Studies on mosquitoes ( Diptera : Culicidae ) and anthropic environment . A-Survey of resting adults and synanthropic behaviour in South-Eastern , Brazil *

FORATTINI, O.P. et al. Studies on mosquitoes (Diptera: Culicidae) and anthropic environment. 4Survey of resting adults and synanthropic behaviour in South Eastern, Brazil. Rev. Saúde Pública, 27: 398-411, 1993. Resting adults Culicidae were collected from January 1992 through January 1993 in several habitats of the Ribeira Valley region. The diversity of species found among them suggested that the vegetation remaining within human settlements favored the survival and the population increase of some mosquitoes. Among there are: Ae. scapularis, Ae. serratus, Cx. (Culex), Cx. nigripalpus and Cx. (Melanoconion) such as Cx. ribeirensis. That preservation role may be attributed to the rearing of livestock and the consequent increase in the number of blood sources. These species may be classified as hemisynanthropes and Cx. oedipus apparently evolving to the eusynanthropic status. On the other hand, An. cruzii showed an asynanthropic behaviour, with a low degree of survival in the modified human environment. Epidemiological implications of the data are mentioned.


Introduction
Outdoor sampling of resting adult Culicidae are usually made to obtain females for blood meal analysis and parity rate determination.Searching for natural outdoor resting places is frequently difficult because mosquito populations are normally widely dispersed even when, as happens in some cases, specialized shelters are found.Nevertheless this is considered to be a biased procedure because it may not provide representative samples of the population (Service 24,25 1977,1993).Thus, systematic evaluation of natural adult resting places is still needed for many mosquito species, especially in the Neotropical region.The great majority of studies so far carried out have focused on malaria transmission, and we have concentrated on anophelines.Mosquito groups, in which some degree of synanthropy exists, were also studied.A problem is that available data are largely observational (Breeland 1,2 , 1972; Navarro et al. 16,17 ,1986, 1987; Gomes and Forattini 11 , 1990; Natal et al. 15 , 1991; Rubio Palis and Curtis 23 , 1992).More recent researches, largely focusing on the culicids as a whole, have been carried out elsewhere (Reisen et al. 22 , 1982; Irby and Apperson 13 ,1992).In the Ribeira Valley region some of these studies have already been undertaken, mainly from the point of view of adult feeding preferences, resting places and dispersal patterns (Forattini et al 4,5,6 ,1987,1990).
Generally speaking, the larger part of the available data are related to artificial shelters sampling for the post appetential flight of female mosquitoes.Up to now few attempts have been made to study the resting places of male adults.Generally, species are more easily and reliably recognized, through their male characters (as also is their taxonomic identification) and the collection of males provides the opportunity to improve biodiversity studies.Following the researches programme on mosquitoes and anthropic environment, this paper presents the results obtained from collecting resting adults, paying particular attention to male.

Study area
Resting mosquitoes were collected in three areas consisting of the Experimental Station (ES), Fonte (FT) and Pariquera Mirim (PQM) district.Detailed descriptions of these study sites have been presented elsewhere (Forattini et al. 5,8,9 , 1987, 1993).From the environmental and landscape points of view, they are representative of the following: a) ES is a large rural human settlement with a great degree of anthropic modifications, which includes the residual woods (RW) representing remnants of the original rain forest.b) FT represents the wild regional environment characteristic of the original Atlantic rain forest ecosystem.c) PQM represents a group of small rural human settlements with secondary growth of bushes that succeded the primitive covering of vegetation.These small sized and low height woods are characterized by the predominance of short slender trees and undergrowth (Figure l, A, B and C).

Material and Method
Resting mosquitoes were collected from vegetation that could be classified as the natural shelters for adults, inside the wooded environment and at the ecotone between this and the open land, with graminous plants and shrubs up to 1.0 meter high.Sampling was diurnal and fortnightly, through out one year (January 1992 -January 1993).Each collection lasted one hour when, in general, five to ten 5minute collections were made.All the habitats were sequentially visited during each sampling occasion.Samples were collected using a battery-powered aspirator as described by Nasci 14 (1981) (Figure 1,D).In order to provide comparisons, artificial resting places were sampled too.They were represented by the peridomestic environment at the Fonte (FT) household, which has already been described in a previous paper (Forattini et al 8 .1993).
After each collection the specimens were identified and counted.Pearson's correlation analysis was performed to identify any relationship between monthly dominance and rainfall levels.Biodiversity was calculated as proposed by Fisher et al. 3 (1943).As for domiciliation, the Povolny 19  (1971) classification concepts and the Nuorteva 18  (1963) synanthropic index (b) were collected, as well as the synanthropic ratio (sr).Data on macroclimatic conditions were obtained from monthly records for the 19561985 period (Instituto Oceanógrafico da USP 12 , 1989).
The collections were carried out as follows: Asp.l residual wood A at Experimental Station (ES).Asp.2 residual wood B at Experimental Station (ES).Asp.3 secondary bush at the Pariquera Mirim (PQM).Asp.4 peridomiciliary dwellings at the Fonte (FT).Asp.5 primitive rain forest at the Fonte (FT).
The taxonomic identifications are presented with the generic and subgeneric names abbreviated according to Reinert 20,21 (1975, 1982).

