Review Article Indian J Med Res 127, January 2008, pp 13-27 Mosquito control by larvivorous fish G. Chandra , I. Bhattacharjee, S.N. Chatterjee & A. Ghosh Department of Zoology, University of Burdwan, Burdwan, India Received June 8, 2006 There is growing of the effects of insecticide used controlling the vectors of human diseases. Manipulating or introducing an auto-reproducing predator into the ecosystem may provide sustained biological control of pest populations. The selection of a biological agent should be based on its selfreplicating capacity, preference for the target pest population in the presence of alternate natural prey, adaptability to the introduced environment, and overall interaction with indigenous organisms. In order to achieve an acceptable range of control, a sound knowledge of various attributes of interactions between the pest population and the predator to be introduced is desirable. Biological larviciding for the control of mosquito borne diseases is feasible and effective only when breeding sites are relatively few or are easily identified and treated. Larval control appears to be promising in urban areas, given that the density of humans needing protection is higher than the limited number of breeding sites. Since 1937, fish have been employed for controlling mosquito larvae. Different types of fish have been used so far in this operational technique. However, use of fish of indigenous origin is found to be more appropriate in this operation. This review presents information on different larvivorous fish species and the present status of their use in mosquito control and provides a ready reference for workers involved and interested in mosquito research. Key words Biocontrol agents - larvivorous fish - mosquito control Mosquito borne diseases continue to be a major problem in almost all tropical and subtropical countries. They are responsible for the transmission of the pathogens causing some of the most life - threatening and debilitating diseases of man, like malaria, yellow fever, dengue fever, chikungunya, filariasis, encephalitis, etc. available than there have been for the last 20 yr1. Further, manufacturers themselves have withdrawn some insecticides due to the high cost of carrying out the additional tests now as per the government norms, in addition to the fact that the production of crop pesticides for the agricultural market is much more lucrative1. The harmful effects of chemicals in mosquitoes as well as on non-target populations, and the development of resistance to these chemicals in mosquitoes along with recent resurgence of different mosquito borne diseases2 have prompted us to explore simple sustainable methods Environmental protection agencies have banned or placed severe restrictions on the use of many pesticides, which were formerly used in mosquito control programmes, and there are now fewer adulticides 13 14 INDIAN J MED RES, JANUARY 2008 of mosquito control. The eradication of mosquito using adulticides is not a prudent strategy, as the adult stage occurs along side human habitation, and they can easily escape remedial measures3,4. Biological control, particularly using larvivorous fish, was important to malaria control programmes in the 20th century, particularly in urban and periurban areas for immediate use in developed and developing countries5. It has a very positive role to play in the integrated control methodologies in which both pesticides and fish or other biotic agents have their own roles6. Biological control refers to the introduction or manipulation of organisms to suppress vector populations. A wide range of organisms helps to regulate mosquito populations naturally through predation, parasitism and competition. As biological mosquito control agents, larvivorous fish (i.e., those that feed on immature stages of mosquitoes) are being used extensively all over the world since the early 1900s (pre DDT era) 7. During the pre DDT era, control of mosquitoes and mosquito vectors of different mosquito borne diseases was undertaken mainly by environmental management, pyrethrum space spraying, use of Paris green, oiling with petrol products and introduction of larvivorous fish. Recognizing the high larvivorous potential of Gambusia affinis, this fish species was purposely introduced from its native Texas (Southern USA) to the Hawaiin Islands in 1905. In 1921, it was introduced in Spain; then from there in Italy during 1920s and later to 60 other countries8. Beginning in 1908, another larvivorous fish, Poecilia reticulata, a native of South America, was introduced for malaria control into British India and many other countries 8. The introduction of the use of DDT in indoor residual spraying for malaria control around the mid - 1940s led to the gradual decline in the use of concepts of environmental management and biological control methods, except in a few programmes in Russia. In the fifties, attention was directed to eradicate mosquitoes using synthetic insecticides until insecticide resistance began to assume prominence. In 1969, the WHO changed its strategy of malaria eradication by spraying houses with synthetic insecticides in favour of the more realistic one for the control of mosquito populations in the larval stages (post DDT era)7. The selection of biological control agents should be based on their potential for unintended impacts, selfreplicating capacity, climatic compatibility, and their capability to maintain very close interactions with target prey populations9. They eliminate certain prey and sustain in such environments (i.e., they eat the prey, when introduced) for long periods thereafter10. However, this will only be possible if the predator possesses extraordinary search efficiency irrespective of the illuminated situation in response to the emergence of prey. It is important to have a sound knowledge of predator’s prey selective patterns and particularly of its mosquito larval selection in the presence of alternate natural prey11,12. In addition, the predator’s adaptability to the introduced environment and overall interaction with indigenous organisms need to be considered prior to introduction13,14. This review presents a brief account of larvivorous fish as mosquito control agents and possible prey/predator interactions in aquatic ecosystems. Definition and criteria for species selection Larvivorous fish are those that feed on immature stages of mosquitoes. According to Job15, larvivorous fish must be small, hardy and capable of getting about easily in shallow waters among thick weeds where mosquitoes find suitable breeding grounds. They must be drought resistant and capable of flourishing in both deep and shallow waters as well as living in drinking water tanks and pools without contaminating the water. They must have the ability to withstand rough handling and transportation for long distances. Larvivorous fish must be prolific breeders having shorter span of life cycle. They must breed freely and successfully in confined waters. Larvivorous fish should be surface feeders and carnivorous in habit and should have a predilection for mosquito larvae even in the presence of other food materials. Another important criterion of all larvivorous species should be its appearance. They should not be brightly coloured or attractive. They should be compatible with the existing fish life in that environment. Above all, they should have no food value, so that the fish-eating people discard them. It is difficult to find a species that satisfies all the above parameters. Hence, the choice usually depends upon those, which satisfy as many of the above qualities as possible. Categorization of larvivorous fish The position of mouth is one of the important characteristics to determine the larvivorous capability of a fish. From the point of view of their efficacy in controlling mosquito larvae, Hora & Mukherjee16 classified the larvivorous fish into the following categories: (i) Typical surface feeders such as CHANDRA et al: MOSQUITO CONTROL BY FISH Aplochelius and Gambusia, which fulfill the characteristic features of larvivorous fish; (ii) Some surface feeders, which are less efficient owing to their mode of life, e.g., Oryzias, Lebistes (Poecilia), Aphanius, etc.; (iii) Sub-surface feeders like Amblypharyngodon