The eastern mosquitofish Gambusia holbrooki is an exotic fish, which is now widespread around the globe and is known to adversely affect native fish through competition and/or predation. G. holbrooki ...
The eastern mosquitofish Gambusia holbrooki is an exotic fish, which is now widespread around the globe and is known to adversely affect native fish through competition and/or predation. G. holbrooki is known to impact upon native fish via competition with similar sized species, predation upon the fry and eggs of native fish, and by attacking all sized fish by aggressive fin-nipping, thereby leaving them susceptible to disease (Arthington, 1991). G. holbrooki can rapidly increase in population size due to its rapid maturation to breeding age (four weeks in summer) and high survival rate of young (Milton and Arthington, 1983; Lloyd et al., 1986; Lloyd, 1990).
This datasheet focuses upon Gambusia holbrooki, although much of the biology is confused between the previous subspecies and now congeners, G. holbrooki and G. affinis. Therefore, unless specified, reference to mosquitofish refers to both species. Data for both species is also presented because both species were both regarded as Gambusia affinis prior to 1988 and even after this date workers often referred Gambusia affinis as a generic name for both species.
The genus Gambusia comprises about 30 species divided into four sub-genera: Orthophallus, Arthrophallus, Heterophallina,and Gambusia. Other gambusine species rarely coexist with G. affinis, but when they do they are usually ecologically segregated. However, this species will hybridise with some of its congeners, including G. heterochir, G. nobilis, and G. marshi (Hubbs, 1955; Rosen and Bailey, 1963; Hubbs and Perden, 1969; Schoenherr, 1974; Yardley and Hubbs, 1976).
Wotten et al. (1988) formally recognised two species - G. holbrooki and G. affinis - which were previously regarded as sub-species of G. affinis (G. a. holbrooki and G. a. affinis). These species are different in a number of meristic characteristics but are definitively differentiated by the morphological differences of the gonopodium (male anal fin) (see Lloyd, 1990c). Work by Lloyd and Tomasov (1985) showed that the Gambusia in Australia are all G. holbrooki.
The male G. holbrooki is about 35 mm standard length whereas the female is larger (up to 60 mm) with a deeper body, the anal fin unmodified and when pregnant, a gravid spot is visible just above the vent (Lloyd, 1987). The fish are mostly translucent grey with a bluish sheen on their sides with a silver belly (Lloyd, 1987). The fins are colourless, with transverse rows of black spots. Some male mosquitofish have irregular black blotching, though some largely melanistic male individuals exist but are uncommon in their native range (Sterba, 1962) and are absent from Australia (Lloyd, 1987). On the male, the anal fin is modified to form a long, thin intromitent organ, the gonopodium, used for sperm transfer (Lloyd, 1990c). The body is slightly compressed with a large and flattened head. The eyes are large, and the mouth is small and terminal (Lloyd, 1987).
Mosquitofish have become the most widely distributed freshwater teleosts in the world (Krumholz, 1948) mainly through deliberate human introductions. The native range of G. affinis is throughout the Mississippi Basin and the tributaries to the northern Gulf of Mexico; G. holbrooki ranges from New Jersey in the east to the Gulf of Mexico and northern Florida.
The species (both G. affinis and G. holbrooki) have been introduced across the USA and much of southern Canada, parts of South America, Australia, New Zealand, Papua New Guinea, Indonesia, Many Pacific Islands, SE Asia, China, Japan, India, throughout Europe and the Middle East, South Africa, and parts of Northern Africa (Lloyd, 1987).
Because G. affinis and G. holbrooki were treated as subspecies of G. affinis prior to 1988, it is not always possible to discern which species is referred to in earlier literature. It was confirmed by Pyke (2005) that records across Australia were of G. holbrooki.
The distribution in this summary table is based on all the information available. When several references are cited, they may give conflicting information on the status. Further details may be available for individual references in the Distribution Table Details section which can be selected by going to Generate Report.
Mosquitofish have become the most widely distributed freshwater teleosts in the world (Krumholz, 1948) mainly through deliberate human introductions (e.g. Lintermans, 2004). Throughout the world, it has been widely distributed to aid mosquito control in rice paddies and natural waters (Krumholz, 1948; Lloyd et al., 1986; Arthington and Lloyd, 1989). Worldwide introduction of Gambusia has occurred since the first introduction into Hawaii in 1905 (Krumholz, 1948).
It is now known that G. holbrooki was introduced to Mediterranean Europe and Australia, whereas G. affinis was introduced outside its natural range to the western United States, Hawaii, and countries in Africa (Rehage and Sih, 2004).