Results
A total of 22,140 adult mosquitoes (12,790 females and 9,350 males) were collected, and these include 11,805 (53.3%) from the ES; 5,563 (25.1%) from PQM and 4,772 (21.6%) from FT.The identifications obtained are presented in Tables 1 and 2. Of that total, 77.2% (60.5% of the females and practically 100.0% of the males) were identified to species level and 39.5% as belonging to various taxonomic groups, or as yet unknown taxa.
Residual woods (ES) -Of the total specimens collected there, 75.3% were identified as species and the pattern of the most prevalent was as follows (Asp.l and Asp.2): The monthly percentage distributions of the total number of each of the three most prevalent species are presented in Figure 2 A,B,C.In February a marked decrease was observed for all of these mosquitoes.The particularly dry weather that occurred during that period could, at least to some extent, explain this.Despite the proximity of the irrigation system at ES, the species An. albitarsis was virtually absent.
Secondary bush (PQM) -After the collections, 79.6% of the specimens caught were identified to species, and the most prevalent were classified as follows (Asp.3): As it can be seen, the first three species were the same as those found in the ES residual woods.In addition, Ae. nubilus was significantly present in this secondary bush habitat.The monthly percentage distributions are presented in the Figure 2 A,B,C,D.
Peridomestic environment (FT)-Of the resting specimens collected in this habitat, 93.0% were identified to species, the most prevalent being the following (Asp.4): As expected, Culex quinquefasciatus was predominant.In addition, Cx. oedipus figures as a significant mosquito found in this environment and therefore deserves of attention.The monthly percentage of regarding these two species is presented in Figure 3A.
Primitive forest (FT) -71.4% of the specimens collected there were identified to species, among them the most predominant were as follows (Asp.5): The monthly percentage distributions of Ae. scapularis and Ae.serratus collections are presented in Figure 3B, and suggest a similar trend between them throughout the year.
Biodiversity -The overall biodiversity indices (a), with 95.0% confidence intervals of the number of resting mosquito species (n) collected from many habitats, were 16.0 ± 0.8 for ES plus PQM, and 14.0 ± 0.7 for FT.The comparison among the several sites (Tables 1 and 2) resulted in indices as follows: A quantitative difference was recorded between the residual woods and the secondary bush samples (Asp.l,Asp.2, Asp.3).However, analysing the data from woods A and B, shows that the number of species was 82, of which 29 were found only at A, 14 only at B and 39 were common to both sites.Thus, no quantitative difference was found when the total numbers of species collected were compared with those found in the secondary bush (Asp.3),where 83 species were recorded.
As for the FT region, of the total number of 79 species identified, 14 were found only in the peridomestic environment, 35 only in the primitive forest and 30 in both habitats.As expected, a quantitative difference was found among the peridomestic (Asp.4) and the primitive forest (Asp.5)environments.Figure 4 shows the main species percentages on the total number of resting mosquitoes collected at several sites.
Synanthropy -To estimate the degrees of synanthropy, a comparison among the species collected in three different environments was made.Two of them resulted from anthropic activity as settlements, and so were represented by the ES residual woods (Asp.l+ Asp.2), jointly with the PQM secondary bush (Asp.3) and the FT household surroundings (Asp.4).The third was represented by the wild environment of the FT primitive forest (Asp.5).The synanthropic indices (s) for the species collected in these three habitats are presented in Table 3.As can be seen some species showed a higher degree of synanthropy than others, as regards their survival capacity and consequent adaptation to the artificial human environment.An explanation of the meaning of these indices is given in Figure 5.
The synanthropic ratios (sr) were estimated by comparing ES + PQM/FT data, and consequently the results of Asp.l + Asp.2 + Asp.3 were divided by those of Asp.5.Taking into consideration those species of which at least 100 specimens were collected, the resulting ratios were as follows: With regards to the household environment represented by peridomestic collections (Asp.4),Cx. quinquefasciatus was found only there.Because it was confined to that habitat, the synanthropy of this mosquito was clearly at the highest degree.Although Cx. oedipus was not found inside the FT forest environment (Asp.5), its synathropy ratio can nevertheless be estimated in the light of lhe FT peridomestic environment (Asp.4),just as the human domicile and the ES residual woods jointly with the PQM secondary bush (Asp.l+ Asp.2 + Asp.3) constitute the anthropic rural settlement.Thus sr = FT/ES + PQM gave a ratio of 7.9, indicating a degree of synanthropy for that mosquito.