mola, Danio, Rasbora, etc.; (iv) Column feeders like Puntius spp., Colisa, Chanda, Anabas, etc., which feed on mosquito larvae when chance permits; (v) Fry of carps and mullets, which are helpful in controlling mosquito larvae; (vi) Predatory fishes like Wallago, Channa, Notopterus and Mystus whose fry may destroy mosquito larvae but whose adults may predate upon other fish including larvicidal fish species. Potential indigenous larvivorous fishes as biocontrol agents 1. Aphanius dispar (Ruppell), 1828 (Common name: Dispar topminnow) Size: Approximately 7.5 cm (3 inches). Distribution: India, Kutch; Pakistan; Sind; Ethiopia: Palestine, along the shores of the Red sea. Ecology: A. dispar is a good larvivorous fish and thrives both in fresh and brackish waters where it breeds freely. It is a delicate species and does not stand transport well. It is suitable for drains and polluted water bodies and useful for stagnant water bodies, disused wells, cesspools, etc. Field trials: Natural habitats - Shallow channels: In 1981, AtaurRahim17 reported the natural occurrence of A. dispar in shallow channels near Riyadh where it was reported to successfully control mosquito larvae. Fish were applied at about 3 fish per square meter water surface. Man made habitats - Experiments using A. dispar in man made artificial containers have shown successful results. It has been reported that A.dispar is a suggested larvivorous fish for the control of vectors of Bancroftian filariasis namely Culex quinquefasciatus Say, 1823 in any kind of stagnant water containing organic pollution18. Wells - Louis & Albert19 reported that in an urban area in Djibouti, the indigenous fish, A. dispar, effectively suppressed the breeding of Anopheles arabiensis and An. gambiae breeding in wells, cisterns and barrels and containers by 97 per cent. Further Fletcher et al20 reported that in an urban area in Ethiopia, the indigenous fish, A. dispar, effectively suppressed An. culicifacies adanensis breeding in wells. 15 Cisterns - A. dispar effectively suppressed the breeding of An. arabiensis and An. gambiae breeding in cisterns when the experiment was conducted in an urban area in Djibouti19. Barrels and containers - A. dispar effectively reduced the breeding of An. arabiensis and An. gambiae breeding in barrels and containers by 97 per cent throughout the experiment in an urban area in Djibouti19. Fletcher et al20 who reported that in an urban area in Ethiopia, the indigenous fish, A. dispar, effectively suppressed An. culicifacies adanensis breeding in containers. However, its impact on malaria transmission could not be assessed. 2. Aplocheilus (McClelland), 1839 (i) Aplocheilus blockii (Arnold), 1911 (Common name: Dwarf panchax) Size: Approximately 9 cm (3.6 inches). Distribution: East Coast of India, from Madras northwards as far as the Pennar system in Andhra Pradesh. Ecology: The fish is a strictly fresh water form and inhabits stationary and sheltered waters of tanks, small streams and rivulets overgrown with thick vegetation. Suitable for ponds and impounded water bodies where carnivorous fish are present, wells and abounded water bodies. Also useful for introduction in overhead tanks, ornamental pools, streambeds and margins, reservoirs and wells for malaria disease vector control. Field trials: Natural habitats - Studies conducted by Kumar et al21 showed that predation by A. blockii reduced the larval population of An. stephensi by 75 per cent along the coastal belt of Goa. Man made habitats - A.blockii is a potential larvivorous fish controlling the spread of chikungunya fever by controlling Aedes albopictus Skuse 1894. The experiment was conducted in tanks and bigger cisterns and barrels18. (ii) Aplocheilus lineatus (Valenciennes), 1846 (Common name: Malabar killie) Size: 10 cm (4 inches). Distribution: Widely distributed in Peninsular India and Sri Lanka. Ecology: Common in tanks, paddy fields, canals, and even in tidal waters. It is suitable for fishponds where carnivorous food fish are present and useful for 16 INDIAN J MED RES, JANUARY 2008 introduction in overhead tanks, artificial containers, cisterns and fountains to control urban malaria, and in pools, streambeds, margins and marshes in rural areas. Distribution: Assam and Upper Burma (now Myanmar) to Punjab and North West Frontier Provinces (Pakistan) and throughout peninsular India. Field trials: Ecology: Fresh water rivers, lakes and estuaries generally preferring weedy environment. It is suitable for brackish waters, backwaters and lagoons and especially in lakes, swamps, ponds, etc., with overgrowth of aquatic plants. Man made habitats - A.lineatus is a potential biocontrol agent. It was reported to control dengue fever vector namely Ae. aegypti Linnaeus, 1762. The experiments were conducted in the breeding habitats of the vector, which included water storage tanks, cistern, and barrels17. (iii) Aplocheilus panchax (Hamilton-Buchanan), 1822 (Common name: Panchax minnow) Size: Approximately 9 cm (3.5 inches); the average size in Bengal is about 5 cm (2 inches). Distribution: India: Bengal, Bihar, Orissa, Assam, Punjab, Uttar Pradesh, Madhya Pradesh, Rajasthan; Sri Lanka; Malaya; Burma (now Myanmar); Thailand; and Indonesia. Ecology: The species is quite hardy and active and inhabits clear shallow fresh and brackish water at low altitudes. Suitable for water bodies where carnivorous fish occur; also wells, marshes, lagoons and polluted storm water drains and any other stagnant water bodies containing organic pollution for filariasis vector control. Field trials: A. panchax is a potential larvivorous fish in controlling several vector species in different types of natural and man-made habitats 18. It controlled An. culicifacies in breeding habitats like rain water pools, irrigation channels, sluggish streams with sandy margins and little vegetation, river bed pools, borrow pits, cemented tanks, swimming pools, freshly laid rice fields etc. An. sundaicus was controlled in brackish waters with algae, behind embankments protecting rice fields, tanks, cleared mangroves and lagoons, ponds, lakes and borrow pits in coastal areas. A. panchax also controlled Cx. quinquefasciatus in cesspools, drains, chocked severs, storm water drains, ponds, polluted waterways, septic tanks, disused wells, well, manure pits etc., and Cx. vishnui in rice fields, marshes, ponds, pools, streams, ditches, borrow pits, irrigation channels, field wells etc. 3. Colisa (Cuvier), 1831 (i) Colisa fasciatus (Schneider), 1801 (Common name: Giant gourami) Size: Approximately 12.5 cm (5 inches). Laboratory trials: C. fasciatus, a locally available indigenous fish collected from stone quarries of Shankargarh block of Allahabad district and ponds/ pools of Dadraul block of Shahjahanpur district (U.P.)22 was evaluated for control of mosquito breeding. The number of larvae consumed per day by a fish of an average length of 5.3 cm collected from stone quarries ranged from 86 to 96. However, larval consumption rate of a fish of an average length of 5.9 cm collected from ponds/pools was 85 to 101. No significant difference was observed in the number of larvae consumed with and without fish food throughout the study period. Field trials: It has been reported that C. fasciatus controls vectors of brugian filariasis and malaria namely Mansonioides indiana Edwards, 1930 and Anopheles (C.) annularis Van der Wulp, 1884 respectively18. Man made habitats: Wells - Field trials carried out in the wells of Dadraul PHC22 revealed that experimental wells with 50 and 75 larvivorous fish did not attain negativity even after three weeks of fish introduction. However, wells with 100 larvivorous fish become negative in one to two weeks of fish introduction. The negativity persisted for 4 months, suggesting that C. fasciatus can be used for controlling mosquito breeding in wells if introduced in sufficient numbers22. (ii) Colisa lalia (Hamilton-Buchanan), 1822 (Common name: Dwarf gourami) Size: Approximately 5 cm (2 inches). Distribution: Northern India, Assam, Bengal, Bihar and Uttar Pradesh. Ecology: It inhabits slow moving streams, rivulets and lakes with plenty of vegetation. Suitable for water bodies where carnivorous food fish are present. Useful for lakes, tanks, etc. CHANDRA et al: MOSQUITO CONTROL BY FISH Field trials: Natural habitats - C. lalia is a good biocontrol agent. It has been reported to control An. (C.) annularis Van der Wulp, 1884 thereby preventing the spread of malaria to a considerable extent. Breeding habitat of the vector include clear weed grown stagnant waters, margins of lakes, tanks, dead rivers, borrow pits, and rice field18. (iii) Colisa sota (Hamilton-Buchanan), 1822 (Common name: Sunset gourami) Size: Approximately 4 cm (1.5 inches). Distribution: Northern India, Assam, Bihar, Bengal and Uttar Pradesh. Ecology: In fresh water rivers and lakes living amidst water plants for protection. Suitable for fresh water bodies where carnivorous food fish are also present. Also useful for introduction in clearweed grown stagnant waters for malaria vector control. Field trials: Natural habitats - C. sota is also a potent biocontrol agent with natural habitats similar to C. lalia18. 