Journal of the American Mosquito Control Association, 12(2):155_166, 1996 Copyright @ 1996 by the American Mosquito Control Association, Inc.

FORUM

ADVERSE ASSESSMENTSOF GAMBUSIA AFFINIS: AN ALTERNATE VIEW FOR MOSQUITO CONTROL PRACTITIONERS'

HENRY R. RT]PP

1440 Mohawk Road, North Brunswich NJ 0g902_I4I l ABSTRACT' Adverse opinions on the introduction of Gambusiaaffinis for the control of larval mos- quitoes are reviewed. The sourcesspan a period of some 59 years un"d"o-" from a variety of sources. The principal opposition to the introduction of G. ffinis clmes from ichthyologists, although some mosquito researchers have expressedconcerns about the environmental impJct oi placing *rJ fisrr in habitatsto which it is not native. Questionsconcerning the appropriatenessofusing the fish are presented.

Although mosquito research and control per_ ical Control of Mosquitoes, Laird (19S5) pointed sonnel have been almost unanimous in their ap_ out that biocontrol of mosquitoes had been al_ proval of the use of Gambusia affinis for mos_ most exclusively concerned with G. affinis pior quito control, members of the ichthyological to commercial production of Bacillus thurin_ community have viewed introduction of tho fish giensis subspecies israelensis. Early successes into nonnative habitats with alarm because of with the fish had led to widespread dispersal. real and potential damage to these ecosystems. Laird pointed out concern by the World Health This paper presents a brief review of "recommendation idverse Organization leading to a that opinions so mosquito control personnel can have G. affinis not be introduced into new areas but another perspective to consider when planning that indigenous species be researched and to utilize the fish outside its native habitat. used." He noted application ofpredatory agents Meisch (1985) published a l4-page chapter on "destruction ^ should not result in of tirese aI_ Gambusia affinis affinis in the basic text for ready-present natural enemies of mosquitoes.', mosquito control workers, Biological Control of Although there are many larvivorous fish, Gam_ Mosquitoes. Writing from b mosquito corrt ol- busia has been the backbone of biocontrol for perspective, Meisch included a section on ..Nes_ one-quarter of a century, and Laird remarked ative Aspects," although these negativ" "ori_ that any other biological control agents could ments were vitiated by insertion of positive also have adverse impacts. The effect on non_ statements about Gambusia. Bay (19g5) wrote a "Other target organisms should be reduced. Thus. Laird chapter on Larvivorous Fishes', in this recommended cost/risk/benefit studies as an es_ same book. In the concluding chapter of Biolog_ sential part of any control program and conclud_ ' ed, most realistically, that no adverse impacts In the August 1995 issue of ttre AMCA Newsletter. would mean no mosquito control. President John Edman suggested that we might have a 1967 Bay had reviewed positive torum section in the Journal of the AMCA to promote ]n and neg_ ative aspects of the use scientific debate. Fortuitously, a controversiai manu- of Gambusia and later (1972) script was received from Henry Rupp, entitled ,.Ad_ reviewed opposition to use of Gambusia verse assessments of Gambusia affinis: an in West Africa, citing the anti-Gambusia alternative "most feel- view for mosquito control practitioners." This manu_ ings of ichthyologists.,, script appeared to be ideal for such a forum. Com_ Another source, long familiar to mosquito ments about Mr. Rupp's manuscript "A were solicited control workers, is guide to the use of the lrom a number of mosquito control or wildlife soe_ mosquito fish, Gambusia afinis, for mosquito cialists who were asked to give their views on the pios control" by Scholdt et al. (1972). Scholdt or cons of Mr. Rupp's contention that the use of dam_ "the warned. that impact busia for mosquito control may have more negative of the fish on t}te aquat_ rc envlronment effects than positive ones. These comments ,roi, *" cannot be underestimated as being published with the above paper. The response of there is good evidence that the indiscriminate our readers to a Forum section and suggestions for use of mosquito fish can be as detrimental as the topics would be appreciated.-R. frture Al Ward, nd- misuse of pesticides." Many of the same sentiments were echoed bv r55 156 Jouwar oF THEAMERIcAN Moseurro CoNrnor Assocrmot Vor-. 12, No. 2

"Time Laird (1977) who wrote, has proved that on Gambusia in Ecology and Evolution of Live- mosquitofish eventually became harmful in bearing Fishes, where they state: some areas to which they were introduced half "In summary, mosquito fish a century ago-the harm ranged from eating the almost invariably present a multitude of problems when introduced eggs of economically desirable fish, to endan- beyond their native range and offer no real compen- gering rare indigenous species." Later, in 1988, satory or biological control advantages. The species Laird cited Miura's work in the rice fields of should not be used as a larvivore, with native spe- California as indicative of the lack of adverse cies much preferable in that role whenever possible effects on one environment, an environment (e.g. Lloyd 1986). Mosquitofish are far too aggres- Harrington (personal communication) referred to sive and predatory to be indiscriminately spread "artificial." as Laird (1988) cited L6gendre's throughout the world without recognition of dangers 1937 article in the Bulletin Economique Indo- to native biota. An international ban on their use as a control agent is biologically chine, pointing out that the threat to indigenous appropriate and war- ranted." fish from Gambusia predated Myers's 1965 ar- ticle, which referred to events of the mid-l93Os, What are the reasons presented for this con- but seemed unaware of Sweetman's The Biolog- demnation? Let us look at some of the evidence ical Control of Insects (1936), which had pre- Courtenay and Meffe present in their review of ceded L6gendre by a year and would have been the literature dealing with the adverse affects of more readily accessible to mosquito control the use of Gambusia. practitioners in this country. Howeveq Laird re- The first complaint raised is that Gambusia is "Too minds us, much should not be expected, not really that effective in mosquito control and though, of such indigenous fish in natural waters that better control has been achieved with native of high taxonomic diversity," and indicated, in species (Table I ). As far back as 1949 Bates was "The relation to Aplocheilus panchax, omnivorous noting, success of practical operations fish can maintain themselves well in nature with- along these lines [biological control] has not out mosquito larvae, and Gambusia is indeed an been very remarkable." ln 1967 Harrington "very omnivore. He also cited Russian feeling that (personal communication) said, few en- "the general effectiveness of this fish has been tomologists have sound evidence of the alleged disappointing." efficacy of Gambusia where it has been intro- Speaking of biocontrol programs, Laird duced." (1988) commented on use of Aplocheilus latipes In Australia, Allen (1989) remarked on Gam- in Russia following tests of larval consumption busia's lack of efficacy in mosquito control, and rates of the fish in aquaria. in 1993 Dennis C. Haney of the National Bio- logical Survey of the U.S. Department of the "It "I is submittedthat biocontrol introductions of this Interior wrote (personal communication), kind, based upon a mosquito larvivore that is very think you will find that there is little or no evi- probably of far more catholic tastesthan univariate dence for Gambusia being particularly effective laboratory experimentscould reveal, are only likely in controlling mosquito larvae. In fact, almost to prove successful where mosquito production is all the evidence indicates that Gambusia is no unusually high in rather short-lived larval habitats better at controlling larvae than are native fish." harbouring a limited flora and fauna; or in more Recognizing there is testimony complex aquaticecosystems where preliminary hab- on both sides itat manipulation is undertakento give, for example, of the issue and Gambusia may not be so uni- Aplocheilus unrestrictedaccess to dense concentra- versally successful as we have been led to be- tions of larval Culicidae." lieve, let us tum to what ichthyologists consider a more serious issue. The failure to effect larval Service (1983) noted adverse environmental control is a concern of the mosquito control effects resulting in increased mosquito produc- community, but Gambusia's impact on nontar- tion from the use of Gambusia, and in 1995 he get organisms affects a broader community of reaffirmed his doubts about the value of Gam- interests. busia in control for disease prevention purposes. The concern with environmental impacts goes In 1985 Mahmoud, in a study of the efficacy of back more than half a century when in 1936 Gambusia for malaria control in the Sudan, con- Harvey Sweetman warned in his pioneering The cluded Gambusia, although effective in canals Biological Control of Insects: "an and ditches, was not efficient mosquito con- "Finally, it should always be held in mind that trol measure during the peak season of malaria " the introduction of any foreign animal is apt to cause transmission. repercussionson the native fauna in unexpected These caveats are mild compared to the opin- ways. Gambusia has been spread far and wide in ion presented by ichthyologists Courtenay and anti-mosquitowork, frequently in ignoranceofvalu- Meffe (1989) in the conclusion to their section able native species. Gambusia affinis and G. hal- JUNE 1996 FoRtrM: Aovense AssEssMENTsoF GAMBUSTA t57

