Sandflies ( Diptera : Psychodidae ) in rural and urban environments in an endemic area of cutaneous leishmaniasis in southern Brazil

Leishmaniasis ranks second among the six most common infectious and parasitic diseases in the world and approximately 350 million people are at risk for infection. As many as 12 million people are believed to be currently infected, with approximately one-two million estimated new cases occurring every year (WHO 2010). With the exception of Oceania and Antarctica, all continents have reported human cases of leishmaniasis. Cutaneous leishmaniasis constitutes a public health problem in 88 countries on four continents (the Americas, Europe, Africa and Asia) (WHO 2010). In Brazil, American cutaneous leishmaniasis (ACL) is a dermatological disease that calls for special attention due to the consequent risk of deformities and psychological effects. ACL affects both the social and economic lives of patients. ACL is widespread in Brazil and cases have been reported in all regions of the country. It affects both sexes and all age groups; however, 90% of ACL cases occur in patients over 10 years old and 74% in males. ACL is considered an occupational disease (MS 2007). A total of 260,486 ACL cases were reported in Brazil between 2000-2009: 6,625 (2.6%) occurred in the southern region and 6,010 (90.7%) of those occurred in the state of Paraná (PR) [Health Informatics Department of the Brazilian Ministry of Health (datasus.gov.br/DATASUS/index.php)] (MS 2010). In this state, the majority of the cases have been concentrated in the northern areas since 1940 when an extensive clearing of native vegetation (dense and transitional tropical/subtropical forest) began (Teodoro et al. 1991). More recently, the disease has been associated with areas of intense soybean, corn and sugar cane cultivation and pastures near vestigial forests (Lonardoni et al. 2006). Studies of the sandfly fauna in PR have been undertaken more extensively in the northern region, where the greatest numbers of ACL cases have been registered (Gomes & Galati 1977, Teodoro et al. 1991, 1993a, b, 2001, 2006, Teodoro & Kuhl 1997, Oliveira et al. 2000, Massafera et al. 2005). Nyssomyia neivai (Pinto, 1926), Nyssomyia whitmani (Antunes & Coutinho, 1939), Pintomyia pessoai (Coutinho & Barretto, 1940), Pintomyia fischeri (Pinto, 1926) and Migonemyia migonei (França, 1920) are the most commonly suspected vectors of Leishmania sp. found in PR (Zanzarine et al. 2005). Ny. neivai (denoted as Lutzomyia intermedia, s.lat. in many studies) and Ny. whitmani are consistently present in high numbers in peridomestic and domestic areas in most parts of northern PR (Teodoro et al. 1993a, b, Teodoro & Kuhl 1997, Oliveira et al. 2000, Massafera et al. 2005). The high urban prevalence of ACL cases (81/123; 66%) between 2000-2009 in the municipality of Bandeirantes (Cruz 2010) created the need to investigate the behaviours of these insects in both urban and rural environments to better understand the epidemiology of the disease.

online | memorias.ioc.fiocruz.br Mem Inst Oswaldo Cruz,Rio de Janeiro,Vol. 108(3): 303-311, May 2013 Leishmaniasis ranks second among the six most common infectious and parasitic diseases in the world and approximately 350 million people are at risk for infection. As many as 12 million people are believed to be currently infected, with approximately one-two million estimated new cases occurring every year (WHO 2010).
With the exception of Oceania and Antarctica, all continents have reported human cases of leishmaniasis. Cutaneous leishmaniasis constitutes a public health problem in 88 countries on four continents (the Americas, Europe, Africa and Asia) (WHO 2010). In Brazil, American cutaneous leishmaniasis (ACL) is a dermatological disease that calls for special attention due to the consequent risk of deformities and psychological effects. ACL affects both the social and economic lives of patients. ACL is widespread in Brazil and cases have been reported in all regions of the country. It affects both sexes and all age groups; however, 90% of ACL cases occur in patients over 10 years old and 74% in males. ACL is considered an occupational disease (MS 2007).
A total of 260,486 ACL cases were reported in Brazil between 2000-2009: 6,625 (2.6%) occurred in the southern region and 6,010 (90.7%) of those occurred in the state of Paraná (PR) [Health Informatics Department of the Brazilian Ministry of Health (datasus.gov.br/DATA-SUS/index.php)] (MS 2010). In this state, the majority of the cases have been concentrated in the northern areas since 1940 when an extensive clearing of native vegetation (dense and transitional tropical/subtropical forest) began (Teodoro et al. 1991). More recently, the disease has been associated with areas of intense soybean, corn and sugar cane cultivation and pastures near vestigial forests (Lonardoni et al. 2006).

