Laboratório de Identificação e Pesquisa em Fauna Sinantrópica, Centro de Controle de Zoonoses, Coordenação de Vigilância Saúde, Secretaria Municipal de Saúde Prefeitura Municipal de São Paulo, Rua Santa Eulália, 86, São Paulo, 02031-020, Brazil
Laboratório de Protozoologia, Instituto de Medicina Tropical de São Paulo, Universidade de São Paulo, Av. Dr. Enéas de Carvalho Aguiar, 470, São Paulo, 05403-000, Brazil
In order to assess the epidemiological potential of the Culicidae species in remaining areas of the Brazilian Atlantic Forest, specimens of this family were collected in wild and anthropic environments. A total of 9,403 adult mosquitoes was collected from May, 2009 to June, 2010. The most prevalent among species collected in the wild environment were Anopheles (Kerteszia) cruzii, the Melanoconion section of Culex (Melanoconion), and Aedes serratus, while the most common in the anthropic site were Coquillettidia chrysonotum/albifera, Culex (Culex) Coronator group, and An. (Ker.) cruzii. Mosquito richness was similar between environments, although the abundance of individuals from different species varied. When comparing diversity patterns between environments, anthropic sites exhibited higher richness and evenness, suggesting that environmental stress increased the number of favorable niches for culicids, promoting diversity. Increased abundance of opportunistic species in the anthropic environment enhances contact with culicids that transmit vector-borne diseases.
Remnants of the Atlantic Forest in southeastern Brazil constitute a biome of high diversity and multiple niches that favor mosquito (Diptera: Culicidae) proliferation (Forattini et al. 1990, Guimarães et al. 2000). A portion of São Paulo state contains a mountainous coastal area dominated by an Atlantic Forest fragment, which plays an essential role in the dynamic balance of temperature and humidity in the metropolis, affecting the local quality of life. However, humans are increasingly invading these areas, causing environmental disturbance and degradation.
The Capivari-Monos Environmental Protection Area (APA) was established in the south of São Paulo municipality to ensure the preservation of an Atlantic Forest area. A significant part of the APA is still covered by primary vegetation, corresponding to 1/6 of the municipal area. However, as a result of vegetation suppression and selective plant cutting throughout land use history, secondary forest is currently the main formation in the APA, producing an assemblage of plant species at different ecological succession phases. Classes of native vegetation present in the APA are dense, upper-mountain and lower-mountain rainforest (Ombrophilous Forest). Riparian forests, swamps, bogs, and flood plains are also found at different recovery stages (PMSP 2011).
Silviculture in the APA is limited and consists of Pinussp followed by Eucaliptus sp planting. Activities such as pasture growing for livestock production and cropping are also developed, forming anthropic sites with low plant diversity. Subsistence agriculture is the main productive activity of APA inhabitants, although they are not landowners. Public policies attempt to implement sustainable agriculture practices among these farmers; however, the area is at risk of environmental impact (PMSP 2011).
Environmental heterogeneity of the Capivari-Monos APA promotes biodiversity. Development of Culicidae fauna in particular benefits from the abundance of suitable breeding sites and blood sources. In this scenario, direct contact between humans and mosquitoes facilitates enzootic cycles of infectious and parasitic disease agents (Iversson 1994, Figueiredo 2007).
The study of mosquito ecology is essential in order to understand the epidemiology of different diseases. These investigations provide useful knowledge for vector control, allowing monitoring of these biological indicators of environmental changes caused by man. In fact, environmental degradation promotes the proliferation of mosquito species with adaptive plasticity, which can develop in suburban areas, showing a domiciliation tendency and ability to carry and transmit pathogens to humans and animals (Forattini et al. 1989, Forattini et al. 1991).
In light of human proximity to forest areas and consequent interaction with disease vectors, autochthonous malaria outbreaks have been reported in the Capivari-Monos APA since 2006. This triggered interest in studying the previously unknown mosquito fauna in the area. However, in addition to Anopheline mosquitoes and malaria risk, the situation described is a potential public health problem since other local vectors can carry a number of arboviruses, causing diseases such as yellow fever, dengue fever, and encephalomyelitis. Aggravating this situation, humans are accidentally inserted into the natural cycle of some pathogens, eventually participating as primary, secondary, or dead-end hosts (Marcondes 2009).