G. holbrooki were first introduced to Brisbane in 1925 and Sydney the following year (Wilson, 1960). The distribution of the Australian populations has continued to expand through new invasions (such as Northern Tasmania) or in filling of locations within catchments. Gambusia was actively introduced by health authorities to new locations in Australia until the 1990s (Lloyd, 1987).
Only G. holbrooki has been identified in the current field records and museum collections that exist in Australia (Lloyd and Tomasov, 1985). Stocks of mosquitofish have been found in many isolated water bodies in central Australia, suggesting that natural events such as floods aid their dispersal (Lloyd, 1987; Chapman and Warburton, 2006). However, it is human activity that has most aided the dispersal of G. holbrooki, largely through projects aimed at mosquito control on a local, national and global scale (Arthington et al., 1983).
G. holbrooki is highly invasive because of its ability to move and colonise new habitats, its high fecundity, high survival of juveniles and rapid population growth (Moyle and Light, 1996; Williamson and Fitter, 1996; Alemadi and Jenkins, 2007). Mosquitofish can disperse through waters as shallow as 3 mm, which is only half of average body depth (Alemadi and Jenkins, 2007) and uses drains and natural channels to disperse between water bodies. Rehage and Sih (2004) link dispersal behaviour to the ‘invasiveness’ of a species, with G. affinis dispersing faster than G. holbrooki due to the movement of females into new territories. However, other important characteristics, such as fecundity, and maximum population growth rates suggest that G. holbrooki is ’a superior invader‘ (Rehage and Sih, 2004). Eastern mosquitofish can rapidly increase in population size due to their rapid maturation to breeding age (four weeks in summer) and high survival rate of young (Milton and Arthington, 1983; Lloyd et al., 1986; Lloyd, 1990a).
The native habitat of mosquitofish is the lowland ponds, lakes and streams of southern USA (Casterlin and Reynolds, 1977). Mosquitofish are particularly adapted to exploiting inundated floodplains (Ross and Baker, 1983). G. holbrooki is abundant in near-shore environments, close to dense vegetation, and prefers sluggish waters to running water (Lloyd, 1987). Other habitat preferences include: shallow water, dark-coloured substrata, and subsurface vegetation (providing lateral rather than vertical concealment) (Casterlin and Reynolds, 1977; Arthington and Marshall, 1999). They are very adaptable and will live in almost aquatic habitats from fresh to hyper-saline, cold temperate to tropical waters (and artificially heated waters), inland, coastal and estuarine waters, and both still and slow-flowing waters. The species does seem to be poorly adapted to fast flowing waters which inhibits its ability to develop large populations (Lloyd, 1987).
Disturbed environments are prone to invasion from this species due to the fact that disturbed habitats can support large populations of invertebrate species (which tend to be pests, e.g. chironomids) and often lack other fish “due to harsh physical conditions” (Lloyd, 1987). Mosquitofish are typically more resistant to pollutants, including organic wastes, herbicides, insecticides, rotenone, phenols, heavy metals and radiation, than most other fish (Lloyd, 1987; Edwards, 2005). The ability of the species to survive genetic ‘bottle-necks’ and their ability to physiologically and genetically adapt to different environments is also important in their spread and success.
Environmental Requirements
Mosquitofish have invaded a wide range of habitats throughout the world including: hot springs, rivers, streams, lakes, swamps, billabongs, cooling pondages, rice fields, ornamental ponds, estuaries, near-shore marine habitats and salt lakes (Lloyd, 1987; Arthington and Lloyd, 1989).
Mosquitofish are found in waters from 0.5°C to 39°C, with a preference for warm waters of about 25°C (Otto, 1974), though Pyke (2005) suggests that mosquitofish prefer water temperatures between 31-35°C. Juvenile mosquitofish are more thermally tolerant than adults, allowing them to colonise and exploit warm patches of the environment with increasing growth, survival, and maturation rate. Mosquitofish can inhabit ice-covered waters (Hirose, 1976;Sasa and Kurihara, 1980) in Japan and hot bores (over 37°C) in central Australia (John Glover, personal communication, cited in Lloyd, 1987).
The broad salinity tolerance of mosquitofish allows them to colonise environments, such as salt lakes, estuaries, near coastal marine environments (Lloyd, 1987). The salinity LD50 for mosquitofish is more than 58g/L and they can tolerate direct transfers to salinity differences of up to seawater (35 g/L) with few mortalities (Chervinski, 1983).
Mosquitofish can tolerate oxygen concentrations as low as 1.3 mg/L, without access to surface water but can withstand virtual anoxia by utilising of the oxygen-rich surface water/air interface because it has a dorso-ventral mouth (Lewis, 1970).
