Introduction

The Culicidae family is represented by nearly 3,610 species, among which approximately 150 species belonging to the genus Culex, Aedes, Anopheles and Haemagogus are involved in transmission of pathogenic agents to animals and humans (^{GUEDES, 2012}). The immature forms of these culicids develop in aquatic environments that are present in domestic, peridomestic and forest areas. Some species, such as Culex (Culex) quinquefasciatus, may develop in water containing decomposing organic material, such as in small pits and sewage areas, while others, such as Stegomyia (Stegomyia) aegypti and Stegomyia(Stegomyia) albopicta, use a variety of objects (e.g. tires, cans, bottles, plates and vases, water tanks, buckets, bamboo cavities and bromeliad tanks) containing water with smaller quantities of decomposing organic material and salts, as their breeding grounds (CONSOLI; OLIVEIRA, 1994).

Culicids have great medical-veterinary importance because they are vectors of different pathogens and parasites that cause diseases (e.g. arboviruses, protozoa and helminths) (^{CARVALHO et al., 2013}; ^{DIBO et al., 2011}; FORATTINI, 2002; ^{PINTO et al., 2011}). In Brazil, the state of Pernambuco is considered an endemic area for several pathogens transmitted by mosquitoes. For example, human filariasis, caused by Wuchereria bancrofti, has been a serious sanitary problem in Recife (the state capital) and adjacent areas for a long time (^{FONTES et al., 2012}; ^{MEDEIROS et al., 2006}). Similarly, canine filariasis caused by Dirofilaria immitis, a parasite with zoonotic potential, is frequently reported in northeastern region of Brazil, including the state of Pernambuco (^{ALVES et al., 1999}; ^{CARVALHO et al., 2008}, ²⁰¹³; ^{SILVA; LANGONI, 2009}).

Close contact between culicids and human/animal populations, in regions characterized as fragments of forests embedded in urban or rural environments, facilitates transmission of pathogenic agents, being the humans easily involved in enzootic cycles of infectious and parasitic diseases (^{MONTES, 2005}). Considering the sanitary importance of these invertebrates, studies on their occurrences in certain areas are essential for providing information about the periods of risk to populations, and also for gathering data that is useful in effectively controlling these vectors. Therefore, the aim of this study was to evaluate occurrences of immature forms of culicids, over an one-year period, in a specific area of northeastern region of Brazil.

Materials and Methods

Collection and identification of culicids

Third and fourth-instar culicid larvae were collected every week from artificial breeding grounds, between October 2012 and September 2013, totaling 48 collections during the whole study. For this, 23 traps constructed from plastic (PET) bottles that were cut and painted black, were distributed in domestic (n = 5), peridomestic (n = 5; three on the ground and two at the treetops) and forest environments (n = 13; six on the ground and seven spread out across the treetops) (Figure 1), with distances of 20 meters between the traps. The treetop traps were positioned at a height of 2 m from the ground.

Figure 1. Trap constructed from plastic (PET) bottle, located in the forest environment on the ground. 

The larvae thus collected were transported in plastic vials containing water and were then identified in the laboratory. They were mounted between a slide and a coverslip, and viewed through an optical microscope (at magnifications of 10X and 40X). Larval samples (10%) from each breeding ground was maintained in a colony (at temperature conditions of 25 ± 2 °C and relative air humidity of 80 ± 5%) until adults were obtained, in order to confirm the species. The larvae and adults were identified in accordance with previously proposed dichotomous keys (^{ALMIRÓN; HARBACH, 1996}; CONSOLI; OLIVEIRA, 1994; ^{FORATTINI, 1965}, 2002).

The abbreviations used for the genus and subgenus are presented according to ^{Reinert (2009)}.

Results

A total of 12,718 culicid larvae (average of 1,059.83 specimens/month) were collected during the study, and they were detected in all months of the year (Figure 2). The highest number of specimens was detected in September 2013 (n = 4,084) and the lowest in December 2012 (n = 86). In particular, 2,441 (19.2%), 7,098 (55.9%) and 3,179 (24.9%) larvae were collected in the domestic, peridomestic and forest areas, respectively. With regards to the positioning of traps (peridomestic and forest areas), 8,585 specimens (67.5%) were collected from the ground and 1,692 (13.3%) from treetops.

Figure 2. Occurrence of immature forms of culicids at Cabo de Santo Agostinho from October 2012 to September 2013. 

Ten different species of culicids were identified, being S. (S.) albopicta(46.45%; 5,908/12,718), Culex(Culex) maxi (35.56%; 4,523/12,718) and Limatus durhamii (12.58%; 1,600/12,718) the most frequent (Table 1).