Table l. Reports of unsuccessful mosquito control by Gambusia affinislholbroati (from Courtenay and Meffe !9891; see this publication for reference citations).

Locality Comments Reference Australia ft lr *g quito control than some native fishes. "I Australia believe their effect on mosquitoes has been negli- Lake (1971) gible. " california Mosquitofish can increase mosquito populations by Moyle (1976) eliminating other mosquito predators. California Experiment. Pupfish more effective mosquito predator Danielson (196g) in emergent vegetation; mosquitoes a problem in Owens Valley after Owens pupfish eliminated, de_ spite mosquitofi sh introduction. California Experiment. Native Cyprinodon macularis is equal in Walters and Legner osquito control and not as dangerous. (l9g0) Iraq Native fish also consumes mosquitoes; mosquitofish Sharma and Al-Daham lose efficiency in presence of other organisms. (lg7g) Males are poor consumers. Italy Mosquitofish unsuccessful in eliminating Anopheles Hildebrand (1930) from running waters. Japan Mosquitofish rEduced the number of larval, but not Hirose et al. fl9g0) adult mosquitoes. Missouri Mosquitofish are little more effective in mosquito pflieger (lgls) control than the natives they replace.

brookii are for their size, among the most voracious tion of Gambusia (Table 2). Lynch (1991) re- and destructive of fishes, and although no reports of ported on the impact of Gambusia on the plains damage to the young of valuable food fishes in area" topminnow, Fundulus sciadicus, in Nebraska into which they have been introduced have ap_ in peared, it is possible that introduction an experimental release program. His comments into certain ..Most places will prove to have been a mistake.,' on such research are instructive: experi- ments are done under controlled circumstances Myers, writing in 1965, noted some 30 years "the where the experimenter has some notion of what previously crew ofthe California State Fish to expect .... The release of self-replicating and Game Department's black-bass hatchery at agents into the environment is fundamentallv Friant had to discontinue using 'forage Gambusia is a different. because as soon as they are released, fish' with which to feed the bass. Gam_ any controls are lost." busia was destroying a large proportion of the Gambusia's pugnacious and omnivorous na_ young bass!" He also reported removing Gam_ ture is not helpful to other species, reproduction busia from a pond shared with goldfish und,e"_ (Myers 1965). Eggs and fry, even of intensive_ ing the goldfish population double and redouble care fish such as the Centrarchidae, are grist for in 2 years. their reproductive mill (Myers 1965), and com- The effectiveness of Gambusia as a predator petition for resources is not the problem; it is as well as its omnivorous feeding patterni makes predation (Courtenay and Meffe 1989). it a hazard to native species and any other aquat_ Australian ichthyologists talk of Gambusia al_ ic organisms that can be eaten (iourtenav and most as Australian agriculturists speak of the Meffe f 989). Cambusia does not specialize in rabbit. Brought in with good intentions, the fish mosquito larvae and pupae (Harrington and Har_ has spread widely across the continent and is rington 1961, Myers 1965, Washino and Hoka_ viewed as a threat to native species (Allen t9g9, ma 1967, Meisch 1985). Myers (1965) reported Arthington and Lloyd 1989); and, in coastal the loss of 5 species subsequent to the intioduc_ regions of southeast Asia, Bardach et al. 0g7Z) tion of Gambusia: Poeciliopsis (USA), sp. cite Gambusia as well as the common guppy as Ap,locheilus panchax and, phenacoitethus sp. pest species. Wildekamp (1993) notes rhe impact (Thailand), Gulaphallus sp. (philippines), and of the introduction of Gambusia on 4 speciei of Micropanchax sholleri (Lower Nilel. Courtenav Aphanius in the first of a multivolume iet about and Meffe (1989) listed other species and oi_ killifish of the world: A. anatoliae transgradiens ganisms reduced or eliminated by the introduc_ (Turkey), A. apodus (Algeria), A. fasciatus 158 JoURNAL oF THE AMERTCIN Mosqurro CoNrnol AssocrenoN Vol. 12,No.2

Table 2. Examples of negative impacts of nonnative Gambusia affinislholbroolci populations on local fishes (from Courtenay and Meffe [1989]; see this publication for reference citations). Native species Effect Reference 2O taxa: Cyprinidae (l), Cyprinodontidae (6), Reduction or elim- Schoenherr (1981) Poeciliidae (9), Neostethidae (3), Gasterostei- ination dae (1). Cyprinus carpio No reproduction Sreenivasan and Natara- jan (1962) Crenichthys baileyi Reduction Deacon et al. (1964) Cyprinodon calaritanus Elimination Missiroli (1948) Cyprinodon bovinus, Gambusia gagei, Lepidome- Partial elimination Miller (1961) da mollispinis pratensis due to mosqui- tofish Rhinichthys osculus Elimination Deacon et al. (1964) Hype re leotris galli, Me lanotaenia fluviatilis Reduction Arthington et al. (1983) Eleotridae, Ambas sus castelnaui, Nannope rca Reduction Lloyd (1984) australis, Craterocephalzs spp., Melanotaenia fluvitalis, Pseudomugil signifer, Retropinna se- moni

