Studies on mosquitoes ( Diptera : Culicidae ) and anthropic environment . 10-Survey of adult behaviour of Culex nigripalpus and other species of Culex ( Culex ) in South-Eastern Brazil *

A survey of adult behaviour of Culex (Culex) species was carried out from August 1992 through December 1993 in a human modified (anthropic) environment in the Ribeira Valley, S.Paulo State, Brazil. Culex nigripalpus dominated the catches at several sites and it's tendency to increase in the anthropic environment became quite clear. Nevertheless no high level of synanthropy was demonstrated. So it seems that the mosquito may have a restricted role in natural arbovirus cycles. Nonetheless, Cx. nigripalpus must be considered a potential vector of arboviruses, especially St. Louis encephalitis virus outside dwellings.


Introduction
From the epidemiological point of view Culex (Culex) is a group of mosquitoes that includes several arbovirus vector species.They are essentially nocturnal-crepuscular tending to be active throughout the night, but most are active just after dark.In general, these mosquitoes are mainly ornithophilic and thus they cycle virus agents among avian hosts.Nevertheless, important differences exist according to weather and peculiar environmental conditions that may influence their role as an arbovirus vector.In North America, species of the Culex subgenus are important in the transmission of mosquito-borne arboviruses of public health importance such as eastern (EEE), western (WEE) and St. Louis (SLE) encephalitis.Culex nigripalpus is considered to be an important SLE vector in the southeastern United States, and perhaps EEE virus as well (Chamberlain et al. 2 , 1964; Dow et al. 4 , 1964; Morris 27 , 1992).
In Central and South America, Cx. nigripalpus was considered as a potential vector of EEE virus in the Dominican Republic (Mitchell et al. 25 ,1979).In addition, an isolation of vesicular stomatitis New Jersey (VSNJ) virus was obtained from this species in Guatemala (C.H.Calisher apud Webb and Holbrook 29 , 1988), and isolations of others virus were reported from this mosquito in Panama (Dutary etal. 5,1984).Isolations of SLV from Cx. nigripalpus were obtained in Jamaica and Guatemala (Belle et al. 1 ,1964; Cupp etal. 3 ,1986).In Argentina, natural infections by SLE virus were reported in Cx. quinquefasciatus and undetermined Cx. (Culex) specimens (Mitchell et al. 26 ,1985).In the brazilian Amazon region, SLE virus was isolated from Cx. coronator, Cx. declarator and unidentified Cx. (Culex) mosquitoes which also provided Mucambo virus isolations (Vasconcelos et al. 28 , 1991).The relative importance of all these neotropical culicids in the maintenance and transmission of these arboviruses, particularly SLE virus in Latin America, remains to be defined.
Observations on mosquito behaviour were made during a research program in the Ribeira Valley region of S. Paulo State, subsequent to an encephalitis epidemic started in 1975-1976 thereafter decreasing until 1983 (Iversson 21 , 1988).Culicidae were studied regarding their dominance, biting habits, synanthropy and geographic distribution (Forattini et al. 6,7,8,10,11 , 1981, 1986, 1989,  1990).More recently, observations were performed at a rural man-made environmental sites where artificial irrigation is used for rice cultivation, focusing on several species including some Culex (Culex) found there (Forattini et al. 12,13,14,15 1993,  1994).A continuation of these researches in that anthropic environment focused on mosquito biting activities and occurrence at several places.During 1992-1993, adults were collected through the use of human bait and Shannon-type traps.The results regarding the subgenus Culex species are presented here for the purpose of revealing epidemiological evidence about their potential hypothetical vector role.

Study areas and methods
Adult sampling followed the same study areas and methods have been described elsewhere (Forattini et al. 16,17 , 1995).To make it easier to understand the results presented, the sampling sites are noted again.Biting activity was estimated through the Williams' mean (Xw), the domiciliation degree estimation was made by the synanthropic index(s) comparing it through the synanthropic ratios(sr).Besides, considering the several sites as representative of development phases, the index of change (IC) was calculated for the abundance comparison at each of these phases.
Data about the climatic conditions during the period were obtained directly from the climatology Section of the Campinas Agronomic Institute (Instituto Agronômico de Campinas) of S. Paulo State.

Results
A total of 1,880 adult mosquitoes (

Shannon Traps
The monthly distribution of Cx. nigripalpus and of the Coronator Group species, are presented in Tables 2 and 3. Cx. nigripalpus incidence in the open land (Shan E) of the anthropic environment (ES), that correspond to the rice fields margin, showed a February-April peak.In that agricultural environment of the ES area, 90.5% of the Cx.nigripalpus females was caught in the February-April period.Nevertheless, it seems that adult occurrence was explosive at that time, coinciding with higher levels of precipitation and less temperature fluctuations that occurs during that time period (Fig. ).The monthly distribution of the Coronator Group was more regular, apparently not influenced by environmental factors in any particular period of the year.
The data obtained in the forest patches (Shan 1 and Shan 2) were statistically compared with those from the open land (Shan E) (Table 2).Through the Mann-Whitney test the p values were 0.8918 for Cx.nigripalpus and 0.0464 for the Coronator Group, resulting in some significance for this second group of mosquitoes.
In the partially disturbed environment, represented by the GA site, irregular variations were obtained.As a general view, what may be considered a small peak occurred in March for both mosquitoes groups (Table 3).

Domiciliary Environment
Through the use of human bait at the BC dwelling, a total of 66 Cx. (Culex) females was caught.From that total 58 (87.9%) belonged to Cx. nigripalpus and only 8 (12.1%) to the Coronator Group.According to the sampling sites, the results obtained were as follows (Table 1 and 4

):
Comparing the sites sampled, through the Mann-Whitney test, the p value was 0.524 for Cx.nigripalpus showing no significant differences between indoor and outdoor collections.Regarding the Coronator Group, the samples obtained were too small and thus insufficient for the statistical test (Table 4).