4. Chanda nama (Hamilton-Buchanan), 1822 (Common name: Elongate glass perchlet) Size: Approximately 7.6 cm (3 inches). Distribution: India, Bangladesh and Burma (now Myanmar). Ecology: Widely distributed in fresh waters with thick vegetation and breeds freely in confined waters as well as rivers. The species is useful as a larvicidal fish for introduction into forest pools, streams, tanks, ditches, etc., overgrown with vegetation for control of malarial vector. Field trials: Natural habitats - C. nama has been reported to control the population of An. (C.) culicifacies Giles, 1901; An. (C.) balabacensis balabacensis Baisas, 1926; and An. (C.) varuna Iyengar, 1924, which, in turn are potent malarial vectors inhabiting slow moving fresh water17. 5. Oryzias melastigma (McClelland), 1839 (Common name: Estuarine ricefish) 17 Ecology: It is a carnivorous, surface feeder found in both still and running waters. Though primarily as estuarine and brackish water fish, it is found inhabiting fresh waters such as ponds, lakes, rivers, canals and creeks, in large number. An excellent larvivorous form suitable for open shallow water stretches especially in rice fields for control of mosquitoes causing Japenese B encephalitis. Laboratory trials: Predation potential of O. melastigma (measuring 2.4 to 2.5 cm) was experimented in glass containers (measuring 20 ´ 17 ´ 20 cm) against IV instar larvae of Anopheles sp.22. Study on four subsequent days revealed that O. melastigma consumed 98 IV instar larvae of Anopheles per day. Field trials: Natural habitats- O. melastigma is a potential larvivorous fish controlling the occurrence of Japense B encephalitis by restricting populations of Cx. (C.) vishnui Theobald, 1901. Breeding habitats of the vector include rice fields, marshes, ponds, pools, streams, ditches, borrow pits, irrigation channels, field wells, rain water in fallow lands18. A field-based experiment was carried out to evaluate the efficacy of O. melastigma in controlling mosquito breeding in rice fields22 rich in Anopheles sp. and Culex sp. The rice field was divided into nine quadrates of a square meter surface with 15 cm deep water. O.melastigma were released in three quadrates each at the rate of 5 fish per quadrate; three quadrates served as control. Larval density in each quadrate was monitored on day 2, 4, 6, 12, 18, 24, 30, 36 and 42 respectively. Percentage reduction in the density of larva and pupa was calculated. On day 6, O. melastigma lowered the density of III and IV instar larvae and pupae by 76.2 per cent and on subsequent days the percentage reduction ranged from 98.3 to 100 per cent. From day 12 onwards, 100 per cent reduction in larval and pupal densities was recorded. 6. Danio rerio (Hamilton-Buchanan), 1822 (Common name: Zebra danio) Size: Approx. 5 cm (2 inch) Size: 4 cm (1.5 inches). Distribution: All over northern India, Bangladesh; and Burma (now Myanmar). Distribution: Widely distributed in Bengal, Orissa and Tamil Nadu. Ecology: They are surface feeder both in slow and moving streams and ponds, common in rivulets at 18 INDIAN J MED RES, JANUARY 2008 foothills; also useful for introduction in clear water, slow moving stream with grassy margins and shallow earth wells, seepages for control of malaria. Laboratory trials: A laboratory based experiment on predation potential of D. rerio (measuring 2.4 to 2.5 cm) in glass containers (measuring 20 ´ 17 ´ 20 cm) against IV instar larvae of Anopheles sp. was done. The trial was repeated on four subsequent days and the average number of larvae consumed per day by each fish was recorded. Study revealed that D. rerio consumed 52 fourth instar larvae of Anopheles per day22. Field trials: Natural habitats - A field-based experiment to evaluate the efficacy of D. rerio in controlling mosquito breeding in rice fields rich in Anopheles sp. and Culex sp. was carried out21. On day 6, D. rerio lowered the density of III and IV instar larvae and pupae by 86.8 per cent and on subsequent days the percentage reduction ranges from 92.4 to 99.3 per cent. From day 12 onwards, 100 per cent reduction in larval and pupal densities was recorded. 7. Macropodus cupanus (Valenciennes), 1831 (Common name: Spiketailed paradise fish) Size: Approximately 7.5 cm (3 inches). Distribution: Eastern India, Sri Lanka, Western Burma (now Myanmar), Malay Peninsula and Sumatra. Ecology: M. cupanus is a good larvivorous fish thriving both in fresh and brackish waters of the low lands; also found in ditches, paddy fields and shallow waters. It breeds freely in stagnant waters and is tolerant to low content or even deficiency of oxygen. It is also suitable for brackish waters, marshes, lagoons, polluted canals and ditches. Laboratory trials: Mathavan et al23 carried out an experiment with M. cupanus collected from paddy fields. The collected fish were grouped into three weight (W) classes (80, 270 and 570 mg live weight) and maintained in separate glass aquaria. They were acclimated to laboratory conditions (27 ± 1°C) and fed ad libitum on the fourth instar larvae of the mosquito Cx. fatigans. To evoke different levels of hunger, individuals of each W class were deprived of food for 6, 9, 12, 24 or 48 h before commencing the feeding experiments. Significant results were obtained which proved that fish when deprived of food for equal duration, a larger individual becomes hungrier than the smaller ones. Further, prey searching activities of larger individuals increase their hunger level. Potential exotic larvivorous fish as biocontrol agents 1. Carassius auratus (Linnaeus), 1758 (Common name: Gold fish) Size: 200-460 mm. Distribution: Naturally found in China, Korea, Taiwan, Japan, Europe, Siberia, East Asia, Campuchia, etc., and introduced as an aquarium fish in India. Ecology: Aquariums and ornamental ponds. Laboratory trials: Chatterjee et al24 reported the biocontrol efficacy of gold fish under experimental conditions. Under laboratory conditions, one C. auratus was allowed to feed on 200 IV stage larvae of each of An. subpictus, Cx. quinquefasciatus and Ar. subalbatus in separate containers. The number of larvae consumed was 193,188 and 132 per day respectively. Field trials: Man made habitats - Unused reservoirs: Under field conditions, C. auratus efficiently fed upon An. subpictus larvae in unused water reservoirs in Hooghly, West Bengal24. There was a remarkable reduction in the per dip density of An. subpictus larvae from 34.5 to 0.02. 