1981). He recommended that research be undertaken on the efficacy of snail-eating fish, either alone or in combination with herbivorous fish species. But even in 1995, only a few well documented field trials existed, which would report varying degrees of success with species such as black carp (Mylopharynodon piceus), shellcracker sunfish (Lepomis microlophus), and African cichlids Astatoreochromis alluaudi and Haplochromis mellandi (Slootweg, 1995). M. piceus was effective in controlling nuisance snails in reservoirs in Israel, H. mellandi, a good table fish, was successful in controlling Bulinus and Biomphalaria in fish ponds, irrigation canals and ricefields in Zaire (but Slootweg comments that this information dates from 1956), and L. microlophus successfully controlled snails in Puerto Rico. Slootweg also gives examples of the impact of introduced Astatoreochromis alluaudi on snails in small water bodies in various parts of Africa. In Kenya, where the fish was introduced in earthen dams in 1955, it initially reduced particularly Bimphalaria pfeifferi, but less Bulinus spp. B. pfeifferi formed the principal diet of A. alluaudi (Slootweg et al., undated manuscript). When the dams were revisited in 1986/87, the fish were still present in 5 dams, but snails were found in abundance. In a field trial in Cameroon, which started in 1988, A. alluaudi was stocked together with catfish Clarias gariepinus and tilapia Oreochromis niloticus. Neither catfish, a known snail eater, nor A. alluaudi had any influence on resident snail populations in ponds. The conclusion of the study was that fish culture under good nutritional regimes enhances growth and reproduction of snails. Because of a lack of competition for food the so-called moluscivorous fish prefer to eat “easier” food items, readily available in fish ponds. Field trials in drainage canals and in a rice-fish culture experiment gave additional evidence that the fish were not capable of controlling snails. Slootweg (1995) gives three reasons for failure of Astatoreochromis alluaudi as: foraging behaviour and prey choice, plasticity of pharyngeal teeth, and snail ecology. He believes that other species of snail-eating cichlids will also fail to control molluscs for the same reasons.
Laboratory experiments showed that Oreochromis mossambicus and especially Tilapia rendalli may be promising biological control agents of schistosomiasis by acting as predators of eggs and young Biomphalaria glabrata less than 10 mm in diameter (Graber et al., 1981). Another approach on how to control pulmonate gastropod hosts of schistosomiasis has been suggested by Daldorph and Thomas (1991). Eutrophicated water bodies are characterized by a retreat of macrophytes and their replacement by phytoplankton. While eutrophication is generally considered to be detrimental to conservation and water management in the developed world, it could possibly be of benefit in tropical countries: a collapse of the macrophyte community due to eutrophication could cause a decline in pulmonate snail population. How this would work in tropical waters is not quite clear at present. There is anecdotal evidence that Bulinus truncatus, a host of schistosomiasis in Egypt and the Sudan, is adversely affected by increased levels of eutrophication, but Biomphalaria alexandrina in the same water bodies appears to be increasing and hence benefiting from the ecological changes associated with eutrophication (Daldorph and Thomas, 1991).
Slootweg (1995) noted that if fish were to be used in snail control, it should be limited to permanent habitats and in combination with other control measures. The role of fish must be seen as part of an integrated approach where habitat alterations and appropriate water management can reduce snail breeding and refuge sites, and where natural or introduced competitors and predators put further pressure on snail population. Obeng (1978) suggested control of snail hosts of Schistosoma spp by using herbivorous fish such as grass and common carp to control aquatic plants and hence removing the substrate with which the snails are usually associated. Clearing of aquatic weeds using grass carp in irrigation canals in Egypt had a significant reducing effect on snail populations. Clearing of aquatic weeds reduces the amount of food and also exposes snails to predators that might be naturally present (Slootweg, 1995). A combined use of malacophagous and herbivorous fish is one of the promissing approaches, as applied already in some water bodies for the control of mosquitos. In the irrigation system of Mansuriya District, Egypt, Van Schayck (1985) noticed that grass carp had a positive effect on reducing the number of Bulinus truncatus and Biomphalaria alexandrina. Slootweg (1995) pointed out that future research activities should concentrate on integrated research, rather than hoping to find a snail eating fish that will fully eradicate intermediate hosts of schistosomiasis and other trematode parasites.