(Rhone Delta-{amargue), and A. iberus (Spain duced by introduction of exotics, then cultiva- and Algeria). tion and stocking of native egg-laying fish, it is More recently, Gratz (personal communica- assumed, take more effort. Live-bearers are sup- tion) wrote of a fish expert for FAO coming into posedly easier to raise than egg-layers; fry are his WHO office complaining about the stupidity largec free swimming, and feeding at birth; they of people releasing Gambusia into an area where grow more quickly and become predators faster. he was trying to develop fish populations to feed However, in this writer's experience, egg-laying indigenous populations only to find the Gam- minnows (Pimphales promelas "Rosy") pro- busia eating the eggs of the desired fish species. duced thousands of fry in a lo-ft. garden pond It is instructive to compare the comrnentary over the course of a summer with minimal pa- on Gambusia in 2 standard field guides to fresh- rental predation. Gambusia have proved easy to water fishes of North America. The Audubon use, require minimal training for application, guide (Boschung et al. 1983) notes Gambusia and, more importantly, have been thought safe. has been widely introduced for use in mosquito These fish have good public relations value control. The Peterson guide (Page and Burr (Duryea 1993), showing our ability to not be "Introductions l99l) is much more specific: of totally reliant on pesticides in a period of this species, often for mosquito control, have chemophobia. Further, their easy availability has caused or contributed to the elimination of many allowed us the luxury of not having to seek oth- populations of flshes with similar ecological re- er tools. quirements. Introductions into western drainages One question, never effectively raised, be- have been especially deleterious to the survival canse Gambusla was a biological organism, was of certain rare fishes." Is it possible the Audu- the question automatically asked about any pes- bon Society with its known antipathy to pesti- ticide: What are the nontarget effects? Were a cides is willing to tolerate the use of Gambusia label sought for use of Gambusia as a pesticide as the lesser of two evils? today, one suspects it might well prove unac- Are hazards posed by use of Gambusia real? ceptable to the Environmental Protection Agen- One suspects arguments presented by ichthyol- cy because of adverse environmental impacts, ogists must have substance, and the evidence particularly as those effects would be revealed presented is real. Critics of Gambusia ask why in the review process. Being a highly goal-ori- native species are not used to control mosqui- ented community, one suspects mosquito re- toes. Several reasons may be posited. If a fish is search and control people have looked at Gam- in its native habitat and is an effective predator, busia with rose-colored glasses. Another ques- such as saltmarsh killies, the need to institute a tion we have not asked ourselves is that if Gam- biocontrol program is unnecessary. However, busia is so effective a predator, how is it there when it comes to habitats where native fish are are so many mosquitoes in areas that are its na- not as numerous or their numbers have been re- tive habitat? I Jur.rEree6 FORUM: ADVERSE AssEssMEl.ffs oF GAMBUSIA 159

Mosquito research and control people find In: H. C. Chapman (ed.). Biological control of mos- their origins in the disease-prevention programs quitoes. Am. Mosq. Control Assoc. Bull. Volume 6. of the past. Their concerns have been oriented Fresno, CA. toward human well-being, so it is not surprising Boschung, H. T., Jr., J. D. Williams, D. W. Gotshall, D. K. Caldwell and M. C. Caldwell. 1983. The Au- they welcomed a tool such as Gambusia. But, if dubon Society field guide to North American fishes, health concerns are given as a reason for intro- whales & dolphins, pp. 516-517. Alfred A. Knopf, duction of Gambusia, it should be remembered New York, NY. that Gambusia is not an effective control against Courtenay, W. R., Jr. and G. K. Meffe. 1989. Small vector species such as Aedes aegypti, Ae. albo- fishes in strange places: a review of introduced poe- pictus, Coquillettidia perturbans, or Culiseta ciliids, pp. 301-331. 1n.'G. K. Meffe and E F Snel- melnnura. One should also note that if disease son, Jr. (eds.). Ecology and evolution of livebearing is a factor, then control should be effected as fishes (Poeciliidae). Prentice Hall, Englewood promptly and as completely as possible. It is Cliffs, NJ. Duryea, R. 1993. County experience with mosquito- more responsible to use an insecticide whose en- fish. Proc. N.J. Mosq. Control Assoc. 80:5-9. vironmental breakdown is a known factor rather Harrington, R. W and E. S. Harrington. 1961. Food than a biological agent whose environmental selection among fishes invading a high subtropical fate can only be speculated. salt marsh: from onset of flooding through the prog- The recognition that there is disagreement ress of a mosquito brood. Ecology 42:646-666. about the :use of Gambusia ard the willingness Laird, M. 1977. Enemies and diseases of mosquitoes. of researchers (e.9., Laird, Meisch, and Service) Their natural regulatory significance in relation to pes- to recognize and present opposing points of ticide use, and their future as marketable components view make it obligatory that studies, other than of integrated control. Mosq. News 37:331-339. Laird, M. 1985. Conclusion, pp. 216-218. ln; H. C. specifically scientific work on Gambusia, take Chapman (ed.). Biological control of mosquitoes. Am. into consideration alternative opinions. Other- Mosq. Control Assoc. Bull. Volume 6. Fresno, CA. wise such studies risk compromising their integ- Laild, M. 1988. The natural history of larval mosquio rity. Our thinking over the years has been con- habitats, pp. 467 -47 2. Academic Press, I-ondon. ditioned by what mosquito researchers, not ich- Lynch, J. D. 1991. A footnote to history: a tale of two thyologists, have told us about Gambusia and its fish species in Nebraska. Nebraskaland July l99l: efficacy, particularly about the numbers of lar- 50-55. vae consumed in aquaria or other artificial hab- Mahmoud, A. A. 1985. Mosquitofish Gambusia afin- itats. We need to expand our horizons to learn is holbrooki as a malaria vector control agent in Ge- zira irrigation canals of the Sudan. J. Am. Mosq. what other specialists in the aquatic world have Control Assoc. l:524-526. to say about Gambusia, and we need to factor Meisch, M. V. 1985. Gambusia affinis ffinis, pp.3- in that knowledge so we can avoid ecological 16. In: H. C. Chapman (ed.). Biological control of mishaps is areas not already destablized by in- mosquitoes. Am. Mosq. Control Assoc. Bull. Vol- troduction of exotics. ume 6. Fresno, CA. Myers, G. S. 1965. Gambusia, the fish destroyer. Tiop. Fish Hobbyist 13:31-32, 53-54. REFERENCES CITED Page, L. M. and B. M. Burr. 1991. A field guide to Allen, G. R. 1989. Freshwaterfishes of Australia, p. freshwater 'lFH fishes: North America north of Mexico 214. Publications,Neptune City, NJ. (The Peterson Field Guide Series@), pp.235-237. Arthington, A. H. and L. N. Lloyd. 1989. Introduced Houghton Miflin Co., Boston, MA. poeciliids in Australia and New Zealand, pp. 333- Scholdt, L., D. A. Ehrhardt and A. G. Michael. 1972. 348. In: G. K. Meffe and F, E Snelson, Jr. (eds.). A guide to the use of the mosquito fish, Gambusia Ecology and evolution of livebearing fishes (Poeci- affinis, for mosquito control. Navy Environmental liidae). PrenticeHall, Englewood Cliffs, NJ. and Preventive Medicine Unit No. 2., Norfolk, VA. Bardach, J. E., J. H. Ryther and W. O. Mclarnay. Service, M. W. 1983. Biological control of mosqui- 1972. Aquaculture: the farming and husbandry of toes-has it a future? Mosq. News 43:113-120. freshwater and marine organisms, pp. lo9-l lO. Service, M. W. 1995. Can we do without pesticides? John Wiley & Sons, New York, NY. A summary. J. Am. Mosq. Control Assoc. ll:29o- Bates, M. 1949(1970). The natural history of mos- 293. quitoes,pp. 158-159,3O'7. Peter Smith, Gloucester, Sweetman, H. L. 1936. The biological control of in- MA. sects, pp. 3Zl-325. Comstock Publ. Co., Ithaca, NY. Bay, E. C. 1967. Potential for naturalist control of Washino, R. K. and Y. Hokama. 1967. Preliminary mosquitoes.Proc. Pap. Calif. Mosq. Control Assoc. report of the feeding pattern of two species of fish 35:34-37. in a rice field habitat. Proc. Pap. Calif. Mosq. Con- Bay, E. C. 1972. A preliminary assessmentof the po- trol Assoc. 35:84-87. tentialities of larvivorous fishes for Anopheles con- Wildekamp, R. H. 1993. A world of killies: atlas of trol in West Africa. WHONBC/72.4O3. World the oviparous cyprinodontiform fishes of the world, Health Organization,Geneva, Switzerland. Volume l, pp. 22-56. The American Killifish As- Bay, E. C. 1985. Other larvivorous fishes,pp. 18-24. sociation. Mishiwaka. IN. 752 JounNel oF THEAMERrcet Mosqurro Covrnor_ Assocreuorl Vol. 12.No.4