Discussion
The results reported here arc concerned with regular sampling of mosquito resting sites.To a certain degree, the results arc a reflection of breeding places near which the adults, subsequent to resting after emergence, dispersed through the environment.Therefore the data obtained may suggest the habitats where mosquitoes carry out their activities.
Comparing these results with those previously reported (Forattini et al. 10 , 1993) it is remarkable that An. albitarsis was practically absent at the resting sites sampled in the two residual woods of the same Experimental Station (ES).On the other hand, Ae. scapularis was consistently collected at those sites, as well as at the irrigation system.
Significant monthly correlations with rainfall were found for Ae.nubilus (r = 0.54; P<0.05) and Ae.serratus (r = 0.41; P<0.05), both collected at PQM secondary bush, where they showed a peak associated with the dry season (August) (Figure 2 A and D).Otherwise, the end of the rainy season (March to May) was significantly correlated to Cx. declarator (r = 0.37; P<0.05) collections in the same PQM environment (Figure 2D).Nevertheless, these relations varied according to the habitats and no correlations with rainfall were found in the monthly distribution of these species in the ES residual woods (Asp.l+ Asp.2) and in the FT primitive forest (Asp.5).In addition, Ae. scapularis and Cx.nigripalpus showed no significant correlations with their monthly collections through the year.In the FT peridomestic environment, significance was found for Cx.quinquefasciatus (r = 0.32; P<0.05) and for Cx.oedipus (r = 0.46; P<0.05), correlated to the wet season (Figure 3A).These results agree with those previously reported (Forattini et al. 1993 9 , 1993 10 ) which show Ae.scapularis as a mosquito occurring throughout the year, both in the anthropic and the primitive environment (Figures 2B, 3B).As for Culex species, adults were collected in larger numbers when the rainy season was finishing or starting.This is probably associated with the greater abundance of larval habitats, which tend to be flushed out during heavy summer rains.
The ES plus PQM data indicated higher biodiversity than that observed in the FT primitive environment.The explanation may be found in the influence of vegetation that, even in the human settlements, harbours mosquito species favouring blood sources from cattle and other domesticated animals (Figure 1B).Obviously mosquito species diversity was lower in the household environment, where Cx. qiunquefasciatus is the dominant species.
It seems evident that a mosquito species group which includes Ae. nubilus, Ae. scapularis, Ae. serratus, Cx. nigripalpus, Cx. ribeirensis and others (Figure 5) is favored by the installation of human settlements.Natural environmental alterations result in a patchy remnant vegetation that provides resting places for many mosquito species.Inadiction, domestic livestock furnish blood sources.Thus, such mosquito species may find resources good enough to enable them to survive and their population to increase.According to the Povolny 19 (1971) classification, these species may be considered as hemisynanthropes with respect to their synanthropic indices (s) and ratios (sr) (Table 3; Figure 5).The present results support previous observations on Ae. scapularis and Culex species, such as Cx.ribeirensis and others (Forattini et al. 4,6,7 , 1987, 1990, 1991; Gomes et al. 11 1990).At the opposite extreme An. cruzii may be classified as an asynanthrope mosquito which decreases significantly in accordance with the development of the anthropic alterations to the natural environment.Even when visiting households regularly seeking blood adult females do not remain there, but instead fly after feeding to their natural habitats (Forattini et al. 6,8 1990,1993).
As for the peridomestic environment, Cx. quinquefasciatus must obviously be considered as an endophilous eusynathropic mosquito.The results obtained here strongly suggest that Cx. oedipus may be a local species developing exophilous eusynanthropic behaviour.That is, it may be associated with the manmade environment but without necessarily requiring human habitation.In any case, it is a species deserving further attention.