In an attempt to better understand the epidemiological potential of mosquitoes in the Capivari-Monos APA, the present study describes their diversity profile and similarity within this region and discusses the potential risk of vector contact with the local human population.
MATERIALS AND METHODS
The study was conducted in a remnant of Atlantic Rainforest in the Capivari-Monos APA, located in the subdistrict of Parelheiros, south of São Paulo and almost 60 km from the city center. The protected area includes part of the Billings and Guarapiranga hydrographic basins, and the Capivari and Monos River basins, important water sources for the city of São Paulo (PMSP 2011).
Entomological collections were carried out in two areas. The first was a wildlife zone (WLZ) close to the São Paulo- Santos railway line at the Evangelista de Souza station (S 23° 56′ 14,09″/W 46°38′ 09,07″). According to current legislation for the Capivari-Monos APA, the WLZ is an integrated conservation area for the preservation of the biota and water resources. The region is predominantly covered by advanced stage secondary forest, with limited human presence and influence. The second area, denominated the anthropic environment, is located 15 km from the WLZ in an agricultural zone (AZ) in Embura (S 23° 53′ 14.36″/W 46°44′ 25.96″). The AZ was formed to promote sustainable development among APA inhabitants with adequate agricultural management and land use. Rural areas are close to forest fragments, which suffer intense anthropic pressure. Predominant plant cover formation is forests at initial or intermediate succession stage. Occurrence of malaria has been recorded in the AZ and culicids at this site were collected in domiciliary and peri-domiciliary environments.
A high occurrence of epiphytic bromeliads was detected in both environments. The characteristic leaf shape of these plants form natural breeding sites that house immature culicids of different species. Climate in the area is predominantly super-humid oceanic tropical, typical of the eastern portion of the Serra do Mar (mountain range) (PMSP 2011).
To estimate diversity patterns, mosquitoes were captured once a month from May, 2009 to June, 2010 with CDC-type automatic traps, aspirator and Shannon traps. Six CDC-type traps containing dry ice as attractant were installed at three fixed points in the anthropic environment and three other points in the wild area, operating for 12 h (from dusk to dawn) during each collection, totaling a 1,008-h sampling effort in each environment. The aspirator (12V) captured mosquitoes at rest by suctioning for 20 min during each sampling, with a total sampling effort of 840 min per area. The Shannon traps, fed with a gas lamp to attract the mosquitoes, operated for 3 h (beginning at dusk) during each collection, with a 42-h sampling effort for each environment.
Mosquitoes were identified in the Entomology Laboratory at the Public Health Faculty of the University of São Paulo. Abbreviations used for species names were in accordance with Reinert (2001) and identifications were determined following Forattini (2002).
Data for each environment were analyzed irrespective of capture techniques used. According to Magurran (2004), the number of species and their relative contribution to the community are important in assessing diversity. As such, given that richness and evenness are complimentary parameters, they are the most suitable for analyzing diversity. Specific richness (S) and Shannon-Wiener (H') indices were used to assess diversity, the Simpson's index (DS) to determine dominance, and Pielou's index (J') to evaluate evenness. Similarity between environments was established using the Jaccard's index (IJ) (Magurran 2004). Indices were applied using version 1.88 of the PAST software package (Hammer and Harper 2009).
The diversity pattern of environments was also used for comparisons. This procedure compares richness and evenness using a diversity ordering technique known as Rényi diversity profiles:
where, Hα is the diversity index value for factor α (α≥ 0, α≠ 1) and p1, p2, p3, …, pn corresponds to the number of individuals from species 1, 2, 3 … S (Melo 2008).
The classification established by Ott and Carvalho (2001) was applied to determine species dominance, using the formula D%= (i/t) × 100, where i = number of individuals of a same species and t = number of individuals captured. According to the D values obtained, five dominance classes were established: eudominant (over 10%), dominant (5 to 10%), subdominant (2 to 5%), recedent (1 – 2%), and subrecedent (under 1%).
The constancy index was calculated with the following formula: C = p/N × 100, where p = number of samples belonging to a given species; P = total number of samples analyzed. Species were then classified into three constancy categories: constant (C > 50%), accessory (25<C<50%), and accidental (C < 25%) (Dajos 1983).