ERRATUM

"Adverse In my Comments on Assessments of Gambusia affinis" that appeared on pages 160-16l of theJune 1996issue(Vol. 12,No.2) of theJournalof theAmericanuosquttoControtAssociation,the "the phrase beginning on line 3l of the second paragraph, highest number of mosquito larvae were in ". paddies stocked with Mf;" should have read . stocked with blackfish" tEd Nbte: This alters the "a meaning of the remainder of the sentence, severe indictment of the degree of control provided "1. by the species.

N. G. Gratz JoURNAL oF THE AMERIclr.r Mosquno CoNTRoL AssocrlrroN Vol. 12, No. 2

COMMENTS ON "ADVERSE ASSESSMENTSOF GAMBUSIA AFFINIS"

Perhaps the greatest service that Mr. Rupp has culture of the species difficult, Cech and Moyle done in writing this article is to show the im- (1983) suggested that 2 species of fish native to portance of periodically scrutinizing any meth- California, Orthodon microlepidotus and Lavi- od, material, or approach to vector control, even nia exilcauda, which have high fecundities, one whose efficacy has been taken for granted could be possible alternatives for rice field mos- as long as has the use of predacious fish. Vir- quito control, although more research was need- tually every manual, brochure, or guide on mos- ed. Bence (1988) observed that in some of his quito control urges the use of predacious flsh, studies, the introduction of mosquitofish in- usually Gambusia affinis, wherever possible; creased the abundance of mosquito larvae. For these reviews are usually written by vector con- 3 years Todd and Giglioli (1983) monitored dai- trol specialists, most of whom have little de- ly the number of Gambusia in a mangrove tailed information on the population dynamics swamp on Grand Cayman Island and found that of G. affinis or of the scattered literature quali- Gambusia and other small fish were unable to tatively assessing the efficacy of the control ef- control Aedes taeniorhynchus. Other studies fected by the species or its effect on nontarget could be referred to that resulted in similar con- organisms. clusions. With the abundance of papers from dif- In addition to the references provided by Mr. ferent geographical and ecological areas that re- Rupp on the adverse impact of mosquitofish on port poor or no control of mosquito larvae (if other fish species as well as on other mosquito not an increase in larval numbers) after the predators and on the failure of the mosquito fish stocking of G. ffinis, one must ask, as Mr. Rupp to provide a satisfactory level of mosquito larval does, why continue to use mosquitofish at all? control, some additional references are given be- Overreliance on laboratory studies is part of the low that illustrate the failure of G. affinis to pro- reason but part must be ascribed to the sheer vide satisfactory control of mosquitoes in dif- inertia of control workers who, having used fish ferent larval habitats. Many of these papers also for a long time, take it for granted that they are observe that use of G. affinis puts other species effective and are quite unaware of the effect on of fish or invertebrate predators at risk (Bence nontarget fauna. Mr. Rupp's question as to whether the use of G. affinis for mosquito 1982, Walton and Mulla l99l). A field trial car- con- be must receive ried out by Blaustein (1992) in California re- trol should continued a nesative reply. sulted in inadequate control of rice field-breed- ing of Anopheles and Culex mosquito larvae by G. affinis and Lepomis cyanellus, the green sun- REFERENCES CITED fish. The fish depressed the number of notonec- Bellini, R., R. Veronesi,and M. Rizzoli. 1994. Etfi- tid mosquito predators to a level that may have cacy of various fish species(Carassius auratus, Cy- to the failure contributed of the fish to control prinus carpio, Gambusiaaffinis lBaird and Girardl) mosquitoes. The author concluded that the con- in the control of rice field mosquitoes in northern ventional wisdom that fish are efficient mosquito Italy. Bull. Soc. VectorEcol. l9:87-99. predators in rice fields in not necessarily correct Bence, J. R. 1982. Some interactions of predacious and that they probably consume efficient mos- insects and mosquitofish (Gambusia affinis)'. a re- quito predators. Mulla et al. (1979), observed view of somerecent results. Bull. Soc.Vector Ecol. 7:41-44. that G. affinis induces drastic alterations in the Bence,J. R. 1988. Indirecteffects and biologicalcon- ecosystem, some of which are irreversible. Cech trol of mosquitoes by mosquitofish. J. Appl. Ecol. and Linden (1987) stocked paddies with both 25:5O5-521. mosquitofish (MF) and Sacramento blackfish Blaustein, L. 1992. Larvivorous fishes fail to control (BF) and then drained the paddies 12 wk later mosquitoesin experimental rice plots. Hydrobiolo- to count the remaining fish and invertebrates; the gia 232:219-232. highest number of mosquito larvae were in pad- Cech, J. J., Jr. and A. L. Linden. 1987. Comparative dies stocked with ME a severe indictment of the larvivorous performancesof mosquitolish, Gambu- juvenile degree of control provided by the species. Bel- sia affinis and Sacramentoblackfish, Or- thodon microlepidotus in experimental paddies. J. lini et al. (1994) stocked 3 species of fish in rice Am. Mosq. Control Assoc.3:35-41. fields in northern ltaly, including G. affinis, and Cech,J. J., Jr. and P B. Moyle. 1983. Alternativefish found that none of them had any significant im- speciesas predatorsfor rice field mosquitoesin Cal- pact on mosquito densities. As the low fecundity ifornia.Bull. Soc. VectorEcol. 8:107-l lO. and internal fertilization of G. afinis make mass Mulla, M. S., G. Majori, and A. A. Arata. 1979. Im- JUNE1996 Fonuv: ADVERSEAssEssMEl.,rrs oF GAMBUSIA