In the traps distributed in the domestic environments, the species S. (S.) albopictapredominated (95.58%; 2,333/2,441), while in those positioned in the peridomestic areas, the species C.(C.) maxi predominated on the ground (73.22%; 4,523/6,177) and S.(S.) albopicta in the treetops (97.94%; 902/921). In the traps located in the forest areas, several species were found: L. durhamii predominated on the ground (50.33%; 1,212/2,408), while S.(S.) albopicta was also found in large numbers on the ground (38.12%; 918/2,408) and predominated in the treetops (55.71%; 430/771). Ochlerotatus(Ochlerotatus) scapularis, Anopheles (Kerteszia) bellator, Anopheles(Nyssorhynchus) braziliensis and Mansonia (Mansonia) titillans were found exclusively in the peridomestic area, whereas C. (C.) quinquefasciatus was found in the forest area (Table 1).

The results showed that S.(S.) albopicta colonized all three environments, predominating in the domestic and peridomestic areas, which did not differ statistically from each other (x² = 2,417; p = 0.1200) but differed from the forest environment (domestic: x² = 263,042; p = 0.0000; peridomestic: x² = 215,632; p = 0.0000). The species L. durhamii also colonized all three environments, but predominantly in the forest environment, which differed statistically from the other environments (x² = 41,397; p = 0.0000). S. (S.) aegypti was detected in the domestic and peridomestic areas, but predominantly in the domestic environment (x²= 24,490; p = 0.0000).

In comparing the culicids' preference for trap location (ground or treetops) in the three environments, it was observed that S. (S.) albopictapresented greater frequency in the traps installed on the ground in the peridomestic areas and also in the forest, with a statistical difference (x² = 26,591; p = 0.0000). L. durhamii was also predominantly observed in traps placed on the ground, especially those installed in the forest (x² = 4,357; p = 0.0369).

The species were more abundant at medium temperatures ranging from 24.6 °C to 27.2 °C (Figure 3), with relative air humidity between 72.0% and 80.8%. The precipitation varied greatly, from 44.8 mm to 440.1 mm per month. In correlating occurrences of species with climatic variables, a weak correlation was found between precipitation and relative air humidity (r < 0.33 and r < 0.30, respectively). With regard to temperature, the correlation was negative and moderate (r < 0.50).

Figure 3. Climatic conditions observed in the study area from October 2012 to September 2013. 

The average monthly temperature ranged from 25.1 °C (October 2012) to 28.8 °C (December 2012). The relative air humidity ranged from 68.7% (November 2012) to 80.8% (June 2013). The driest months were November 2012 and February and July 2013. The average monthly precipitation oscillated from 9.9 mm (November 2012) to 440.1 mm (July 2013), such that the months of November 2012 and February 2013 were the months with least precipitation, and the months of June and July 2013 were the most rainy. During the study, a period of intense drought occurred in the area, thus making the climatic conditions atypical.

Discussion

This study evaluated occurrences of culicid larvae in an area with a tropical climate (northeastern Brazil) over an one-year period. Larvae were found during all months of the year, and were most frequent in the month of September 2013. In general, the study showed larval population peaks between the months of January and April, and between June and September.

The population fluctuations of culicid species in this study were probably related to the different environments from which they were collected (domestic, peridomestic and forest), and to the particular features of each species, such as their adaptation and better development under different climatic conditions. Considering that in this area the temperature and air humidity variations are minimal over the course of the year, it is most likely that precipitation levels play an important role in the population of culicids. The influence of climatic parameters on culicid larval abundance has been reported previously, presenting different results in many cases. For instance, the abundance of C. (C.) quinquefasciatus in the city of Córdoba (Argentina) was not correlated with variables such as temperature and rainfall (^{ALMIRÓN; BREWER, 1996}), while in Buenos Aires, the larval plethora of Culex (Culex) pipiens was positively associated with the mean weekly temperature (^{FISCHER et al., 2000}; ^{FISCHER; SCHWEIGMANN, 2004}).

It is known that the risk of pathogen transmission is higher when the culicid population increases. In this study, the species S.(S.) albopicta presented two populational peaks, firstly from February to April and secondly from June to September, coinciding with the populational peaks of C.(C.) quinquefasciatus and S. (S.) aegypti, respectively. It is likely that the larval abundance of culicids was not determined by a single climatic parameter, but by an association of variables such as environmental degradation and the diverse distribution of rainfall.

Although larvae were collected from all three study areas (i.e. domestic, peridomestic and forest), they were predominantly found in the peridomestic area. In addition, the specimens were more frequently found in the traps positioned on the ground than in those on treetops. These variations were probably related to the biological behavior of each species. Interestingly, O. (O.) scapularis, A. (K.) bellator, A. (N.) braziliensis and M.(M.) titillans were found exclusively in the peridomestic area, whereas C. (C.) quinquefasciatus was found in the forest area. Notwithstanding, larvae of O. (O.) scapularis are predominantly found in natural breeding sites at ground level, collection of these larvae from artificial sites has already been reported (FORATTINI, 2002). The retrieval of C.(C.) quinquefasciatus exclusively from forest areas may be related to the fact that this species is collected predominantly from shaded breeding sites. Thus, the forest environment presents suitable conditions for this species to develop.