The individual-based rarefaction curve was used to analyze sample sufficiency and compare environments (Gotelli and Colwell 2001). Rarefaction curve, error bar (Sobs Mao Tau) and Jackknife estimates were calculated with EstimateS 8.2 (Colwell 2009).
Monthly data on mosquito captures from May, 2009 to June, 2010 (Table 1) showed higher mosquito abundance from December, 2009 to February, 2010 in the anthropic environment and from October to December, 2009 in the wild environment (Figure 1).
Table 1. Adult mosquitoes collected in different traps in the wild and anthropic environments in the subdistrict of Parelheiros, municipality of São Paulo, from May, 2009 to June, 2010.
We collected 9,403 adult mosquitoes from 91 taxonomical categories. Of these, 6,911 (73.5%) were captured in the wild environment and 2,492 (27.5%) in the anthropic environment. The most frequent species in the former was Anopheles (Kerteszia) cruzii (n = 5,371), followed by the Melanoconion section of Culex (Melanoconion) (n = 373) and Aedes serratus (n = 246). In the anthropic environment, the most common species were Coquillettidia chrysonotum/albifera (n = 407), Culex (Culex) from the Coronator group (n = 321), and An. (Ker.) cruzii (n = 304) (Table 1).
Shannon traps yielded the highest An. (Ker.) cruzii capture efficiency in both environments (n = 4,827; 90%). CDC traps captured the highest number of Cx. (Mel.) Melanoconion section in the wild area (n=299, 75%) and Cx. (Cux.) Coronator group in the anthropic environment. Cq. chrysonotum/albifera species, collected in Shannon traps, were abundant in the anthropic environment (Table 1).
The most constant and dominant (eudominant) species was An. (Ker.) cruzii, occurring in all captures, with relative abundance above 10%. Melanoconion section of Cx. (Mel.), Ae. serratus and Cq. chrysonotum/albifera were classified as constant, but were subdominant since abundance was below 5% (Table 1).
The rarefaction curve for species (Figure 2) showed sampling efficiency in both wild and anthropic environments, while the plot revealed higher species richness in the anthropic environment. The jacknife technique indicated that both environments have a tendency towards a same richness with increased sampling effort. However, mean richness (and standard deviation) observed differed from the mean jackknife value (and standard deviation), indicating that total richness was not estimated for these environments.
The diversity pattern (Figure 3) shows similar species richness between the areas, whereas diversity and evenness were greater in the anthropic environment. Low evenness in the wild environment reflects high species dominance, since 75% (n = 5,371) of individuals captured corresponded to An. cruzii (Table 2). Richness of the samples captured in CDC traps was higher than in samples collected by the other methods, irrespective of the environment. However, Shannon traps captured the highest number of individuals. The samples collected with CDC traps and aspirator exhibited the highest diversity and homogeneity (Table 3).
Table 2. Mosquito diversity in the wild and anthropic environments in the subdistrict of Parelheiros, municipality of São Paulo, from May, 2009 to June, 2010.
An. (Ker.) cruzii
Table 3. Diversity of mosquitoes collected in the wild and anthropic environments using different capture techniques, in the subdistrict of Parelheiros, municipality of São Paulo, from May, 2009 to June, 2010.
Jaccard's Index demonstrated similarity of 55% between the wild and anthropic environments, revealing a high similarity in species composition. Of the species collected, 50 were found in both environments, 19 were exclusive to the anthropic environment, and 22 to the wild environment. The primary exclusive species in the anthropic environment was Cx. (Mel.) delpontei (n = 48), and Ocellatus section of Cx. (n = 35) in the wild environment.
We identified substantial Culicidae richness in the Capivari-Monos APA. This area, on the outskirts of the city of São Paulo, has approximately 11 million inhabitants and is recognized as an important economic center in Brazil. Factors such as intensive urban growth along with disordered encroachment of spring and forest areas favors contact between humans and several species of mosquitoes, facilitating the transmission of arboviruses and other parasites. Some species captured deserve special attention due to their epidemiological importance. Although species richness in both environments was similar, they differed in terms of abundance.
In the present study, captures with CDC traps and aspirator provided samples of higher diversity than those obtained with Shannon traps. The highest richness index values in anthropic and wild environments were obtained in CDC trap samples. However, mosquitoes captured with the aspirator exhibited the highest diversity. According to Brown et al. (2008), these results are important for comparing different types of traps since richness and diversity are not necessarily associated.