pact of biological and chemical mosquito control indigenous species of fish (Service 1983, Lloyd agents on nontarget biota in aquatic ecosystems. 1987). Residue Rev. 7 l:l2ll-1273. Whether this fish leads to desired levels of Todd, R. G. and M. E. C. Giglioli. 1983. The failure mosquito control in many situations is still de- of Gambusia puncticulata and other minnows to bated. Probably an accurate assessment is re- control Aedes taeniorhynchus in a mangrove swamp (1930) ". on Grand Cayman. W.I. Mosq. News 43:419-425. vealed in a statement by Kligler that . . Walton, W. E. and M. S. Mulla. 1991. Integrated con- their usefulness as larvae-destroyers under local trol of Culex tarsalis tsing Bacillus sphaericus and conditions where vegetation is abundant and mi- Gambusia affinis: effects on mosquitoes and nontar- crofauna rich enough to supply their needs with- get organisms in field mesocosm. Bull. Soc. Vector out great trouble, is limited. In moderately clear Ecol. 16:2O3-221. canals, on the other hand, or in pools having a limited food supply, they yielded excellent re- Norman S. Gratz sults...." 4 ch. du Ruisseau contemporary uses of and successes with 1291 Commugny Gambusia were discussed in detail by Legner Switzerland (1995). It was concluded that although the mos- quitofish has been useful for control of mosqui- toes in a number of situations, clearly there A variety of fish species are used for the bi- are environmental drawbacks to its use. This fish ological control of mosquitoes, which together probably never would have been intentionally constitute the major successes in biological con- introduced into foreign areas if today's environ- trol of these Diptera. However, their application mental concerns existed in the early 1900s. A is limited to relatively permanent bodies of wa- major objection to mosquitofish has been their ter where their impact on the target species var- direct impact on native fishes through predation, ies with climate and locality. Bay et al. (1976) or their indirect impact through competition. noted that many kinds of fish consume mosquito More than 30 species of native fish have been larvae, but only a few species have been manip- adversely affected by the introduction of Gam- ulated to manage mosquito populations (note: busia. Introductions of Gambusia have also re- please refer to Legner [1995] for references cit- duced the numbers ofother aquatic invertebrates ed here). coinhabiting the same waters. However, there The mosquitofish, Gambusia affinis, is the are no reports ofthis species, through its feeding best-known biological mosquito control agent. on zooplankton, causing algal blooms outside of Native to the southeastern United States, eastern the aquarium environment. The deployment of Mexico, and the Caribbean area, this species fish species other than Gambusia in mosquito was first used as an introduced agent for mos- control often has been restricted by inadequate quito control when transported from North Car- supplies, or the cost of tropical and semitropical olina to New Jersey in 1905 (Lloyd 1987). Later species obtainable from commercial sources has it was introduced to the Hawaiian Islands to con- been too high for stocking large acreages. trol mosquitoes, and during the next 7O years to Research into the biological control of mos- over 5O countries. The mosquitofish is the most quitoes has established that the natural enemy widely disseminated biological control agent in component is frequently responsible for signifi- the world. Many of the introductions were made cant population reduction and is indispensable to control Anopheles species that were carrying to integrated control. Further advancements in malaria. Hackett (1937) described the usefulness the control of mosquitoes must continue to em- of the mosquitofish in malaria programs control brace a sound appreciation for the natural con- in Europe, noting that the fish had a definite im- trol component and nurture ways to allow its pact on the suppression of the disease. Tabib- maximum expression. zadel et al. (197O) reported an extensive release program in Iran and concluded that the fish was an important component in malaria eradication. REFERENCE CITED Although Sasa and Kurihara (1981) and Service Legner, E. E 1995. Biological control of Diptera of (1983) judged that the fish had liule impact on medical and veterinary importance.J. Vector Ecol. the disease and that most evidence was circum- 2O:59-12O. stantial, Inci et al. (1992) observed a 5OVo re- E. F Legner duction in malaria cases in southeastern Turkey Department of Entomology after the fish became established. Gambusia no University of California longer is recommended by the World Health Or- Riverside, CA 92521-0314 ganization for malaria control programs, primar- ily because of its apparent harmful impact on t62 JounNar-or rHr Avnnlcat Moseurro Corqrnol AssocrerroN Vor. 12. No. 2