2. Gambusia affinis (Baird & Girard), 1853 (Common name: Top minnow) Size: Male - 3.5 cm, Female - 6 cm. Distribution: A native of coastal waters of United States from New Jersey southwards, introduced into India about 40 years ago from Italy and Thailand. Ecology: Found in freshwater, brackish water and salt marshes with high salinity. Feed on aquatic and terrestrial insects. Terrestrial insects that fall in the water show preference to mosquito larvae. Laboratory trials: Chatterjee & Chandra25 reported the biocontrol efficacy of G. affinis under experimental conditions in the laboratory. G. affinis consumed per day 48, 51 and 31 larvae of An. subpictus, Cx. quinquefasciatus and Ar. subalbatus respectively. The fish was more active during CHANDRA et al: MOSQUITO CONTROL BY FISH 04.00 - 10.00 h. Feeding rate increased with the increase in prey and predator densities. Feeding rate decreased with the increase in water volume (search area). Field trials: Natural habitats - Uncultivated land: Hackett26 described the usefulness of the mosquito predatory fish in malaria control programmes in Europe. According to him, G. affinis, when employed in an area of about 21 km2 on Istrain peninsula, resulted in the reduction in malaria rates from 98 per cent in 1924 to 10 per cent in 1980. In the same way, Menon & Rajagopalan27 studied on the habitat predation rate and larvivorous potentiality of 14 species of fish found in Pondicherry (now Puducherry). In this experiment, each Gambusia fish showed an average predation rate of 65.7 per day on larvae of An. subpictus. In an experiment conducted by Singaravelu et al28 (where they studied the predatory efficiency of G. affinis on the larvae of Ae. aegypti), the role of predation was found to be dependent on prey density. Marshy areas - G. affinis holbrooki were introduced from Italy into the Ghazian marshes during 1922 and were successful in combating malaria29. Man made habitats: Wells - In Hyderabad city, India, an operational release of G. a. holbrooki in 1967 controlled the breeding of An. stephensi in hundreds of wells in about 2 years30. 31 Casuarina pits - Rao et al carried out a study to assess the feasibility of controlling mosquito breeding in casuarina pits in four coastal villages of Pondicherry (now Puducherry) using G. affinis. An. subpictus is the predominant species breeding in the casuarinas pits. A drastic reduction was noted in the number of pit breeding mosquitoes and the maximum control achieved was about 96 per cent. In the check villages where no fish was noticed, the percentage of pit breeding mosquitoes ranged from 55.5 to 91.6 per cent. Overhead tanks - Rajnikant et al32 through a series of experiments, showed that G. affinis was the best predator of the larvae of An. stephensi breeding in overhead tanks. Rice fields - G. affinis is the most widely used species in anti-malarial programmes. It has been used worldwide. G. affinis, when introduced at a rate of 46 fish/m2 water surface in the rice fields, brought about a sharp reduction in the anopheline larval densities 19 and vector biting rates 33 . The experiments were performed in the Kunder valley of Afghanistan. According to the report of Tabibzadeh et al34, Gambusia sp. substantially reduced anopheline larvae in habitats in Iran and contributed to a reduction in malaria transmission and was found to be an important component in malaria eradication. When rice fields had been stocked with 250 to 750 G. affinis per hectare, there was a 95 per cent and a 40 per cent reduction in the immature density of An. freeborni and An. pulcherrimus respectively. In an almost similar experimental set up, Das & Prasad 35 evaluated the mosquito control potential of G. affinis in the rice fields in Shahjahanpur district of Uttar Pradesh, India. At a stocking rate of 5 fish/sq.m, G. affinis significantly reduced the larval and pupal densities in experimental fields as compared to control fields during the entire observation period of 42 days. Control of mosquito breeding in rice fields through fish seemed to be promising. Prasad et al 36 reported that G. affinis survived well in submerged rice fields and provided 87.8 per cent mosquito larval control in Shahajahanpur district, Uttar Pradesh, during 1991. Rajnikant et al31 showed that mosquito fish, G. affinis was the best predator of the larvae of An. culicifacies and An. subpictus breeding in rice fields. 3. Poecilia (Lebistes) reticulata (Peters), 1859 (Common name: Guppy) Size: Male - 2 cm (0.75 inch); Female - 4 cm (1.5 inch). Distribution: It is originally from tropical America. The native distribution includes The Netherlands, West Indies and from Western Venezuela to Guyana. It was imported to India more than once, and restricted to south India and some other parts. Ecology: Poecilia cannot tolerate low temperature. A prolific breeder in tropical waters requiring a temperature between 22 and 24°C, Poecilia lives on artificial food and prefers mosquito larvae. It has been found to tolerate pollution more than Gambusia. Laboratory trials: According to laboratory experiments, an adult and a fingerling of P. reticulata can consume 32 and 18 IV stage An. subpictus larvae in 24 h37. Field trials: Natural habitats - Menon & Rajagopalan27 studied the habitat, predation rate and larvivorous potentiality of 14 species of fish found in Pondicherry (Puducherry). 20 INDIAN J MED RES, JANUARY 2008 Average predation of P. reticulata per day was 53.1 and range of consumption was from 15 to 100. Rice fields - Nalim & Tribuwono38 studied the rice field breeding mosquito An. aconitus in central Java and their effective control using P. reticulata through community participation. They also noticed a sharp decline in the number of malarial cases after introduction of effective biocontrol procedures with larvivorous fishes. Man made habitats: Cisterns and washbasins - Sabatinelli et al39 reported that in Grand Comore Island, the indigenous fish, P. reticulata, effectively suppressed larval and adult population of An.gambiae in washbasins, and cisterns by 85 per cent in a single year using 3-5 fish in a water surface of 1 m2. Containers - Gupta et al40 reported that in India, P. reticulata effectively reduced the breeding of An. stephensi and An. subpictus population breeding in containers, by 86 per cent using 5-10 fish in a water surface of 1 m2. 41 Drains - Saha et al studied on the use of guppy (P. reticulata) as a powerful biocontrol agent in the field of mosquito eradication. They selected twenty mosquito breeding surface drains in the outskirts of Kolkata; ten were observed to contain both mosquito larvae and guppy fishes and the remaining ten were used as control. Per dip larval and pupal densities of Cx. quinquefasciatus varied remarkably than the corresponding densities in the drains without guppy fish. Wells - The role of P. reticulata in the control of mosquito breeding in Ghaziabad district villages near Delhi was judged by the density of immatures and larvivorous fishes; mosquito breeding was found effectively controlled in wells provided the fish did not die or were not prevented from feeding on larvae due to debris. Guppies survived and multiplied in wells over the twenty two week period of observations22. Malaria was a major problem in a sericulture area of Karnataka, south India, where An. culicifacies s.l. and An. fluviatilis s.l. were considered to be the main vectors. Sibling species complexes of these two species were analyzed in three ecologically different villages. Among An. culicifacies, only sibling species A and B were found. In Puram, a village with 22 wells, species A predominated; species B predominated in a village with four wells and a stream, and in a village with a stream and no wells. P. reticulata fish were introduced into all wells and streams in the villages, and after one year no vectors were found in Puram, and all, or nearly all, An. culicifacies were species B in the other two villages. All An. fluviatilis belonged to the sibling species T. Before the introduction of fish, the annual parasite incidence for malaria was high in Puram, but much lower in the other two villages. From 1998 (over one year after release of fish) until 2003, no malaria cases were detected in the three villages42. 4. Nothobranchius guentheri (Pfeffer), (Common name: Killi fish: Egg laying toothed carp) Size: 7 cm (2.5 inch), females are smaller than males. Distribution: East Africa: Mombassa to the Pangani River in Tanzania. Ecology: N. guentheri is a fast growing fish, growing from egg to spawning adult in four weeks. Female lays about 20-100 eggs per day for the whole life which may be until the pool dries up during the dry season. Field trials: Natural habitats - Vanderplank43 brought into light the fact that N. guentheri was the most suitable antimalaria fish available for use when Panama Canal was under construction. 