Joumal of the American Mosquito Control Association, l3(l):18-23, 1997 Copyright A 1997 by the American Mosquito Control Association, Inc. EVALUATION OF CARIBBEAN STRAINS OF MACROCYCLOPS AND MESOCYCLOPS(CYCLOPOIDA: CYCLOPIDAE) AS BIOLOGICAL CONTROL TOOLS FOR THE DENGUE VECTOR AEDES AEGYPTI S. C. RAWLINS, R. MARTINEZ, S. WILTSHIRE, D. CLARKE, P PRABHAKAR AND M. SPINKS Caribbean Epidemiology Centre (CAREC), P. O. Box 164, Port-of-Spain, Trinidad ABSTRACT. Fifteen Caribbean strains of copepods were assessed for their predation ability against mosquito larvae. Macrocyclops albidus from Nariva, Mesocyclops aspericornis from Oropouche, and Mesocyclops lon- gisetus from El Socorro, Trinidad, were most effective against Aedes aegypti but not against Culex quinquefas- ciatus. Mesocyclops longisetus and Me. aspericornis prevented any mosquito survival over 25 wk of observation despite weekly challenges with Ae. aegypti. The copepods were tolerant to dosages of the insecticide temephos that are usually toxic to mosquito larvae. This indicated that copepods could be incorporated into an integrated control system. To determine whether pathogenic microbes might be introduced with copepods into drinking water, microbial studies were done on the copepods. These showed the presence of only Aeromonas .sobria, Pseudomonas sp., Alcalignes sp., and gram-positive bacilli. Although none of these are highly pathogenic to humans, the application of these copepods has not yet been recommended for use in drinking water. INTRODUCTION 1992). ln contrast, the larger M. longisetu,s killed IOOVo of Ae. aegypti, Anopheles farauti Laveran, The continuing endemicity of dengue serotypes ard Culex quinquefascialas Say. 1, 2, and 4 in the Caribbean region together with Vasconcelos et al. (L992) in small-scale field tri- the risk of introduction of dengue type 3 from Cen- als in Brazil found that M. longisetus controlled Ae. tral Arnerica has greatly emphasized the need for aegypti in containers that produced more than 8O7o efficient managementof the vector Aedes aegypti of the mosquito. They found the involvement of the (Linnaeus)(Nathan 1993, Pan American Health Or- community in such a project to be very helpful. In ganization 1994). There is evidence of failure of Mexico, M. longisetus selected from 15 species of source reduction practices for controlling this con- copepods collected in Nuevo Leon and Coahila was tainer breederdue to the reluctanceof some Carib- equally predaceous on both Ae. aegypti and Cx. bean householdersto discard most potential mos- quinquefasciaras (Rodriguez 1992). In the Carib- quito-producing containers (Rosenbaum et al. bean islands of Puerto Rico and Anguilla, M. as- 1995). Also, the development of insecticide resis- pericornis controlled but did not eliminate Ae. ae- tance in several Caribbean populations of Ae. ae- gypti in drum habitats (Suarez 1992). gypti (Rawlins and Hing Wan 1995) has forced The prospect of incorporating copepods into an mosquito control authoritiesto seek alternative sus- integrated control system for mosquitoes has at- tainable systemsfor managementof this mosquito. tracted the attention of several mosquito control Biological control tools such as the mosquito workers. Thus, the combination of copepods with Toxorhynchitesmoctezuma (Dyar and Knab) have Bacillus thuringiensis var israelensis (B.t.i.), Ba- been evaluated for management of Ae. aegypti cillus sphaericrrs, and methoprene (Tietze et al. (Rawlinset al. 1991,Tikasingh 1992). This hasmet 1994), or with insecticides (Marten et al. 1993), has with only minimal success due to failure of the improved the performance of various copepod spe- predator to flnd and colonize habitats of its prey in cies. urban environmentswhere Ae. aegypti is known to This paper evaluates the larval consumption thrive. characteristics of a number of Caribbean copepod Recently, certain cyclopoid copepodshave prov- taxa and their capability of surviving in insecticide- en to be capable of controlling Ae. aegypti larvae treated habitats and their potential for incorporation in peridomestic breeding containers (Marten 1990, into an integrated control program for Ae. aegypti. Marten et al. 1994). Marten et ^1. (1992) reported The present study provides data from such work that one species, Mesocyclops longisetus (Thi€' performed in Port-of-Spain, Trinidad. baud), was capable of reducing Ae. aegypti larval populations by more thar99.9Vo. MATERIALS AND METHODS Other authors have evaluated some Nonh Amer- ican copepods against container-breedingmosqui- Collection of copepods: Various copepod sam- toes such as Aedes albopictus (Skuse)(Schreiber et ples were collected from freshwater bodies in dif- d. 1993). Brazilian strains of Mesocyclopsasperi- ferent Caribbean locations. The copepods were cornis (Daday) showed Potential as biological con- taken to the entomological laboratories at the Ca- trol agents against Ae. aegypti but were not as ef- ribbean Epidemiology Centre (CAREC; Port-of- fective against Anopheles and Culex (Kay et. al. Spain, Trinidad) and were reared