The paper by Henry Rupp is appropriate and But this means that we are addressing the wrong long overdue. For decades, ichthyologists and problem. Rather than scattering mosquitofish ecologists have known, written, and spoken across a degraded landscape, we should be using about the clear environmental hazards of Gam- those resources and energies to restore the nat- busia affinis and Gambusia holbrooki when ural environments, including their native larvi- transplanted outside their native ranges, but the vores. We are heaping one problem on top of mosquito control community seems to either not another, while ignoring the underlying causes. have heard the message or has chosen to ignore Mosquito problems are frequently an indication it. I support the comments and observations of of environmental degradation, and should be Mr. Rupp in their entirety, and wish to add a treated as such. few observations of my own. In conclusion, I reiterate my earlier claim, cit- First, I applaud and admire the mosquito con- ed by Henry Rupp, that G. affinis and G. hol- trol community for wanting to find safe alter- brooki shorld not be used beyond their native natives to pesticide spraying. We are in agree- ranges, and their intentional movements should ment that such spraying is to be avoided when- in fact be banned. Yes, we need to address mos- ever possible, and that pesticides have damaging quito control in some areas, but the use of Gam- effects, including dedth of nontarget species. busia is often ineffective and has repeatedly However, the use of mosquitofish has the same been shown to be dangerous to other biota. effect (although not as immediate and visible), Gambusia is not the answer to the problem. and it does not break down over time. Yes. an Gary K. Meffe alternative to spraying is much desired, but I am Savannah River Ecology Laboratory afraid that Gambusia is not the answer. There University of Georgia may be no easy answer. But wanting something Drawer E to be true does not make it so. It would be won- Aiken, SC 29802 derful if Gambusia was a silver bullet with no harmful side effects, but that is not the case. I simply ask you to look at the evidence in an The bottom line to Henry Rupp's article on unbiased way. Gambusia affinis is that, as in all professions, Second, many argue that Gambusia are safe mosquito control practitioners should always "contained" when used in the and artificial en- know their tools. I too have extensively re- vironments in which they often are stocked, and viewed the literature cited by Mr. Rupp up until will not escape to the wild. This is untrue. Mos- 1973, and am personally familiar with some of quitofish have an uncanny ability to move or be the work and observations later published. What moved around-by people, by floods, perhaps is noteworthy is how few, speculative, and in- by fish-eating birds who drop them in flight- consequential are the statements in Mr. Rupp's but they do move. For example, I and others tables related to the argument of adverse effects. have found them in extremely remote areas of Given the concerns of ichthyologists I find this the desert southwestern USA that were thought lack of documentation over the past 22 years of to be safe and unreachable. They always seem renewed biological control emphasis, including to move, eventually. the use of larvivorous fishes, to be remarkable. Third, let's examine the efficacy of mosqui- More significant is what the documentation says tofish as larvivores. Do they control mosquitoes? about the failures of Gambusia. Voluminous as I suspect that under certain circumstances, es- it is. the literature on larvivorous fish cites rel- pecially in artificial or highly disturbed habitats, atively few well-documented successes of con- they are effective. But if they are effective, then sequence. by definition they are good predators. Because Tfy to breed mosquito larvae in a water we know they eat many items besides mosquito trough populated with Gambusia. [t can't be larvae, then obviously they eat a lot of other done! Try to reduce biting populations of adult things in their new homes. By definition then, mosquitoes in regions of puddled swails, marshy introduced mosquitofish are dangerous because habitats, intermittent flooding, tree holes, and a they are predators on many invertebrate and ver- host of other larval sources by introducing Gam- tebrate species beyond the target species. busia. lt won't happen! Somewhere in between Fourth, I have seen it argued that Gambusia these extremes Gambusia meets its limits. are used primarily where native species are not Gambusia has been variously described by "most present. This means either an artificial habitat, ecologist Frank Wilson as the widely dis- or a natural habitat so disturbed or degraded that seminated natural enemy in the history of bio- native larvivores have disappeared. Because logical control," and by Louis Krumholtz as "the Gambusia affinis and G. holbrooki are so toler- most widely distributed of any freshwater ant. thev can survive where other fishes cannot. fish." With so few documented cases of adverse JUNE 1996 Fornu: AovEnsg AssEssMENrs oF GAMBUSIA r63 effects by Gambusia over more than a half cen- [eds.]. Fishes in North American deserts. Wiley tury one wonders at the validity of ichthyolo- Interscience. New York. NY). gists' concerns. Even the noted poeciliid expert, Since the turn of this century, Gambusia af- the late Carl Hubbs, wrote to Marston Bates in finis has been introduced into about 60 countries 1968 "We have convicted the little criminal in efforts to control mosquitoes and as a con- lGambusial on circumstantial evidence to a sequence has become the world's most widely large degree." Perhaps the most cited indictment distributed fish species. There can be no disput- "Gambusia, is George Myers's paper the fish ing that the mosquitofish can cause ecological destroyer," again referred to by Mr. Rupp. My- damage, and since about 1982 the World Health ers relates that eliminating Gambusia from a Organization has not supported its introduction goldfish pond resulted in a stable population of as a biocontrol agent. With hindsight it should 1l goldfish increasing to 50 fish within 2 years. never have been exported outside the USA nor The key word should be stable. The 2 species disseminated within the USA. had been coexisting and Myers simply reduced Many publications state that G. affinis reduces the species diversity. mosquito larval numbers in breeding places, but Granted, there are isolated habitats into which to be an effective control agent it must reduce Gambusia should never be placed, for example, biting densities of the adults andlor reduce dis- desert pools with relict pupfish populations and ease prevalence. In the great majority of control in questionable habitats where introduction of programs employing larvivorous fish there has Gambusia would contribute negligibly to overall been no attempt at such critical evaluation. Nev- mosquito suppression. On the other hand, Clark ertheless, there are reports that fish, sometimes Hubbs, in 1972, speaking of recreational fish- G. affinis, have reduced malaria transmission in eries in artificial impoundments commented parts of Afghanistan, China, Ethiopia, Iran, Ko- "The available fauna may be limited and mal- rea, Somalia, Turkey, and the Ukraine. But in adapted, making supplementation by exotics most of the trials anopheline breeding has been beneficial and inevitable." The same logic restricted to discrete more or less permanent lar- should also apply to species used to aid human val habitats such as water reservoirs in dry ar- health and comfort. eas, control has been very local, and there is Gambusia has had more than ample oppor- usually no evidence of sustained disease control. tunity to star in the cavalcade ofintroduced spe- In other instances fish have been used in inte- cies disasters, but it has never been more than a grated control operations so it becomes difficult bit player. Prudent mosquito control practitioners to evaluate claims that fish have played an im- need only cast this player in its proper role. portant role in reducing biting densities or dis- ease transmission. Ernest C. Bay In a recent symposium on the future of vector 617 5th Avenue NW control without chemicals (Service. 