5. Xenentodon cancila (Hamilton-Buchanan), 1822 (Common name: Fresh water gar fish) Size: 30 cm (12 inch). Distribution: Pakistan, India, Bangladesh, Sri Lanka, Burma (now Myanmar) and Thailand. Ecology: This is an elegant surface living fish, which attains a length of 40 cm TL. In North Bengal, it occurs in clear, gravelly, perennial streams and ponds of Terai and Duars. It is fairly common in the GangaBrahmaputra system. Laboratory trials: Chatterjee & Chandra44 reported the efficacy of X. cancila as biocontrol agent against fourth stage larval form of An. subpictus, Cx. quinquefasciatus and Ar. subalbatus under laboratory conditions. Its average consumption rate during 24 h study period was appreciable. Three specimens of X. cancila separately consumed an average of 31, 28, 21 of An. subpictus, Cx. quinquefasciatus and Ar. subalbatus respectively during 24 h study period. CHANDRA et al: MOSQUITO CONTROL BY FISH 21 6.(i) Oreochromis mossambica (Peters), 1852 (Common name: Mozambique cichlid, Tilapia) Mixed culture of larvivorous fishes as biocontrol agents Size: Up to 32 cm (12-13 inch). In order to obtain high production per ha of water body, fast growing compatible species of fish of different feeding habits, or different weight classes of the same species are stocked together in the same pond so that all its ecological niches are occupied by fish. This system of pond management is called mixed fish farming or composite fish culture or polyculture. Distribution: East Africa; an introduced species in India, Pakistan, Sri Lanka, etc. Ecology: O. mossambica grows fast and attains a maximum large size (approx. 3 kg). It is observed in wild, but stunting is common in culture. It grows well and reproduces under salinities as high as 35 per cent. The lower lethal temperature for this species is 10°C. O. mossambica is a good candidate for hybridization if salinity tolerance is desired in the offspring generation. Field trials: Man made habitats - O. mossambica were effective for controlling mosquitoes in cow dung pits22 when introduced against III and IV instar larvae and pupae of Cx. quinquefasciatus and An. culicifacies at the rate of 5 fish per square meter surface area. (ii) Oreochromis niloticus niloticus (Linnaeus), (Common name: Nile Tilapia) Size: Up to 34 cm (13-14 inch). Distribution: East Africa, West Africa, River Nile. Ecology: O. niloticus niloticus is the fastest growing species in many countries. Maximum size is about 3 kg. It does not tolerate high salinity and has poor cold tolerance. It is highly suitable for farming in tropical climate, fresh water and brackish water systems. The lower lethal temperature is 12°C. Laboratory trials: Ghosh et al 45 performed an experiment and established O. n. niloticus as a strong biological agent against larval mosquitoes in the laboratory. Field trials: Man made habitats - Under field conditions a study revealed a significant decrease in per dip larval density after one and half month from introduction of fishes45. The larval density again increased significantly after removal of fish from mosquito breeding ground. When 20 fish were introduced in field conditions, the per dip larval density reduced to 17.38 and 11.39 after 30 and 45 days respectively from an initial value of 26.78. On the contrary, the larval density increased to 21.2 and 24.37 after 30 and 45 days respectively after removal of fish. Predation under co-existence reveals the significance of predatory efficiency of different predator combinations with reference to prey density and exposure period. Several experimental set ups were initiated by different researchers using a mixture of exotic and indigenous fish and at times even bacteria were included. The main aim was to study the efficacy of such combinations as potential biocontrol agents and thereby control mosquito borne diseases. Laboratory trials: Ambrose et al46 performed a predation experiment using Gerris (A) spinolae, Laccotrephes griseus and G. affinis against fourth stage culicine larvae with varying prey densities. Ranking of individual predatory efficiency showed the following sequence: large Gambusia > medium Gambusia > small Gambusia > female Laccotrephes > male Laccotrephes > Gerris. Predation under co-existence revealed the significance of predatory efficiency of different predator combinations with reference to prey density and exposure period. Further Kim et al47 demonstrated biological control of vector mosquitoes (An. sinensis) by the use of fish predators, Moroco oxycephalus and Misgurnus anguillicandatus in the laboratory. A sustained control was achieved during the study period. It was Hurst et al48 who first reported the predation efficacy of 7 native Brisbane fresh water fish on I and IV instars of the fresh water arbovirus vector Cx. annulirostris when evaluated in a series of 24 h laboratory trials. The predatory efficacy of native crimson-spotted rainbow fish Melanotaenia duboulayi (Melanotaeniidae), Australian smelt Retropinna semoni (Retropinnadae), Pacific blueeye Pseudomugil signifer (Atherinidae), firetail gudgeon Hyposeleotris galii (Eleotridae), empire gudgeon Hypseleotris compressa (Eleotridae), and estuary perchlet Ambassis marianus (Ambassidae) was compared with that of the exotic eastern mosquito fish Gambusia holbrooki (Poeciliidae). This environmentally damaging exotic fish has been disseminated worldwide 22 INDIAN J MED RES, JANUARY 2008 and has been declared noxious in Queensland. M. duboulayi was found to consume the greatest number of both I and IV instars of Cx. annulirostris. The predation efficacy of the remaining Australian native species was comparable with that of the exotic G. holbrooki. Later Ghosh et al49 reported that predation experiments using aquarium fish Betta splendens, Pseudotropheus tropheops, Osphronemus gorami and Ptereophyllum scalarae were conducted against fourth instar Anopheles stephensi larvae with varying prey and predator densities. Ranking of individual efficacy against the larval form showed the following sequence: P. tropheops > B. splendens > O. gorami > P. scalarae. The mean larval feeding rates per day of each of the different fish in 1 liter of water were 247, 238, 180 and 40 respectively in descending order of predation efficacy. The corresponding feeding rates in 2 liter of water were 185, 185, 134 and 30 respectively and were not statistically different whether tasted individually or in a school of six. Predation experiment using C. carpio (Linnaeus 1758), Ctenopharyngdon idella (Valenciennes 1844), O. n. niloticus (Linnaeus 1758) and Clarias gariepinus (Burchell 1822) were conducted against fourth instar An. stephensi (Liston 1901) larvae at varying prey and predator densities50. The relative consumption rates of these four fish species on An. stephensi larvae during 24 h experiments under laboratory conditions were Cl. gariepinus > C. idella > C. carpio > O. n. niloticus. Predatory efficacy was positively related with prey density and inversely related with water volume i.e. search area. A laboratory-based experiment was conducted on larval feeding efficacy of fish against III and IV instar larvae of Anopheles, Culex and Aedes sp. in BHEL industrial complex22 where, 10 fish [5 P. reticulata (length 3.0 ± 0.1 cm) and 5 G. affinis (length 3.5 ± 0.1 cm)] were placed in separate enamel basins. After 30 min of introduction of fish, 25 III and IV instar larvae were let in and the larvae swallowed by fish after 10, 20 and 30 min and every hour up to 6 h were counted. The results of the laboratory studies showed that in the first hour, fish were slow in devouring larvae but became very voracious later on. P. reticulata consumed 76 larvae in 6 h, with an average larvivorous potential of 2.53 larvae per hour per fish. On the contrary, G. affinis was somewhat fast in initial hours but slowed down later and consumed only 50 larvae in 6 hrs with an average larvivorous capacity of 1.66 larvae per hour per fish. Chatterjee & Chandra37 studied the feeding activity of the fish, G. affinis and Lebistes reticulatus on An. subpictus larvae in the laboratory. They observed that the average larval feeding rate of adult G. affinis and L. reticulatus was 48 and 32 per day respectively in the laboratory. Field trials: Natural habitats - Ghosh et al 50 reported that predation experiment using C. carpio (Linnaeus 1758), Ctenopharyngdon idella (Valenciennes 1844), O. n. niloticus (Linnaeus 1758) and Clarias gariepinus (Burchell 1822) were conducted against fourth instar An. stephensi (Liston 1901) larvae. A significant decrease in larval abundance in dipper samples was observed at 30 and 45 days since introduction of fish under field conditions. The