Mosquitofish have become the most widely distributed freshwater teleosts in the world (Krumholz, 1948) mainly through deliberate human introductions (e.g. Lintermans, 2004). Throughout the world, G. holbrooki and G. affinis have been widely distributed to aid mosquito control in rice paddies and natural waters. (Krumholz, 1948; Lloyd et al., 1986; Arthington and Lloyd, 1989). Worldwide introduction of Gambusia has occurred since the first introduction into Hawaii in 1905 (Krumholz, 1948).
Gambusia are cited to be used in the commercial aquarium industry (www.fishbase.org) but poor sales are likely given its noxious status in many countries, its aggressive behaviour, and its poor appearance. They are likely to be under aquaculture in the USA for use for mosquito control in rice fields. Mosquito control authorities have been known to transfer mosquitofish between locations. Children may also collect mosquitofish for bait and help to distribute it when they move from one place to another.
In Australia, and possibly other parts of its introduced range, G. holbrooki faces few predators, parasites, diseases or competitors (Lloyd, 1987). Experiments have shown that several Australian native fish predators actively avoid eating this mosquitofish (Lloyd, 1987).
Rehage et al. (2005a, b) suggest that invasive species of Gambusia (G. holbrooki and G. affinis) are more efficient foragers than their non-invasive relatives (G. geiseri and G. hispaniolae), and that invasive juveniles feed more voraciously and widely than adults. This study suggests that Gambusia, released from their native habitat and therefore native predators, increase in size, and accordingly, have a higher feeding rate which has a greater impact on native species.
G. holbrooki is a voracious predator of invertebrates resulting in a possible increase in mosquito populations, as it targets insects, which are natural predators of mosquitoes. Willems et al. (2005) compared two different fish species on their predation of mosquitoes which found that Gambusia were not as effective as controlling mosquito larvae as the native fish in the study. Studies of small native fish and G. holbrooki diets in the River Murray in South Australia (Lloyd, 1987) showed that the eastern mosquitofish was a poor predator of mosquito larvae compared to most of the small native fish species. Ling (2004) showed that G. affinis may inhabit an unoccupied niche in New Zealand environments and a threat to native invertebrate grazers. G. holbrooki has been shown, through gut contents assessment, to be significant predators of native fish in Australia (Ivantsoff and Aarn, 1999). It can also have an impact on the tadpoles of native frogs (Sadlier and Pressey, 1994; Morgan and Buttemer, 1996; Komak and Crossland, 2000; Pyke and White, 2000). The reproductive rituals, breeding success and growth of native fishes can be impacted by gambusia (Howe et al., 1997).
Threatened Species
Gambusia occupy the specialized dystrophic habitats of one restricted and two endangered Australian freshwater fishes - Rhadinocentrus ornatus, Pseudomugil mellis, Nannoperca oxleyana (all found in South-eastern Queensland). It is possible that G. holbrooki and the three species interact and compete for habitat, food and spawning areas (Howe et al., 1997; Arthington and Marshall, 1999; Knight and Arthington, 2008).
G. holbrooki and G. affinis have been introduced widely for mosquito control. They are used in the commercial aquarium industry (www.fishbase.org) but poor sales are likely given their noxious status in many countries, aggressive behaviour, and poor appearance.
All of these measures have been applied to prevent the spread of Gambusia in Australia. The Australian Quarantine Inspection Service (AQIS) has listed G. holbrooki as a high-risk species, highly likely to establish and spread more widely once introduced to new areas (Arthington et al., 1999). Predation by G. holbrooki has been listed as a key threatening process under the New South Wales Threatened Species Conservation Act 1995. The inappropriate spread of Gambusia by humans for the purpose of mosquito control remains a problem in spite of repeated cautionary advice over the past 20 years (Arthington and Lloyd, 1989). Gambusia has been recognised as a potential vertebrate pest in Australia and as such should be managed under the edicts of vertebrate pest policies, programmes, and legislation (Bomford, 2001; Bomford and Glover, 2004).
Early warning systems
As eradication of fish after establishment as breeding populations is usually impossible, alien fish management should concentrate on preventing new introductions and limiting future spread (Lodge et al., 1998; Lodge and Shrader-Frechette, 2003). The resources to monitor and then act upon eradication of new populations are often limited. Involving the community in monitoring activities is one way to achieve early warning advice to management authorities. Involving the community provides people with a sense of ownership of the pest fish issue and an improved understanding of the complexities of managing alien fish species.
Rapid response
Not known to be a feature of Gambusia control but should be a first line of action (McDowall, 2004).
Public awareness
Gambusia is one of about 29 aquarium species listed as harmful to native fishes, invertebrates and aquatic ecosystem integrity in Australia (Arthington et al., 1999). Scientific publications, information papers, pamphlets and posters have all been used to inform the public about the dangers of alien fishes, especially those declared noxious under state and commonwealth legislation, or quarantine regulations such as those of the Australian Quarantine Inspection Service (AQIS). Involving the community in monitoring activities provides people with a sense of ownership of the pest fish issue and an improved understanding of the complexities of managing alien fish species.
