Revista de Saúde Pública Survey of cyclopids ( Crustacea, Copepoda) in Brazil and preliminary screening of their potential as dengue vector predators in and preliminary screening of as

are to good mosquito specially as regards the larvae of the dengue vectors Aedes aegypti and Ae. albopictus. Method The objective of the study was to survey the local copepod fauna and search for new strains of M. longisetus var . longisetus , comparing the potential of the samples found with the current strain ML-01 against Ae. albopictus larvae, under laboratory conditions. Eleven bodies of water in Campinas, SP, Brazil, were screened for copepods by collecting 1.5 l of water from each of then. The predatory potential of adults copepods was evaluated over 24 h, in the laboratory, for groups of 5 individuals preying upon 30 first instar Ae. albopictus larvae. same species evaluated for comparison. The sample with few M. albidus var. albidus was ranked in second place showing an average 25.9% efficiency. The use of copepods in trap tires as dengue vector controllers is discussed.


INTRODUCTION
Two species of the genus Aedes have been continuously focused on in vector control campaigns in Brazil in relation to dengue and yellow fever epidemics. Aedes aegypti, considered in 1956 to have been eradicated, re-emerged and has been the main dengue vector in the country since the Boa Vista, RO epidemics in 1981/82 13 . The second species, Aedes albopictus, was first recorded in Rio de Janeiro, RJ in 1986 3 . Supposedly introduced from Japan, this latter species has been considered a secondary dengue vector and a possible link between the presently eradicated urban yellow fever and the sylvatic form of the disease, due to its breeding habits in both urban and rural environments 4,14 . In 1994 the distribution pattern of both species in the State of S. Paulo encompassed 184 and 157 municipalities, respectively, for Ae. aegypti and Ae. albopictus, and a total of 209 distinct counties for the two species 22 .
The families Oithonidae and Cyclopinidae comprise predominantly salt water copepod species of the Order Cyclopoida, whereas the third family (Cyclopidae) is almost universally distributed along with mosquitoes in aquatic habitats such as wells, tree holes, lakes, reservoirs, fountains, salt marshes and even discarded tires 12 . Many forms of freshwater zooplankton have a broad diet but cyclopids show a particular interest in first instar mosquito larvae as prey.
Cyclopids are more effective for biological control than other predatory invertebrates because it is common for cyclopids to be numerically abundant even when mosquitoes are not present 10 . The potential of cyclopid crustaceans for the control of mosquitoes has been undergoing evaluation since the early 1980's 6,18,21 . Field trials and even cyclopidbased programs for mosquito control have been carried out in many countries including Australia 2 , the United States 11 , Honduras 10 , Mexico 15 , Colombia 7 and Brazil 23 .
Four different Brazilian strains of Mesocyclops aspericornis from Fortaleza, CE in the Brazilian northeast have shown good potential as biological control agents of Ae. aegypti larvae, but were not as effective as an autochthonous Mesocyclops longisetus var. longisetus strain 5 . In a recent study, a strain of M. longisetus var. longisetus from Campinas, SP (ML-01) showed efficiencies higher than 97% for controlling Ae. albopictus larvae in trap tires in two field trials 19 . Projects on the use of copepods along with planarians and Bacillus thuringiensis var. israelensis are being carried out by the authors in education-based programs using trap tires. The introduction of copepods in natural mosquito breeding places in the county emerges too as a promising follow-up to this study. The aim of the present paper was to survey the local copepod fauna and search for new strains of M. longisetus var. longisetus, comparing the potential of the samples found with the current strain ML-01 against Ae. albopictus larvae, under laboratory conditions.

Survey
Copepods were collected in Campinas, SP during June and July 1994. Eleven approximately equidistant areas were selected covering both urban and suburban locations of the county. In each area the water body possibly serving as a breeding place was screened by visual inspection,

Material e Método
Onze and samples of 1.5 l of water were collected from a depth of from 5 to 15 cm. The number of adult and immature copepods was then assessed in batches under laboratory conditions, the crustaceans remaining in the water in the field, without additional food for 15 to 20 days before the predation trials. The areas sampled were: Area 1-The Municipal park "Parque do Taquaral" (TAQUA) located in an urban area. Collections were made alongside a 17 ha lake, in well insolated places. Area 2-The Municipal park "Parque Ecológico Monsenhor Emílio José Salim" (PEMJS), on the eastern outskirts of the city. Collections were made alongside a 10 ha lake, in well insolated places. Area 3-Forestry resource reserve "Mata de Santa Genebra" (MSGEN) consisting of 240 ha of semi-deciduous and mesophytic forest on the northwestern limits of the city. Those water samples which were positive for copepods were analyzed for temperature, pH, dissolved O 2 , nitrite, nitrate and ammonia.

Predatory Potential
Adult copepods from each positive sample were evaluated for predatory potential in groups of 5 individuals preying upon 30 first instar Ae. albopictus larvae. M. longisetus var. longisetus from a laboratory culture (ML-01) maintained for the last 3 years were also evaluated for comparison. The size of females with egg sacs was used as indicative of adult forms. Some individuals (from each sample) were saved in order to serve as eventual founders of new cultures. Four replicate cultures along with two controls (only copepods and only mosquitoes) were established in the laboratory for each population sampled.
Mosquito larvae were offered to the copepods 24 h after their transfer to plastic vials containing 450 ml of commercial spring water, with no additional food. Evaluations were made by counting mosquito larvae surviving 24 h thereafter 10 . The temperature was 24 -25 o C. Data were corrected by using Abbott's formula 1 and compared by Student's t-test.
After the evaluations all copepods were preserved in glycerinated alcohol (20%) and sent to the Museum of Zoology of the State University of S. Paulo, USP (entry numbers from 12.270 to 12.280).