1995. J. Puyallup, WA 98371 Am. Mosq. Control Assoc. ll:247-293), I was forced to conclude that chemicals will continue to play the dominant role in the foreseeable fu- I agree with much of what Mr. Rupp says ture in vector control operations. In conclusion, about the adverse effects arising from the use of the benefits of fish, including G. affinis, in con- Gambusia affinis to control mosquitoes. [n fact, trolling mosquitoes have frequently been grossly I think a stronger case could have been made exaggerated and often promoted by those with against the fish. Important references reviewing little ecological understanding of population dy- the successes and and failures of G. afinis in namics or pest management. The value of flsh reducing mosquito populations and the adverse as a practical method of mosquito control re- ecological effects that can be caused that were mains, at best, questionable. not cited by Rupp include Lacey and Lacey (1990. J. Am. Mosq. Control Assoc., 6[Suppl. M. W. Service 2l:l-93); Lacey and Orr (1994. Am. J. Ttop. Liverpool School of Tfopical Medicine Pembroke Place Med. Hyg. 5o[Suppl.]:1-159); Irgner (1995. J. Liverpool L3 Vector Ecol. 20:59-12O); Lloyd (1987. pp. 15G- 5QA England 163. In:'L D. St. George, B. H. Kay and J. Blok [eds.] Proc. Symp. Arbovirus Res. in Australia 4. Queensl. Inst. Med. Res., Brisbane); Reuben We applaud the efforts of this journal to pro- et al. (1990. pp. 139-158. In: C.F. Curtis [ed.]. mote scientific debate by presenting controver- Appropriate technology in vector control. CRC sial and critical views on aspects of mosquito Press, Boca Raton, FL); Schoenherr (1981. pp. control (e.g., Rupp's article). As researchers in 173--203. 1ll.' R. J. Naiman and D. L. Stolz fish biology, aquatic ecology, and biological 164 JoUnNel op rne AnmnrcAN Moseutro CoNrpoL AssocIATroN Vor. 12,No.2 control of mosquitoes, we are pleased to com- systems such as rice fields with water/fish bar- ment on Mr. Rupp's assessment of the use of riers that contain relatively simple assemblages mosquitofi sh for biocontrol. of organisms and few naturally occurring larvi- Rupp contends that mosquitofish have been vores. ineffective mosquito control agents. Viewed To address the lack of information, miscon- from a global perspective, this is generally true; ceptions, and controversies discussed by Rupp in many instances, mosquitofish planted for and others, and to provide biological back- mosquito control purposes have not effectively ground and operational guidelines for mosquito reduced mosquito populations. However, the lif control personnel planning to use mosquitofish, erature Rupp cites either contains no original we (in collaboration with R. H. Piedrahita, De- data or quantitative evaluations of biocontrol ef- partment of Biological and Agricultural Engi- ficacy (e.g., Lake 1971, Pflieger 1975, Moyle neering, UC Davis) have recently completed a 1976, Grant 1978, Allen 1989), is difficult to book entitled Mosquitortsh: Biology, Culture, assess(e.g., Danielson 1968, Hirose et al. 1980), and Use in Mosquito Contol. The book was or is misinterpreted or misrepresented (e.g., Hil- published by the California Mosquito and Vector debrand 1930, Mahmoud 1985). Mosquitofish Control Association and the University of Cali- performance as a biocontrol agent deserves hon- fornia Mosquito Research Program. A continued est and quantitative evaluation. Many of the ex- commitment to study, publish, and discuss all amples of biocontrol failure can be attributed to aspects of the use of mosquitofish in biological ignorance of mosquitofish biology and of the control will produce greater understanding and ecology of the stocked aquatic system, and re- further improvements in the use of these fish for sultant stocking of mosquitofish into inappropri- mosquito control. Rupp suggests that the use of ate habitats and,/or use of inadequate or incorrect mosquitofish for mosquito control may have had operational methods. more negative effects that positive ones. This Rupp also contends that mosquitofish have may have been true in the past, but we propose had negative impacts on aquatic ecosystems to that it does not have to be true in the future. which they have been introduced. Ignorance of Christina Swanson mosquitofish biology, unrealistic expectations Joseph J. Cech, Jr. about their performance as biocontrollers, and Department of Wildlife, Fish, and inadequate appreciation of the consequences of Conservation Biology the introduction of a nonnative fish into natural University of California aquatic habitats have been directly responsible Davis, CA 95616 for the very real ecological damage wrought by these fish in many parts of the world. Mosqui- tofish are tough, aggressive, fecund, omnivorous Having been a broad-spectrum ecologist for "in- predators, a classic profile for a successful half a century, I am in accord with much of what vading species." This is why, under the appro- Henry Rupp has to say about the eastern mos- priate circumstances, they can be effective bio- quitofish Gambusia affinis (and its western rel- control agents. The objective of mosquito con- ative G. holbrooki). However, although having trol using mosquitofish should be to select and long recognized that omnivorous mosquitofish apply the fish in those specific circumstances and such nonselective chemical pesticides as where the possibility of control is greatest and DDT have been used unwisely against Culicidae adverse impacts on the aquatic system are min- (always excepting in times of human health cri- imized. The fact that mosquitofish ale not exclu- ses), I reject the urging of some environmental sively mosquito predators and therefore impact extremists that field applications of our only use- nontarget organisms probably contributes to ful pre-Bti biocontrol larvicides must be their success in providing long-term, persistent "banned." "Ban!" control in habitats where mosquito larvae may is one of the more detestable buzz be present only periodically. Mosquitofish im- words of environmental activists. who are un- pacts on native larvivores ztrea serious concern aware of major past successes and future pros- that have not yet been well studied or quantified pects for Gambusia spp. as valuable biological (but see Bence. 1988. J. Appl. Ecol. 25:-5O5- elements in integrated control methodologies for 521), despite frequent anecdotal and interpretive the reduction of populations of mosquito pests comments; this issue deserves further attention and vectors. You will notice that I continue to "control," and exemplifies the potential pitfalls of mosqui- use the word which still figures in "manage- tofish introductions into natural aquatic systems our Association's title, rather than with well-established, complex assemblages of ment" as preferred by an increasing proportion organisms. In general, mosquitofish are most ef- of practical mosquito suppression specialists (in- fective in artificial aquatic habitats and agro-eco- cluding Rupp 1995). JUNE 1996 Fonrna: ADVERSEAssessr,reNrs oF GAMBUSIA t65