Interestingly, the study area presented marked environmental degradation, which contributed to alterations in the natural breeding sites of Culicidae species. For instance, larvae that are commonly restricted to forest environments may be collected in domestic/peridomestic areas (^{GOMES et al., 1999}). This transition between the forest and domestic environments involves several mosquito species, which may play important roles in the wild epidemiological life cycles of pathogens, thus representing an important threat to public health.

Overall, a great variety of culicid species was observed, but three of them were more representative, i.e. S.(S.) albopicta, C.(C.) maxi and L. durhamii. Curiously, S.(S.) albopicta was the only species detected throughout the study (i.e. in all months of the year). The ability of the eggs of this species to tolerate desiccation (FORATTINI, 2002) is likely to have contributed towards maintaining the viability of this species throughout the year in the study area. Two populational peaks (from February to April and from June to September) were observed for this species. Therefore, occurrences of this species seem to be related to rainfall, since the two populational peaks overlap with the period of rains reported here. This species presents great epidemiological importance due to its capacity to be a vector for various arboviruses (^{TAIPE-LAGOS; NATAL, 2003}). Moreover, S. (S.) albopicta is a potential vector for the dengue virus, which causes epidemic disease throughout Brazil, and for D. immitis, the etiological agent of canine dirofilariasis, which also has a zoonotic nature (^{CARVALHO et al., 2013}). This culicid species was detected in traps in all three environments (domestic, peridomestic and forest), thus confirming its capacity to colonize a wide variety of artificial breeding grounds, without necessarily abandoning its natural ecotopes (^{GOMES et al., 1999}).

Although C. (C.) maxi was described in the state of Pernambuco only recently (^{RAMOS et al., 2011}), the great number of larvae detected here indicates that this mosquito species is well adapted to the environmental conditions present in the study area. It was the second most abundant species throughout the study, and its abundance was closely related to the rainy season, as observed in a previous study in Buenos Aires, Argentina (^{FISCHER; SCHWEIGMANN, 2004}). Considering that this species has been reported to be an important vector of arboviruses in other regions (^{FORATTINI, 1998}), its vector capacity in the region studied here needs to be investigated.

On the other hand, the species L. durhamii, which is typical of forest areas (FORATTINI, 2002), was also found in the domestic and peridomestic environments. This finding demonstrates the domiciliation of this species, as previously reported in the southeastern region of Brazil (^{REZENDE et al., 2011}). Similar behavior was observed for O. (O.) scapularis (formerly typical of forest areas), which was detected in peridomestic areas. Studies have indicated that this species is adapted to anthropic environments, especially areas modified by human beings (BRANCO et al., 2009; ^{TAIPE-LAGOS; NATAL, 2003}). The finding of a high number of L. durhamiispecimens during the rainy period corroborates with what was reported by ^{LIRA-VIEIRA et al. (2013)} in the central-western region of Brazil. Similarly, a higher number of O.(O.) scapularis larvae were detected with increasing rainfall and reached zero with reduced rainfall. These data differ from those of a previous study, in which a higher numbers of O.(O.) scapularis larvae were collected during the dry period (^{LIRA-VIEIRA et al., 2013}). The presence of O.(O.) scapularis in this area is of great importance, particularly considering that in previous studies, it was suggested that this species was a vector for D. immitis (^{ALVES et al., 1999}) and W. bancrofti (^{MEDEIROS et al., 2006}) on the coast of Pernambuco (same area of study). Similarly, the presence of A. (K.) bellator in this area is relevant, considering its importance as a vector of protozoa of the genus Plasmodium (^{FORATTINI et al., 1999}). Indeed, the possibility that monkeys might be infected by Plasmodium that affects humans and vice versa has been confirmed, thus demonstrating the zoonotic potential of these parasites (^{ARAÚJO et al., 2013}).

From an epidemiological perspective, development of species originally from forest areas in domestic and peridomestic areas represents a risk for the populations residing in these areas, since they are exposed to different diseases transmitted by culicids, such as dengue, yellow fever, Venezuelan equine encephalitis, human and canine filariasis and Rocio encephalitis (BRANCO et al., 2009; ^{FORATTINI et al., 1995}; FORATTINI, 2002; ^{NUNES et al., 2008}). The last disease mentioned (i.e. Rocio encephalitis) is considered an important emergent zoonosis that is endemic in some regions of Brazil. Although this disease is little known, humans affected by it present rigidity in the neck, mental confusion and balance disorders. Moreover, birds, marsupials and rodents are natural reservoirs of its etiological agent (^{NUNES et al., 2008}).

Although S. (S.) albopicta, the most frequent species studied in the area (i.e. Cabo de Santo Agostinho, Pernambuco), presented two populational peaks – from February to April and from June to September – it is possible to conclude that the animal and human populations residing in the area of study were exposed to culicids and to the agents throughout the year, especially after the rainy period. Therefore, culicid control and prevention measures need to be adopted in the study area in order to reduce the impact on health caused by the endemicity of the species reported here.