Survey
The water from two urban areas (CCONV and TCMEN) showed no copepods during in situ inspection, confirmed by laboratory examination. Additionally, in three areas the number of copepods obtained led to inconclusive identification, being represented by few individuals or only copepodids. The remaining areas showed 1 to 3 copepod species in each sample ( Table 1).
The copepod identification was carried out after the predatory potential test, meaning that in 2 of the samples different species were evaluated together. In PEMJS the proportion was 4 Tropocyclops prasinus to 1 ( M. longisetus var. longisetus) and in the sample RPEDR, Eucyclops sp, Eucyclops solitarius and Macrocyclops a. albidus rated as 3.5/ 1/0.5, respectively.

Predatory Potential
Attempts to breed unidentified copepods in order to evaluate their predatory potential were unsuccessful. Due to the difficulty in unequivocally differentiating adults, late instar copepodids were used and represented 30% and 15% of two of the evaluated samples, respectively TAQUA and PEMJS.
The mortality percentage in the copepod control group was nil while for the 1 st instar Ae. albopictus larvae control group, mortality ranged from nil to 13.3%.
The M. longisetus var. longisetus strain ML-01 confirmed its high potential, preying on up to 92% Ae. albopictus larvae in a 24h period. The samples collected showed a variation ranging from nil to 97.3% (24h), this later performance (UNICA/ paras) being the only one with no significant difference from that showed by the ML-01 strain ( Table 2).

DISCUSSION
The present survey confirms that copepods are common and well distributed in natural water bodies both in urban and suburban areas of Campinas. They were not found in only two downtown artificial pools, probably due to the frequent cleaning that involves complete drainage every two months. In LPARA, also an artificial downtown pool, drainage and cleaning was done 5 months before collections, and at least E. serrulatus could be found in abundance.
Physical analysis of the water where copepods were found showed a pH ranging from 6.5 to 8.5 and dissolved O 2 from 2.0 to 17.0 ppm. Nitrite and nitrate varied from 0.01 to 0.74 ppm and ammonia from 0.06 to 1.28 ppm. Correlation to population density was only found to be "appreciable" for both nitrate (r = 0.46) and nitrite (r = 0.33) and "low" for ammonia (r = 0.25), according to Rugg's table 20 . Water tem-perature ranging from 15 to 19 o C during this study could not account for the absence of copepods in CCONV and TCMEN areas. In UNICA/paras copepods have been frequently found during the winter at temperatures around 15 o C.
According to Reid 16 late instar copepodids are easily mistaken for adults. Rietzler 17 has pointed out that copepods begin to prey on such insects as first instar copepodids, the predation rate rising as they mature. Thus, it may not be supposed that the low observed efficiencies recorded in the present study are to be directly explained by the presence of some late instar copepodids in samples TAQUA and PEMJS.
Statistical analysis of the predation trials showed the best results for the sample containing only M. longisetus var. longisetus (UNICA/paras), with no differences in the trials with the same species cultured under laboratory conditions (ML-01). A lower efficiency was found for two samples: one with both M. longisetus var. longisetus plus T. prasinus (PEMJS) and the other with Eucyclops sp plus M. albidus var. albidus plus E. solitarius (DPEDR). The present results confirm the good potential of M. longisetus var. longisetus and suggest that M. albidus var. albidus was the main predator in this last sample (DPEDR), as both species have been shown to be the most effective against Ae. albopictus larvae 8 .
It may be concluded that the copepods found of the genera Eucyclops, Tropocyclops and Metacyclops have poor predatory potential against Ae. albopictus larvae. Samples containing only copepods of the genus Eucyclops showed predation varying from nil to 3.42%. The sample with only M. mendocinus showed a percentage not significantly different from that range, and the sample with T. prasinus (PEMJS) showed a significantly higher efficiency, but probably due to the presence of one individual of M. longisetus var. longisetus. Copepod efficiency as a mosquito predator can be discussed in terms of the adult size of the different species. In the data presented by Reid 16 a mean length of 0.87 mm is given for adults of the presently evaluated species in Eucyclops, Metacyclops and Tropocyclops, irrespective of sex. Indeed, copepod species represented by small individuals are quite common in mosquito breeding habitats but they are not such good predators as those represented by larger individuals such as those of the genera Macrocyclops and Mesocyclops (mean length 1.45 mm) 9 . While one individual of the largest copepod species (Macrocyclops or Mesocyclops) can prey on as many as 50 larvae/day 10 , the smallest ones attack mosquito larvae in groups, though not compensating in numbers for the disadvantage of their size (C. E. Rocha, State University of S. Paulo, personal communication).
From these findings M. longisetus var. longisetus remains the most promising species among the copepods for use as a biological agent to be introduced in natural breeding places in Campinas. At the present time, no species so far screened shows such good potential as M. longisetus var. longisetus for use in trap tires.

CONCLUSION
Copepods, represented in the present survey by eight species of five genera, are easily found in bodies of water from central urban and suburban locations in Campinas, SP, Brazil.
M. longisetus var. longisetus, a copepod already used in biological control projects, was found in only two out of eleven breeding places, occurring along with Tropocyclops prasinus in a 10 ha lake, and alone in a 1 m 3 rain water reservoir.
Predation trials against 1 st instar Ae. albopictus larvae with samples containing only M. longisetus var. longisetus confirmed this latter as a good mosquito controller. The predatory efficiency of a field collected sample of this copepod did not differ from that of a 3-year old laboratory culture.

ACKNOWLEDGMENT
To Dr. C.E. da Rocha (University of S. Paulo) for identifications of copepods.