Eradication
Attempts have been made to eradicate G. holbrooki from water bodies using the fish poison rotenone. Most native fish are killed by a rotenone concentration of 0.5 ppm but Gambusia can survive this concentration without mortality (Pyke, 2005). Impacts on native fishes and other native fauna have been mitigated by releasing potassium permanganate downstream of the rotenone release point in flowing waterways. Gambusia is more tolerant of the organo-phosphorus pesticide Dursban™ than several native fishes (Pyke, 2005). These observations mean that chemical control methods are highly likely to affect native fish and other aquatic biota well before useful levels of gambusia mortality can be achieved.
Containment/Zoning
May be applicable to Gambusia, but no explicit information available.
Control
Cultural control and sanitary measures
It is well established that many exotic fishes flourish in disturbed habitats, even highly polluted ones in the case of Gambusia. It has been argued that maintaining the natural flow regime and habitat characteristics of aquatic systems should be an effective preventative measure, indeed a management principle (Arthington et al., 1990).
Physical/mechanical control
Gambusia species have been controlled to some extent by draining standing water bodies (especially outdoor ornamental ponds), and by cutting off the pathways for re-colonization (e.g. closing off access drains into and from ornamental ponds). Gambusia is highly invasive and can move through shallow water half their body depth. Barriers to spread must obstruct dispersal pathways (e.g. ditches and channels) as shallow as 3 mm to prevent local spread of Gambusia (Alemadi and Jenkins, 2007).
Movement control
Declaration of noxious status prohibits movement of Gambusia by humans in some countries (e.g. Australia). When Gambusia is collected as by-catch in ecological research programmes it must be disposed of immediately, never transported or introduced to any other water-body or taken to domestic aquaria.
Biological control
Factors limiting natural Gambusia populations (Courtenay and Meffe, 1989) may include endemic parasites and pathogens, as well as predators. Gambusia is reported to host at least 23 parasite species (LN Lloyd, cited in Arthington and Lloyd, 1989), however, most are not host specific. Pathogens may contribute to suppression of Gambusia, yet reliable data on their role are scarce. In April 2000 a fungal infection killed more than 80% of G. affinis in Lake Naini Tal, Uttaranchal, India (Nagdali and Gupta, 2002). Gambusia in its native range may be partially controlled by predators including topminnows of the genus Fundulus. However, there are no reports suggesting that, beyond its native range Gambusia is other than a minor constituent of the diet of piscivores, even other alien fish species such as trout. If Gambusia is a major part of fish diets in some locations then manipulation of aquatic food webs may provide a means to control Gambusia via predation pressure. Unmack (unpublished data) and others have observed that Gambusia and larger galaxiids rarely coexist and galaxiids eagerly devour Gambusia in aquaria. In small water-bodies diadromous galaxiids such as the common galaxiid (Galaxias maculatus) could be introduced at high densities to prey on Gambusia. However, after 2-4 years the galaxiids would die of old age, leaving opportunities for Gambusia to re-establish viable populations. Repeated releases of the galaxiid would be needed to sustain high Gambusia mortality.
Chemical control
Attempts have been made to reduce the abundance of G.holbrooki from water-bodies using the poison rotenone. Impacts on native fishes and other fauna have been mitigated by releasing potassium permanganate downstream of the release point in flowing waterways.
IPM
The integration of science (e.g. alien species biology) with applied management actions will improve environmental outcomes, yet there does not appear to be any example of an integrated management programme to control Gambusia.
Monitoring and Surveillance
The presence and abundance of Gambusia species are being monitored as part of formal stream and river health monitoring programs in Australia (e.g. Kennard et al., 2005).
Ecosystem Restoration
Scientists have long advocated that the maintenance of healthy, functional aquatic ecosystems and the natural processes that support native biodiversity is the best way to suppress alien fishes (Arthington et al., 1990), i.e. maintain ecosystem resistance and resilience. Damaged ecosystems could be restored in ways that are likely to suppress Gambusia, e.g. by protecting riparian corridors, minimizing habitat loss, restoring natural flow regimes or releasing environmental flows to flush out mosquitofish and so constrain their spread and population increase (e.g. Bunn and Arthington, 2002). Issues of alien species being moved about by inter-basin water transfers have been addressed by Todd et al. (2002).
Angela Arthington, Australian Rivers Institute, Room 1.09A, Environ. 2, (Building N13), Griffith School of Environment, Griffith University, Nathan QLD 4111, Australia
Lance Lloyd, Lloyd Environmental Pty Ltd, PO Box 3014, Syndal, Victoria, 3149, Australia