efficacy of the fish under field conditions was also established by significant increase of larval mosquito abundance at the 30th and the 45 th day since removal of fish from mosquito breeding spots. Willems et al51 reported the predation rate of Pseudomugil signifer and G. holbrooki against four larval instars of Cx. annulirostris, three prey densities, and three vegetation densities. In simulated vegetation trials, P. signifer performed marginally better than G. holbrooki in medium to high-density vegetation (0.3 stems/cm2 and 0.6 stems/cm2, respectively). Rice fields - Yu & Lee52 studied the biological control of malaria vector (An. sinensis Wied) by the combined use of larvivorous fish (Aplocheilus latipus) and herbivorous hybrid fish (Tilapia mossambicus niloticus) in the rice fields of Korea. Stocking of A. latipus and T. mossambicus in weed infested rice fields resulted in an 80-82 per cent reduction in the immature density of An. sinensis from fifth week onwards. Initially the reduction in the larval density was in the range of 70.8 and 73.5 per cent. Later, Kramer 53 tried both fish (G. affinis and Tilapia nilotica nilotica respectively) and Bacillus thuringiensis together and found that both the control agents were compatible in controlling An. freeborni and An. franciscanus (for the former) and An. gambiae (for the latter) breeding in rice fields. In China, Wu et al54 found that stocking rice paddies with edible fish (Cyprinus carpio, Ctenopharyngodon idella, Tilapia spp.) improved rice yield, supported significant fish production, and greatly reduced the number of malaria cases by reducing population of An. sinensis within 150-170 days. Wee et al55 studied the advantages of mosquito control by stocking edible fish in rice paddies. In various field studies conducted in China, it was found that stocking edible fish viz., C. carpio, C. CHANDRA et al: MOSQUITO CONTROL BY FISH idella and Tilapia species at the rate of 60-800/0.07 hectare in rice fields controlled breeding of An. sinensis significantly, which, in turn was correlated with decrease in malarial transmission. Cost - benefit analysis indicated that this approach provided considerable economic advantages and gave incentive to the community to participate in vector control programmes. Biological control of vector mosquitoes (An. sinensis) was demonstrated by Kim et al47 using fish predators, Moroco oxycephalus and Misgurnus anguillicandatus in the semifield rice paddy. A sustained control (53.8 55.2%) was achieved during the study period. Composite culture of edible fishes in rice fields, resulted in 81 per cent reduction in immature population of An. subpictus and An. Vagus 56. Bellini et al 57 showed efficacy of various fish species (Carassius auratus, C. carpio, G. affinis) in the control of rice field mosquitoes in northern Italy. Ponds - Chatterjee & Chandra37 studied feeding activity of the fish, G. affinis and Lebistes reticulatus on An. subpictus larvae in field conditions in the Hooghly district, West Bengal and observed that larval density decreased from 25.7 to 0.36 and 23.7 to 0.5 per dip in the presence of G. affinis and L. reticulatus respectively. Peridomestic habitats - Integrated control of peridomestic larval habitats of Aedes sp. and Culex sp. was studied58, and mosquito abundance before and after treatment with P. reticulata and polystyrene beads was compared to the abundance in an untreated village. Entomological indices from human bait collections and larval surveys indicated that mosquito populations were reduced significantly compared with concurrent samples from the untreated control village and that mosquito control remained effective for 6 months since treatment. Man made habitats: Tanks - Martinez-Ibarra et al59 conducted studies on the indigenous fish species for the control of Ae. aegypti in water storage tanks in Southern Mexico. Five indigenous fish species, Lepisosteus tropicus (Gill), Astyanax fasciatus (Cuvier), Brycon guatemalensis (Regan), Ictalurus meridionalis (Gunther) and P. reticulata (Valenciennes) were used as mosquito control agents. The study revealed significant efficacy of indigenous fish against larval mosquitoes. Ditches - Marti et al60 reported that two neotropical fish species, Cnesterodon decemmaculatus (Poeciliidae) and Jenynsia multidentata (Anablepidae), were 23 collected from human made ditches, a common habitat of Cx. pipiens in La Plata, Argentina. C. decemmaculatus and J. multidentata adults consumed IV instar larvae of Cx. pipiens but the consumption rate varied with prey and predator densities. Eradication of Cx. pipiens from a ditch, where densities had averaged 250 immatures per dip, was achieved 17 days after the introduction of 1700 C. decemmaculatus. Open blocked drain, underground tank, effluent ponds and cement tanks: A field-based experiment on larval feeding efficacy of fish (P. reticulata and G. affinis) in four different habitats22. P. reticulata were introduced in open blocked drains and factory underground tanks, whereas G. affinis were released in effluent pond and cement tanks in the township. Four similar habitats were selected as control. The larval density of each habitat was monitored at weekly intervals using a 250 ml dipper. Fish at the rate of about 50 fish/sq. meter were introduced. Field studies on the impact of P. reticulata on mosquito breeding in drains showed that it can control heavy Cx. quinquefasciatus breeding in about 3 months, the larval density being reduced from 145/ dip to 20/dip. The impact of P. reticulata was highly pronounced in factory underground tanks supporting heavy breeding of Culex sp. There was 100 per cent elimination of breeding of Culex sp. in about four weeks and no breeding was recorded in subsequent weeks. On the other hand, in case of G. affinis, breeding was controlled up to 90-95 per cent in effluent ponds in about 3 months where it was able to control 100 per cent breeding of An. culicifacies and An. stephensi in about 2 wk in cement tank inside township21. In India, uses of larvivorous fish constitute one of the important components of the Urban Malaria Schemes (UMS) and have been incorporated as an important component of selective vector control strategy in the Enhanced Malaria Control Project (EMCP) launched in 1998 with World Bank support in the tribal areas of Gujarat and other parts of the country though UMS faced with operational constraints during application61. Due to emerging threat of malaria and dengue in urban areas, Malaria Research Centre (MRC) (now National Institute of Malaria Research) Field Station at Nadiad in collaboration with the Ahmedabad Municipal Corporation took up a demonstration project on the management of malaria and dengue vectors in Ahmedabad city 62 . Impact of larvivorous fish on mosquito breeding was assessed by monitoring the 24 INDIAN J MED RES, JANUARY 2008 larval density before and after the application. The most common mosquito breeding places such as underground cement tanks, ground level tanks, fountains, elevator chambers (lift wells), wells, mill hydrant tanks, cattle troughs and ponds were exclusively monitored. In general, a sharp reduction in the larval densities was observed in most of the habitats during the post application period. Experience of the development of larvivorous fish network for mosquito control in Ahmedabad city demonstrated the feasibility of using larvivorous fish and proved that this can play an important role in mosquito control in urban areas provided a systematic and planned approach is applied. This strategy would be helpful in controlling the re-emergence of certain vector-borne diseases, particularly in urban areas and will reduce the dependence on insecticides62. Discussion & Conclusion Mosquitoes are and will be the major concerns to come. Biological control is expected to play an increasing role in vector management strategies of the future. The technology is challenging as well as difficult. Unlike the chemical pesticides, the results are often unpredictable with biological agents. This calls for a better understanding of the biological interactions with the environment. Developing and acquiring the necessary skills assume paramount importance. Another important consideration is the recognition of the fact that, in