Synanthropy
Comparison of the three sampled environments was made for both mosquito groups using synanthropic indices (s).The results obtained follow: Thus, these data suggest that Cx. nigripalpus retained its exophilic preferences, while the Coronator Group species showed a clear sylvatic behaviour.
Considering the synanthropic ratios (sr) at the ES area, comparisons were made, between the total females caught at the open area (Shan E) jointly with that obtained in the domiciliary environment (BC), and the total caught at the remnant forests (Shan 1 and Shan 2).Doing this, the value obtained was 1.53 conjoin for both groups.The specific values were 1.75 for Cx.nigripalpus and 0.40 for the Coronator Group.These data suggest that the residual forest patches retain greater numbers of Coronator Group mosquitoes than Cx.nigripalpus.Besides, relating the open land (Shan E) data with those of the domiciliary environment (BC), the overall ratio was 13.03 for both mosquito groups considered together, showing greater occurrence outside than inside domiciliary environments.Nevertheless, specifically calculated for each of these mosquito groups were 15.31 for Cx.nigripalpus and 4.1 for the Coronator Group, showing greater occurrence outside than inside the dwelling environment.However, as shown in Table 1, the results obtained with the dwelling catches of Coronator Group were quite negligible.Considering the two groups, separately, the calculated ratios were: In the domiciliary environment (BC) both Cx.nigripalpus and the Coronator Group species showed a lower synanthropy level.A clear tendency to reach dwellings, in a significant manner, was not detected.
Behavioural succession, estimated through the index of change (IC), compares phase 1 (Shan 1 + Shan 2) with the phase 2 (Shan E + BC) (Table 5).These indices were calculated as 0.9853 for Cx.nigripalpus and 0.3433 for the Coronator Group, indicating changes for both these mosquitoes groups.Estimating the changes from phase 1 to phase 2, that mosquito increased 126.31 times while the others increased only 2.04 times.

Discussion
Among the mosquitoes sampled in the collections reported above, Cx. nigripalpus predominated among the Cx.(Culex) species.The Coronator Group species were second in frequency.In previous studies in the same Ribeira Valley region Cx.nigripalpus was constantly poorly represented in collections taken inside the primitive forest environment but readily found in the modified areas represented by remnant and secondary forests (Forattini et al. 7,9,13 ,  1986, 1989, 1993).Probably this pattern may be related to sampling performed at the forest ground, as this mosquito seems to have a good degree of acrodendrophily (Guimarães et al. 19 , 1985).Nonetheless, at least in southern Brazil, Cx. nigripalpus may frequent anthropic modified environments, in some localities (Lourenço-de-Oliveira 22 , 1984).Besides, in the results reported here, no difference was observed between the remnant forest and the open cultivated land of the anthropic environment, regarding collections of that mosquito.The Coronator Group showed a slight preference for the remnant forest environment.An insufficient number of females caught limited the meaning of the statistical analysis.Regarding the data obtained at the partially disturbed environment of the GA site, no difference was found between those two mosquito groups.
The monthly distribution of Cx. nigripalpus in the ES cultivated environment shown a major peak in February-April.This period coincided with the beginning of the fallow uncultivated phase of the rice paddies cycle.Thus, it followed the crop harvesting phase where the empty rice fields and in-creased rainfall favoured larval breeding.However, that association seems quite variable and remains unclear (Forattini et al. 12,15 , 1993, 1994).Anyway, the results seem to correlate with an increase in rainfall and higher temperature levels (Fig. ), as was observed in others regions of southern Brazil (Guimarães and Arlé 18 , 1984; Lourenço-de-Oliveira et al. 23 , 1985).
Considering the domiciliary environment, the data obtained were insufficient to estimate the Williams' means.For Cx. nigripalpus, no significant difference was evident between indoor and outdoor catches.About its synanthropic level, the estimation based on the s indices, gave lower values, around the independence degree (0.0).No clear tendency to enter dwellings was detected, even through the sr ratios.
Finally, the changing pattern of the succession, estimated by the indices of change (IC), indicate that changes occurred for Cx.nigripalpus.These results agree with what was said at the beginning of this discussion.Certainly that mosquito, at least in this southern region of Brazil, is favoured by the man-made environment, even though not reaching, in a significant manner, human habitations.
These data obtained in the Ribeira Valley region support the hypothesis that Cx. nigripalpus may fulfill a role in the local arbovirus cycle.Its biting habits were reported elsewhere and, if not definitive, they suggested that this and other Cx.(Culex) mosquitoes, some having even a great degree of ornithophily, may develop a suitable level of anthropophily (Lourenço-de-Oliveira and Heyden 24 , 1986; Guimarães et al. 20 , 1987; Forattini et al. 9 ,  1989).So, in southern Brazil, Cx. nigripalpus at least deserves attention as a possible vector of arboviruses.Among these viruses, the St. Louis encephalitis maybe transmited ouside the domiciliary environment.

Table 1 .
It may be seen that Cx. nigripalpus comprised 77.9% and jointly with Coronator Group (Cx.coronator and Cx.usquatus) reached 88.1% of the total Cx.(Culex) specimens collected.The behaviour of Cx. nigripalpus, compared to the Coronator Group, will be deserve special attention in this paper.As females of both represented 86.0% of the total adults caught, the following data will relate to them.Specimens determined as Cx.(Culex) sp, could not be identified below subgenus because of damages incurred during collection.