The diversity profile of the anthropic environment was characterized by higher richness and evenness. Man-modified environments are known to eventually become favorable to greater biological diversity. Although these modifications can compromise vertebrates, they stimulate proliferation of invertebrates such as culicids (Hunter 2007). The anthropic environment is an ecotone, a transition area between urban and wild regions. Modification in this environment promotes species dispersion, causing changes in the population of native communities. Moreover, proximity of the forest edge and higher variability of breeders in relation to the wild environment may have resulted in greater richness and evenness, as well as lower species dominance.
Following the same reasoning, environmental changes may have increased the abundance of opportunistic species. Species of the tribe Mansoniini dominated the anthropic environment which, according to Dorvillé (1996), indicates a high level of environmental change. Species from this tribe are attracted by light and displayed marked nocturnal activity (Barghini et al. 2004). They are epidemiologically significant because they are naturally infected by several arboviruses, including transmitters of Venezuelan equine encephalitis (Forattini 2002). In addition to pathogen transmission, culicids can attack and thus disturb local residents and animals (Consoli and Lourenço-de-Oliveira 1994, Forattini 2002).
An important epidemiological finding was the occurrence of 321 specimens of An. cruzii in the anthropic environment over the 14 months of the study. Although this number is lower than that detected in the wild environment, sporadic autochthonous cases of malaria have been notified in the anthropic area.
The presence of An. cruzii and Cx. neglectus is also ecologically significant, given that these species are associated with undisturbed primary environments and can therefore indicate forest recovery. Similar results on mosquito fauna were found in an Atlantic Forest fragment in southern Brazil (Cardoso et al. 2011).
The anthropic environment exhibited man-made modifications such as cropped areas, irrigation ponds, and open areas surrounding the forest. It also contains several endogenous and exogenous bromeliads growing in peri-domiciliary areas, as well as tree assemblages at different succession stages, frequently colonized by epiphytes. Such environmental diversity favors the occurrence of wild and domestic mosquito species owing to the availability of breeding sites, proximity to the wild environment, and human and animal hosts.
Several studies report high dominance of An. cruzii in Atlantic Forest fragments, corroborating this aspect as an indicator of preservation (Forattini et al. 1990, Dorvillé 1996, Guimarães et al. 2000). In the present study, An. cruzii accounted for more than 70% (n = 5,067) of individuals captured in the wild environment, suggesting it is subjected to low anthropic influence, thereby preserving characteristics.
According to Forattini et al. (2000), despite their low synanthropy, An. cruzii are found in anthropic areas since they ensure blood sources for food. However, the species is not adapted to artificial environments and move to the surrounding forest after feeding, where females lay eggs on water corpuses accumulated in the axils of bromeliads. This ensures the epidemiological cycle of bromelian malaria in the Atlantic Forest of south and southeastern Brazil (Downs and Pittendrich 1946).
Close proximity between humans and natural malaria vectors in the study area is alarming because it heightens the risk of malaria occurrence. An. cruzii is a vector of human and simian malaria (Deane et al. 1970, Deane 1992), and some research has reported its occurrence, naturally infected with Plasmodium vivax and its variants, in Atlantic Forest areas of São Paulo (Branquinho et al. 1997) and more recently in Espírito Santo state (Rezende et al. 2009). The occurrence of zoonosis from simian plasmodia (Plasmodium simium and P. brasilianum, which are similar to P. vivax and P. malariae, respectively) in Atlantic Forest areas remains a matter of debate (Duarte et al. 2008). In light of malaria cases recorded, Anopheles sp. occurrence and particularly An. cruzii abundance, research into natural infection of this vector in the areAs shown by the collector's curve and Jacknife estimate, total richness in the environments was not assessed. However, the 91 taxonomical groups found in the two environments were sufficiently high to indicate significant high richness.
Notably, a substantial number of species from the tribe Culicini were captured, particularly in the wild environment. Forty species were identified, including Culex specimens from the subgenus Melanoconion. Some of these culicids can adapt to modified environments and may therefore introduce and transmit several arboviruses to humans and animals within the anthropic habitat, promoting and maintaining enzootic cycles (Forattini et al. 1991, Natal et al. 1998). Furthermore, according to Forattini (2002), one of the main biological vectors of filariasis belongs to the genus Culex. However, since most Melanoconion species are exclusive to wild areas, their occurrence in forest environments has low impact.