on a diet of Par- l8 MrncH 1997 Copppoos Ls ,q.BtocoNrnoI- TooL t9 amecium caudatum, Chilomonas sp. (Suarez et al. of these 100 lst-instar Ae. aegypti larvae were add- 1992), artd lst-instar mosquito larvae. For each ed and the drums were covered with a fine mesh to sample a culture was started from a single gravid prevent entrance of any gravid mosquitoes. The female. The copepods were maintained under lab- drums were left under partially shadedconditions. oratory conditions at ambient temperatures of 2t- At 6-day intervals, all larvae and/or pupae were 34'C until they were evaluated for their predatory removed from each drum by very c:reful sweeping ability against lst-instar mosquito larvae. Copepod with a fine nylon dip net and emptied into enamel samples from each strain were sent to Janet Reid trays. Larvae and/or pupae present were counted, at the Smithsonian Institution, Washington, DC, removed, recorded, and the water containing the who kindly verified the species identifications. A copepods and/or temephos was returned to its re- strain of M. longisetus curvatus (Dussart) was also spective drum. A new set of 100 lst-instar Ae. ae- received from M. E. Suarez. CDC. San Juan Lab- gypti larvae was added to each drum and surviving oratories, Puerto Rico, and was used as a reference larvae, if any, were harvested after 6 days. This predatory strain. processwas continued for 25 wk. Predation studies: First-instar Ae. aegypti larvae Field studies in tire habitats: Discarded auto- were introduced into each well of a tissue culture mobile tires were selectedfor this field study in St. plate each containing one adult copepod in l0 ml James,Port-of-Spain, Trinidad, with a known pres- of water. Six replicates were done for each strain/ ence of both Ae. aegypti and Cx. quinquefasciatus- species. After 24 h, the percent kill was assessed Tires were each treated with one of the following by removing and counting the number of live and regimes: 1) tire 1, IOOM. longisetus in 2 liters of dead larvae. and a new set of lst-instar larvae was water; 2) t$e 2, 20O M. longisetus in 2 liters of exposed to the copepods. This process of evaluation water; 3) tire 3, 10OM. aspericornis in 2 liters of was repeated 5 times and the mean percent mortal- water; 4) ttre 4, 2OOM. aspericorrtis in 2liters of ity of Ae. aegypti larvae for each copepod strain/ water, and 5) tire 5, untreated control, 2 liters of species was assessed. water. Insecticide susceptibility of copepods: Prelimi- Nine different sites were used for each treatment. nary laboratory testing was conducted on 4 strains The tires were left in shaded locations without pro- of copepods: Macrocyclops albidus principalis tection from gravid wild mosquitoesfor 14 wk. The (Herbst) (Nariva), M. longisetus (CDC), M. Iongi- water in the tires was examined by emptying the serus (El Socorro), and, M. aspericornis (Daday) contents into a white enamel bowl and any mos- (Oropouche). From each strain, l0 copepods were quito larvae present were removed with a pipette placed into their respective wells of a tissue culture and taken to the laboratory. This was done at4-day plate with 10 ml of solution containing varying intervals becauselonger periods could permit pu- concentrations of temephos. The 6 concentrations pation and a possible risk of releaseof adult mos- chosen ranged from 0.02 mg/liter (the diagnostic quitoes. The uncounted copepods were returned to dosage for Ae. aegypti larvae) to O.22 mg/liter. their respective tires and the water brought back to Mortality was determined by the presence of mor- its original level. In the laboratory, the larvae for ibund or dead copepods at 24-h intervals over a each mosquito specieswere counted and recorded 2-wk period. No food was offered to the copepods for each tire habitat. in both the controls and test wells during this pe- Statistical treatment: The data were entered us- riod. ing the EPI-Info version 6 program provided by the Copepod/temephos integrated control trials: CDC of the U.S. Public Health Service (Division Two of the more predaceous species of copepods, of Surveillance and Epidemiologic Studies, Epide- Mesocyclops sp. B (near aspericornis) from Chag- miology Program Office, CDC, Atlanta, GA). A re- uaramas and M. aspericornis from North Oropou- peated multivariate analysis of variance was used che, Trinidad, were selected for this tial. Mesocy- to compare regimes, with weeks being selectedas clops longisetus curvatus obtained from CDC in the repeated factor. Verification of the assumptions Puerto Rico was used for comparison to the 2 Trin- for these statistical tests was made through residual idad strains of copepods. analysis. In caseswhere the assumptionswere not TWenty-four drums each containing 200 liters of satisfied, a square root transformation was utilized tap water were allowed to stand covered for 7 days and the test repeated.The usual F statistic was the before any additions were made to tlem. For each criterion applied for assessmentof group compari- strain of copepod, drums were prepared as follows: sons.