Integrated larval control had often been prac- Laird (1968) of highly successful employment ticed in due understanding of possible adverse of mosquitofish against malaria vectors. Thus, consequences of using mosquitofish, long before soon after World War I, during which malaria Myers (1965) so dramatically fingered them as had been a crucial factor in stalemating the Mac- fish destroyers. In the absence of plentiful in- edonian Campaign, Grassi and Sella evaluated vertebrate prey in laboratory tanks and hobby- the role of indigenous larvivorous fish for ists' aquaria, Gambusia spp. may attack the eggs anopheline control in the Pontine Marshes, an (and adults if of suitable size) or other fish shar- extensive area southeast of Rome that had been ing such confined quarters. To a lesser extent notoriously malarious for 2 millenia (Hackett than in the latter, this undoubtedly happens in 1937).ln sharp contrast to Courtenay and Meffe nature too. Howevet concern over such unin- (1989) and Arthington and Lloyd (1989), who tended consequences has lately been overblown, urged the superiority of endemic larvivorous fish especially since Myers (1965), in those heady to that of mosquitofish, Grassi (1923) and Sella early days of environmental activism, used ver- (1926) discovered (prior to importing G. hol- bal overkill of Carson (1962) kind to arouse an- brooki) that little could be expected of native tagonism against mosquitofish, as was just being larvivores. Gerberich and Laird (1968) provided done against DDT following the publication of many annotations on similar successes with Silent Spring. The latter's chapter headings were mosquitofish against malaria vectors in the Adri- carefully chosen for maximal shock-horror ef- atic Islands off the Dalmation coast from 1928; "Elixers "Be- fect, for example, of Death" and in Greece from 1928 when more than a million yond the Dreams of the Borgias." Her book and refugees had crowded into the most malarious her cohorts effectively destroyed most of the regions after World War I, with anopheline in- practical usefulness of DDT for mosquito con- cidence there subsequently declining spectacu- trol within a decade. Ironically this would have larly (Hadjinicolaou 1954); and in Russia during happened anyway, probably before the end of the Great Depression. In the former USSR in the 1970s, by rapidly developing mosquito re- 1962, the late professors W. N. Beklemischev sistance to the best and cheapest synthetic chem- and P A. Petrishcheva detailed to me the exten- ical pesticide that vector controllers have ever sive mass-rearing of G. holbrooki in swimming enjoyed. Sadly, the mushrooming public and pools for annual aerial transportation to large consequently political opposition to DDT was a numbers of anopheline larval habitats from the major factor in the premature cessation of Ukraine to Turkestan and beyond in the late WHO's Malaria Eradication Programme a quzr- 1920s and 1930s. They contended that this ma- ter of a century ago. Consequently, much human jor operation had been instrumental in greatly sickness and death from malaria is ascribable to reducing malaria incidence in a period when the Carson and her followers. economic situation precluded the purchase of A similar campaign is now being waged by adequate amounts of chemical larvicides (oils some without personal experience of the field and Paris green). Surely such reduction of hu- control of mosquito pests and vectors, especially man morbidity and mortality was of far greater in emergency situations, against Gambusia spp. importance in a crisis situation than was any as- Witness the declaration of Arthington and Lloyd sociated adverse environmental impact! "Review (1989) that of the world literature on Courtenay and Meffe (1989) don't seem to mosquito control has not supported the view that appreciate that by the middle of this century G. affinis has reduced mosquito problems or the mosquitofish (including cold-hardy and brack- incidence of mosquito-borne diseases in Austra- ish-water strains) were already established in lia or elsewhere. . . apart from moderate control more than 60 countries around the world (Krum- in parts of California . . . ." Worse, Courtenay holz 1948) and had perhaps become the most and Meffe (1989) display unfamiliarity with numerous of all freshwater fish (Hackett 1937). "Mosquitofish their subject in urging that are far How unthinkably expensive their proposed in- too aggressive and predatory to be indiscrimi- ternational ban on these fish would prove, and nately spread throughout the world without rec- what an exercise in futility! ognition of dangers to native biota. An interna- This is not to decry rational environmental- tional ban on their use as a control agent is bi- ism. Gerberich and Laird (1968) widely circu- ologically appropriate and warranted." Had the lated full notice of adverse viewpoints regarding authors' review even been confined to the very mosquitofish, including those of Myers (1965), widely circulated 719 annotations (including this helping to ensure that greater attention was Myers's) of Gerberich and Laird (1968), their thenceforward given to the impact of mosqui- condemnation could not possibly have been tofish on nontarget organisms-6 fact evident made. from many of the 472 additional titles in Ger- There is abundant evidence in Gerberich and berich and Laird (1985). We must be more alert. r66 JounNar oF THE AMERTcIN Mosquno Coxrnol Assoctauoll Vor-.12, No. 2 though, to the dangers that environmentalist ac- fishes in strange places: a review of introduced poe- tivism and ensuing political consequences pose ciliids, pp. 301-331. In: G. K. Meffe and E E Snel- to rational biocontrol aspects of mosquito sup- son, Jr. (eds.). Ecology and evolution of livebearing (Poeciliidae). pression, including the efforts of South Pacific fishes Prentice Hall, Englewood Cliffs. NJ. countries to keep out unwanted Culicidae. Thus, Gerberich, J. B. and M. Laird. 1968. Bibliography of Senator Patrick J. Leahy (Democrat, VT) ex- papers relating to the control of mosquitoes by the ploded two years ago (International Herald Tri- use of fish. An "very annotated bibliography for the years bune, March 29, 1994) on his unpleasant 1901-1966. FAO Fisheries Tech. Pap. 75, Rome. experience" when exposed to an aerosol disin- Gerberich, J. B. and M. Laird. 1985. Larvivorous fish sectization protocol on approaching Australia. in the biocontrol of mosquitoes, with a selected bib- Howeveq the South Pacific lands wish to main- liography of recent literature, pp. 47--76. In: M. Laird and J. W. (eds.). tain their freedom from current travellers such Miles Integrated mosquito control methodologies, Volume 2. Academic Press, as Aedes albopictus and are not impressed by London and Orlando, FL. the "to Senator's threat introduce legislation to Grassi, B. 1923. Pesci nostrali antimalarici. Atti Ac- protect Americans who travel to those coun- cad. Naz. Lincei Rend. 32:511-513. tries." I strongly counsel members of our As- Hackett, L. W. 1937. Malaria in Europe. An ecolog- sociation having the global environment as ical study. Oxford Univ. Press, London. much at heart as I do, not to keep backing away Hadjinicolaou, J. 1954. Observations on the control from the real issues of control in an increasingly of some insects of medical importance in Greece. Proc. Int. Symp. Control Insect Vectors Dis. 1:160- undisciplined world. Failing our agreement upon 183. what constitutes rational integrated mosquito Krumholz, L. A. 1948. Reproduction in the western control. and then sticking with it, we could col- mosquitofish, Gambusia affinis affinis (Baird & Gi- lectively find ourselves up that familiar creek, rard), and its use in mosquito control. Ecol. Monogr. without a paddle. 18:l-43. Myers, G. S. 1965. Gambusin, the fish destroyer. Trop. Fish Hobbyist 13:31-32, 53-54. REFERENCES CITED Rupp, H. R. 1995. Integrated mosquito management. Wing Beats 6(3):18, 2O, 24. Arthington, A. H. and L. N. Lloyd. 1989. Introduced Sella, M. 1926. (A new means of combating Anoph- poeciliids in Australia and pp. New Tnaland, 333- eles in Italy). C. R. Congr. Int. Paludisme. 16 pp. 348. In: G. K. Meffe and E F Snelson, Jr. (eds.). Ecology and evolution of livebearing fishes (Poeci- Marshall Laird liidae). Prentice Hall, Englewood Cliffs, NJ. Honorary Research Fellow Carson,R. 1962. Silent spring. Houghton Mifflin Co., School of Biological Sciences Boston. MA. University of Auckland Courtenay,W. R., Jr. and G. K. Meffe. 1989. Small Auckland, New Zealand