developing countries like India, success of such strategies depends on developing simple technology backed by a campaign of public education. Interventions targeting vectors of diseases are essentially the most effective strategies to control vector-borne diseases. Much of the current efforts directed at the development of new mosquito control tools are confined to the laboratory scale63. Only biological agents carry the potential for overcoming such obstacles, and the most likely agents are those represented by closely related organisms. Towards this end, we require a programme of biological research aimed towards understanding the factors that limit the number of mosquitoes. The search efficiency of the introduced predator and prey selectivity patterns of larvivorous organisms need to be explored by offering mosquito larvae in combination with other alternate natural prey. The success in measuring the efficacy of the candidate agents depends on a multitude of factors: (i) characterization of natural enemy candidates including ecological, morphological, taxonomical, or genetic markers; (ii) selection of climatically matching candidates; (iii) evaluation of semi-field or field cage conditions following quarantine evaluations prior to proceeding with natural release; (iv) assessment of unintended impacts; and (v) the potential efficacy of existing indigenous agents against larval populations. The two main factors determining the efficacy of the fish are the suitability of the fish species to the water bodies where the vector species breed and the ability of the fish to eat enough larvae of vector species to reduce the number of infective bites. The first factor is best addressed by finding a native fish species that thrives under the conditions prevalent in breeding sites rather than to change breeding sites to accustom the fish, although Wu et al54 have recommended a ditch ridge system for rice fields to better accommodate the fish. It has to be kept in mind that the use of pesticides and fertilizers can negatively influence fish stocked in irrigated fields64. The second factor may be strongly influenced by aquatic vegetation, which in turn, can interfere with fish feeding and can also provide refuge for the mosquito larvae. The effectiveness of larvivorous fish to control mosquitoes may vary due to environmental complexity. It needs to evaluate the efficacy of several indigenous larvivorous fish in seasonal wetlands and larger water bodies and their role in the trophic cascade from the community view point65. There are several disadvantages of using larvivorous fish. Gambusia when stocked in waters outside their native range, often causes serious negative ecological impacts. Gambusia is an opportunistic predator with a highly variable diet that includes algae, zooplankton, aquatic insects, as well as eggs and young of fish and amphibians. Gracia-Berthou66 documented a diet shift from diatoms to cladocerans to adult insect with the maturation of Gambusia. They are voracious and highly aggressive fish that compete with the native fish very successfully for viable food and space. Gambusia essentially depletes all large zooplankton while rotifers and phytoplankton densities increase67,68. Gambusia also consumes a high percentage of the phytoplankton grazers. They indirectly cause adverse ecological changes including increased phytoplankton abundance, higher water temperatures, more dissolved organic phosphorous and decreased water clarity 69. Periodic removal of vegetation may be needed to facilitate the activity of the fish70. In temporary or ephemeral mosquito habitats, other forms of control must be developed as evoking active community participation will be a more difficult task. CHANDRA et al: MOSQUITO CONTROL BY FISH References 1. Collins LE, Blackwell A. The biology of Toxorhynchites mosquitoes and their potential as biocontrol agents. Biocontrol News Information 2000; 21 : 105-16. 2. Milam CD, Farris JL, Wilhide JD. Evaluating mosquito control pesticides for effect on target and non-target organisms. Arch Environ Contam Toxicol 2000; 39 : 324-8. 3. Service MW. Biological control of mosquitoes – has it a future? Mosq News 1983; 43 : 113. 4. Service MW. Importance of ecology in Aedes aegypti control. Southeast Asian J Trop Med Public Health 1992; 23 : 681-8. 5. Gratz NG, Pal R. Malaria vector control: larviciding. In: Wernsdorfer WH, McGregor I, editors. Malaria: Principles and practice of malariology. Edinburgh, UK: Churchill Livingstone; 1988. p. 1213-26. 6. Mulla MS. Mosquito control investigations with emphasis on integration of chemical and biological control in mosquito abatement. 29 th Proc Pap Ann Con Calif Mosquito Contr Assoc 1961. p. 1-4. 7. Raghavendra K, Subbarao SK. Chemical insecticides in malaria vector control in India. ICMR Bull 2002; 32 : 1-7. 8. Gerberich JB. Update of annotated bibliography of papers relating to control of mosquitoes by the use of fish for the years 1965. Geneva, World Health Organization, 1985 (unpublished document VBC/ 85.917. 9. Waage JK, DJ Greathead. Biological control: challenges and opportunities. Phil Trans R Soc Lond 1988; 318 : 111-28. 10. Marten GG, Bordes ES, Nguyen M. Use of cyclopoid copepoid for mosquito control. Hydrobiologia 1994; 292/293 : 491-6. 11. Arthington AH, Lloyd LN. Introduced Poecillidae in Australia and New Zealand. In: Meffe GK, Snelson FF, editors. Evolution and ecology of live bearing fishes (Poecillidae), New York: Prentice-Hall; 1989. p. 333-48. 12. Arthington AH, Marshall CJ. Diet of the exotic mosquito fish Gambusia holbrooki in an Australian lake and potential for competition with indigenous fish species. Asian Fish Sci 1999; 12 : 1-16. 13. Denoth M, Frid L, Myers JH. Multiple agents in biological control: improving the odds? Biol Control 2002; 24 : 20-30. 25 indigenous larvivorous fish (Aphanius dispar) and bacterial toxins. Med Trop 1988; 48 : 127-31. 20. Fletcher M, Teklehaimanot A, Yemane G. Control of mosquito larvae in the port city of Assam by an indigenous larvivorous fish, Aphanius dispar. Acta Tropica 1992; 52 : 155-66. 21. Kumar A, Sharma VP, Sumodan PK, Thavaselvam D. Field trials of biolarvicide Bacillus thuringiensis var. israelensis strain 164 and the larvivorous fish Aplocheilus blocki against Anopheles stephensi for malaria control in Goa, India. J Am Mos Control Assoc 1998; 14 : 457-62. 22. Sharma VP, Ghosh A, editors. Larvivorous Fishes of Inland Ecosystems. In: Proceedings of the MRC-CICFRI Workshop; 1989 Sep 27-28; New Delhi. Malaria Research Centre (ICMR): Delhi; 1989. 23. Mathavan S, Muthukrishnan J, Heleenal GA. Studies on predation on mosquito larvae by the fish Macropodus cupanus. Hydrobiologia 1980; 75 : 255-8. 24. Chatterjee SN, Das S, Chandra G. Gold fish (Carrasius auratus) as a strong larval predator of mosquito. Trans Zool Soc India 1997; 1 : 112-4. 25. Chatterjee SN, Chandra G. Laboratory trials on the feeding pattern of Anopheles subpictus, Culex quinquefasciatus and Armigeres subalbatus larvae by Gambusia affinis. Sci Cult 1997; 63 : 51-2. 26. Hackett LW. Malaria in Europe. An ecological study. London: Oxford University Press; 1937. 27. Menon PKB, Rajagopalan PK. Control of mosquito breeding in wells by using Gambusia affinis and Aplocheilus blochii in Pondicherry town. Indian J Med Res 1978; 68 : 927-33. 28. Singaravelu G, Mahalingam S, Jaya Bharati K. Predatory efficiency of larvivorous fish Gambusia affinis on the mosquito larvae of Aedes aegypti and Anopheles stephensi. Curr Sci 1997; 72 : 512-4. 29. Tabibzadeh GB, Nakhai R. 1970. Use of Gambusia fish in the malaria eradication programme of Iran. Geneva, World Health Organization, (unpublished document WHO/VBC/70.198). 30. Sitaraman NL, Karim MA, Reddy GV. Observations on the use of Gambusia affinis holbrooki to control Anopheles stephensi breeding in wells. Results of two years’ study in Greater Hyderabad city, India. Indian J Med Res 1975; 63 : 1509-16. 14. Carlson J, Keating J, Mbogo CM, Kahindi S, Beier JC. Ecological limitations on aquatic mosquito predator colonization in the urban environment. J Vector Ecol 2004; 29 : 331-9. 31. Rao BUS, Krishnamoorthy K, Reddy CBS, Panicker KN. Feasibility of mosquito larval control in casuarina pits using Gambusia affinis. Indian J Med Res 1982; 76 : 684-8. 