The presence of Psorophora ferox, Aedes scapularis, and Aedes serratus reinforces the epidemiological importance of the Capivari-Monos APA, as well as the need for monitoring possible arbovirus transmission. The Rocio virus, for instance, was found to infect Ps. ferox specimens in natural habitats (Lopes et al. 1981), and Ae. scapularis under laboratory conditions. This last finding was recorded following the encephalitis outbreak in Vale do Ribeira during the 1970s (Mitchell and Forattini 1984, Mitchell et al. 1986). Although these studies only demonstrate the ability of these culicids to carry pathogens, they play a significant role in virus circulation in the natural environment (Forattini 2002). The most abundant species of Aedini was Ae. serratus. Its relevance from a public health standpoint is poorly known, although natural infection records suggest that it has potential to carry arboviruses. Ae. serratus, considered a secondary vector of the Ilhéus virus (Vasconcelos et. al. 1998) was also found to be infected by the Trocara virus in the Peruvian Amazon and in the state of Pará (Travassos da Rosa et al. 2001, Turell et al. 2005). In addition, the occurrence of infected Anopheles, Aedes, Culex, and Uranotaenia mosquitoes with unclassified viruses and uncharacterized viruses of the families Bunyaviridae, Coronaviridae, Flaviviridae, and Rhabdoviridae was found in the West African rainforest (Junglen et al. 2009).
Several authors have suggested that anthropogenic changes should not be underestimated in the recrudescence of vector-borne diseases. For instance, a study investigating a region in France from World War II to 1971 found that the populations of An. Hyrcanus, a vector with potential for malaria transmission, and of the West Nile Virus vector Culex modestus were initially very high in agricultural areas, decreasing after a few years owing to pesticide control in rice cultivation (Ponçon et al. 2007). However, a number of factors caused the population of theses mosquitoes to increase again in 2000, raising the risk of malaria and WNV transmission.
Aedes aegypti was not recorded in captures and Aedes albopictus was scarce in the anthropic environment. However, incidences of dengue have increased over the last two years in other urban areas of the Parelheiros region. In accordance with SUVIS (Health Surveillance Supervision) records, dengue occurrences grow progressively from downtown Parelheiros towards Vargem Grande and Engenheiro Marsilac. According to these data, 37 cases were recorded in 2009, with four of them autochthonous. In 2010, incidences rose to 74, with nine autochthonous cases, and by October 2011 a further 95 occurrences were registered, one of which was autochthonous (personal communication).
In regard to possible reservoirs of wild arboviruses, the APA area contains 364 vertebrate species including amphibians, reptiles, birds, and mammals, corresponding to 67% of the fauna recorded in the city of São Paulo. Avifauna in particular is highly diverse in the APA, with 288 species accounting for 77% of those recorded in the city. The 35 mammal species found in the region correspond to 42% of total municipality records in accordance with an SVMA (Municipal Secretariat of Green Areas and the Environment) survey (SMA 2010). This fauna diversity enables the existence of several arboviroses in the natural environment, including encephalitis.
In conclusion, the study area displays a culicid-rich fauna, with species relevant to public health. Diversity, richness, and evenness components may help to control different pathogens in the Capivari-Monos APA. However, further research in the area is needed, since continuous environmental change caused by human activity results in selective pressure and consequent adaptations of culicid vectors and hosts, which may produce new epidemiological scenarios. Given the existence of mosquitoes able to carry arboviruses and inhabit both wild and anthropic environments, new vectors capable of transmitting pathogens from natural focal points to the anthropic environment may emerge.
The present study provides information to guide local malaria control and prevent metaxenic diseases in Atlantic Rainforest areas subject to human pressure and environmental degradation. As such, we strongly recommend mosquito monitoring by surveillance and control agencies in urban areas surrounded by Atlantic Forest.
We thank the Supervisão de Vigilância em Saúde, Coordenação de Vigilância em Saúde, Secretaria Municipal de Saúde, Prefeitura Municipal de São Paulo for providing data on dengue prevalence; and Instituto Pedro Matajs and Guarda Civil Metropolitana, Prefeitura Municipal de São Paulo for logistic support in field collections. This research was sponsored by FAPESP (2008/52016-0).