15. Job TJ. An investigation of the nutrition of the perches of the Madras coast. Rec Mind Mus 1940; 42 : 289-364. 32. Rajnikant, Pandey SD, Sharma SK. Role of biological agents for the control of mosquito breeding in rice fields. Indian J Malariol 1996; 33 : 209-15. 16. Hora SL, Mukherjee DD. The absence of millions, Lebistes reticulates (Peters) in India. Malaria Bureau No.4, Health Bulletin No.12, Delhi 1938. p. 1-49. 17. Ataur-Rahim M. Observations on Aphanius dispar (Ruppell, 1828), a mosquito larvivorous fish in Riyadh, Saudi Arabia. Ann Trop Med Parasit 1981; 75 : 359-62. 18. Manual of entomological surveillance of vector-borne diseases. Delhi: National Institute of Communicable Diseases; 1988. 19. Louis JP, Albert JP. Malaria in the Republic of Djibouti. Strategy for control using a biological antilarval campaign: 33. Rafatjah HA, Arata AA. 1975. The use of larvivorous fish in antimalaria programmes. Geneva, World Health Organization, (unpublished document MAL/WP/75.6 Rev. 1). 34. Tabibzadeh I, Behbehani G, Nakhai R. Use of Gambusia affinis as a biological agent against Culex tarsalis and Anopheles freeborni in Sacramento valley rice fields. Mosq News 1971; 32 : 146-52. 35. Das MK, Prasad RN. Evaluation of mosquito fish Gambusia affinis in the control of mosquito breeding in rice fields. Indian J Med Res 1991; 72 : 214-7. 26 INDIAN J MED RES, JANUARY 2008 36. Prasad H, Prasad RN, Haq S. Role of biological agents for the control of mosquito breeding through Gambusia affinis in rice fields. Indian J Malariol 1993; 30 : 57-65. 37. Chatterjee SN, Chandra G. Feeding pattern of Gambusia affinis and Lebistes reticulates on Anopheles subpictus larvae in the laboratory and field conditions. J Appl Zool Res 1997; 8 : 152-3. 38. Nalim S, Tribuwono D. Control demonstration of the rice field breeding mosquito Anopheles aconitus Donitz in central Java, using Poecilia reticulata through community participation: 2. Culturing, distribution and use of fish in the field. Bull Penet Kesehatan 1987; 15 : 1-7. 39. Sabatinelli G, Blanchy S, Majori G, Papakay M. Impact de l’utilisations du poisson larvivore Poecilia reticulata sur la transmission du paludisme en RFI des Comores. Ann Parasitol Hum Comp 1991; 66 : 84-8. 40. Gupta DK, Bhatt RM, Sharma RC, Gautam AS, Rajnikant. Intradomestic mosquito breeding sources and their management. Indian J Malariol 1992; 29 : 41-6. 41. Saha D, Biswas D, Chatterjee KK, Chandra G, Bhattacharya A, Bhattacharya S, et al. Guppy (Poecelia reticulata) as a natural predator of Culex quinquefasciatus larvae. Bull Cal Sch Med 1986; 34 : 1-4. 42. Ghosh SK, Tiwari SN, Sathyanarayan TS, Sampath TR, Sharma VP, Nanda N, et al. Larvivorous fish in wells target the malaria vector sibling species of the Anopheles culicifacies complex in villages in Karnataka, India. Trans R Soc Trop Med Hyg 2005; 99 : 101-5. 43. Vanderplank FL. Nothobranchius and Barbus sp: Indigenous antimalarial fish in East Africa. E Afr Med J 1941; 17 : 431-6. 44. Chatterjee SN, Chandra G. Laboratory trials on the feeding pattern of Anopheles subpictus, Culex quinquefasciatus and Armigeres subalbatus by Xenentodon cancila fry. Env Ecol 1996; 14 : 173-4. 45. Ghosh A, Bhattacharjee I, Chandra G. Biocontrol efficacy by Oreochromis niloticus niloticus. J Appl Zool Res 2006; 17 : 114-6. 46. Ambrose T, Mani T, Vincent S, Cyril L, Kumar A, Mathews KT. Biocontrol efficacy of Gerris (A) spinolae, Laccotrephes griseus and Gambusia affinis on larval mosquitoes. Indian J Malariol 1993; 30 : 187-92. 47. Kim HC, Kim MS, Yu HS. Biological control of vector mosquitoes by the use of fish predators, Moroco oxycephalus and Misgurnus anguillicandatus in the laboratory and semi field rive paddy. Korean J Entomol 1994; 24 : 269-84. 51. Willems KJ, Cameron EW, Russell RC. A comparison of mosquito predation by the fish Pseudomugil signifer Kner and Gambusia holbrooki (Girard) in laboratory trials. J Vect Ecol 2005; 30 : 87-90. 52. Yu HS, Lee JH. Biological control of malaria vector (Anopheles sinensis Wied.) by combined use of larvivorous fish (Aplocheilus latipus) and herbivorous hybrid (Tilapia mossambicus niloticus) in rice paddies of Korea. Korean J Appl Entomol 1989; 28 : 229-36. 53. Kramer VL. The ecology and biological control of mosquitoes [Culex tarsalis, Anopheles freeborni and Anopheles franciscanus (Anopheles pseudopunctipennis franciscanus)] in California wild and white rice fields (using Bacillus thuringiensis and larvivorous fish). Dissertation Abstracts International. Bull Sci Engin 1989; 49 : 4670-B. 54. Wu N, Liao G, Li D, Luo Y, Zhong G. The advantages of mosquito biocontrol by stocking edible fish in rice paddies. Southeast Asian J Trop Med Public Health 1991; 22 : 43642. 55. Wee N, Liao GH, Li DF, Luo YL, Zhong GM. The advantage of mosquito biocontrol by stocking edible fish in rice paddies. Southeast Asian J Trop Med Public Health 1991; 22 : 436-42. 56. Victor TJ, Chandrasekaran B, Reuben R. Composite fish culture for mosquito control in rice field in Southern India. Southeast Asian J Trop Med Public Health 1994; 25 : 522-7. 57. Bellini R, Veronesi R, Rizzoli M. Efficacy of various fish species Carassius auratus (L.), Cyprinus carpio (L.), Gambusia affinis (Baird and Girard) in the control of rice field mosquitoes in Northern Italy. Bull Soc Vector Ecol 1994; 19 : 87-99. 58. Lardeux F, Sechan Y, Loncke S, Deparis X, Cheffort J, Faaruia M. Integrated control of peridomestic larval habitats of Aedes and Culex mosquitoes (Diptera: Culicidae) in Atoll villages of French Polynesia. J Med Entomol 2002; 39 : 493-8. 59. Martinez-Ibarra JA, Guillen YG, Arredondo-Jimenez JI, Rodrigu-Lopez MH. Indigenous fish species for the control of Aedes aegypti in water storage tanks in Southern Mexico. BioControl 2002; 47 : 481-6. 60. Marti GA, Azpelicueta MM,Tranchida MC,Pelizza SA, Garcia JJ. Predation efficiency of indigenous larvivorous fish species on Culex pipiiens L. larvae (Diptera: Culicidae) in drainage ditches in Argentina. J Vect Ecol 2006; 31 : 102-6. 61. Pattanayak S, Sharma VP, Kalra NL, Orlov VS, Sharma RS.Malaria paradigms in India and control strategies. Indian J Malariol 1994; 31 : 141-99. 62. Haq S, Yadav RS. Developing larvivorous fish network for mosquito control in urban areas: A case study. ICMR Bull 2003; 33 : 69-73. 48. Hurst TP, Brown MD, Kay BH. Laboratory evaluation of the predation efficacy of native Australian fish on Culex annulirostris (Diptera: Culicidae). Am J Mosq Control Assoc 2004; 20 : 286-91. 63. Spielman A, Kitron U, Pollack RJ. Time limitation and the role of research in the worldwide attempt to eradicate malaria. J Med Entomol 1993; 30 : 6-19. 49. Ghosh A, Bhattacharjee I, Ganguly M, Mondal S, Chandra G. Efficacy of some common aquarium fishes as biocontrol agent of preadult mosquitoes. Bull Pen Kesh 2004; 32 : 144-9. 64. Lacey LA, Lacey CM. The medical importance or riceland mosquitoes and their control using alternatives to chemical insecticides. J Am Mos Control Assoc 1990; 6 (Suppl) : 1-93. 50. Ghosh A, Mondal S, Bhattacharjee I, Chandra G. Biological control of vector mosquitoes by some common exotic fish predators. Turk J Biol 2005; 29 : 167-71. 65. Blaustein L, Chase JM. Interactions between mosquito larvae and species that share the same trophic level. Ann Rev Entomol 2007; 52 : 489-507. CHANDRA et al: MOSQUITO CONTROL BY FISH 27 66. Garcia-Berthou E. Food of introduced mosquito fish: ontogenetic diet shift and prey selection. J Fish Biol 1999; 55 : 135-47. 69. Hurlbert SH, Zedler J, Fairbanks D. Ecosystem alteration by mosquitofish (Gambusia affinis) predation. Science 1972; 175 : 639-41. 67. Hurlbert SH, Mulla MS. Impacts of mosquitofish (Gambusia affinis) predation on plankton communities. Hydrobiologia 1981; 83 : 125-51. 70. Dua VK, Sharma SK. Use of guppy and gambusia fishes for control of mosquito breeding at BHEL industrial complex, Hardwar (U.P.). In: Sharma VP, Ghosh A, editors, Larvivorous fishes of inland ecosystems. Delhi: Malaria Research Centre; 1994. p. 35-42. 68. Bence JR. Indirect effects and biological control of mosquitoes by mosquitofish. J Appl Ecol 1988; 25 : 505-21. Reprint requests: Dr Goutam Chandra, Professor & Former Head, Department of Zoology, Mosquito Research Unit Parasitology Research Laboratory, University of Burdwan, Burdwan 713104, India e-mail: goutamchandra63@yahoo.co.in