MATERIALS AND METHODS
Area of study -The municipality of Bandeirantes is located in northern PR ("Old North"). The town of Bandeirantes is located at 23º06'36"S 50º27'28"W and 420 m above sea level (asl). In 2010, the municipality had 32,182 inhabitants and the urban population was 28,382 [Brazilian Institute of Geography and Statistics (IBGE) (ibge.gov.br/cidadesat/topwindow.htm?1)].
The municipality has an area of 446,301 km², including the urban area of 12 km². The soil is a ferruginous clay composed of basalt, also known as red soil (Panchoni et al. 2008).
The annual rainfall ranges from 1,300-1,400 mm, with approximately 150-200 mm in January and approximately 50-75 mm in June and July. The prevailing winds vary from southwesterly to northwesterly with an average speed of 2.5 m/s. The climate is subtropical with hot summers and infrequent frosts during the winter in the lowlands. High levels of rainfall occur in the summer months, but there is no clearly defined dry season. The average annual temperature is 21ºC, the monthly average in the summer is between 22-28ºC and the monthly average during the winter is between 14-18ºC (Panchoni et al. 2008 Capture sites -Cases of cutaneous leishmaniasis were reported between 2000-2009 in all of the sampling sites. Eleven urban sites were sampled: (1) Itapeva smallholding (near the old piggery), (2) Souto smallholding (henhouse), (3) União suburb (a tree in the backyard and several dogs), (4) Lordani neighbourhood (deposit of recyclable materials), (5) Humberto Teixeira neighbourhood (jaboticaba tree close to dwelling), (6) house in front of the cemetery, (7) Clubhouse of the Athletic Association of the Bank of Brazil (avocado tree near the house of the caretaker), (8) Bela Vista neighbourhood (guava tree in the peridomicile near a sugarcane plantation), (9) Faculties Luis Meneghel (FALM) (tree next to the Nursing department), (10) rural village one (avocado tree) and (11) rural village two (pigsty). Although points 10 and 11 are called "rural villages", both are included in the urban area of Bandeirantes and are located approximately 3 km from the city centre. In the rural area, three ecotopes in each of the following four neighbourhoods were investigated. (1) Jacutinga locality -the ecotopes sampled are situated in a fishing are (a large leisure area open to visitors from Tuesday-Sunday; on weekends the number of visitors is higher and they usually remain until late at night): (1.1) pigsty (10 m from the house and 130 m from the forest edge), (1.2) forest edge (under a guava tree, 120 m from the dwelling and 200 m from the lake) and (1.3) peridomicile.
(2) Tabuleta -this locality consists of a small social centre with a church, soccer field and several plots of land where sugar cane, coffee, alfalfa, soybeans and corn, among other crops, are cultivated; the residence of an ACL case is also located there. The ecotopes sampled were: (2.1) a tree close to the veranda of a dwelling, (2.2) peridomicile, an avocado tree used as a perch by chickens and (2.3) peridomicile, a tree close to the pigsty. (3) Perobinha -in this locality there is a village with approximately 40 plots of land (5,000 m 2 ) where families grow fruits and vegetables and breed pigs, poultry and cattle. This locality also has a social centre with a primary school, church, meeting hall and soccer field and the dwelling of an ACL case is also found there. The ecotopes sampled were: (3.1) peridomicile, under a jaboticaba tree close to the house, (3.2) under an orange tree close to a pigsty and a dwelling and (3.3) a bamboo grove 10 m from the dwelling and next to a pasture, a pond and a small forest. Lastly, (4) Cabiúna -where there are settlements and a small social centre with a bar, school, church and soccer field. The ecotopes sampled were: (4.1) a pigsty close to a dwelling, (4.2) a pigsty situated at the forest edge (50 m from the same dwelling) and (4.3) the forest edge (under a tree and 50 m away from point 4.2).
Collection and identification of sandflies -Sandflies were captured in the urban areas using automatic light traps (Galati et al. 2010) installed fortnightly between 06:00 pm-07:00 am from April 2008-March 2009 in 11 peridomiciles. In the rural areas, two techniques were used: automatic light traps installed one day per month at the 12 points sampled from 06:00 pm-07:00 am from April 2009-March 2010 and at one of these sites (Jacutinga neighbourhood -forest edge of the fishing area), a modified black and white Shannon trap was also used (Galati et al. 2001). This trap was used once during each of the following seasons: spring 2009 (06:00 pm-03:00 am), summer 2010 (06:00 pm-04:00 am) and fall 2010 (06:00 pm-06:00 am). The captures in the spring and summer were interrupted before 06.00 am due to heavy rains with winds.
The insects were sacrificed with chloroform and packed into entomological boxes. At the Public Health Entomology Laboratory/Phlebotominae of the Department of Epidemiology, School of Public Health, University of São Paulo, the sandflies were separated from the other insects captured. After clarification, they were identified in accordance with Galati's keys (Galati 2003).
Statistical analysis -The numerical abundance according to the spatial distribution of sandfly species was calculated based on the Standardized Index of Species Abundance (SISA) (Roberts & Hsi 1979) and the values that are closest to 1.0 represent the most abundant species. The index was calculated for all the species collected using the CDC traps in rural localities. The Williams' geometric average (Haddow 1960), which reflects the frequency and regularity of the species in the collections, was used to evaluate the activity of the most abundant sandfly species. The Shannon diversity index, which indicates the diversity and frequency of the sandfly species at particular capture sites, and Pielou's evenness index were obtained in accordance with the formulas presented by Service (1993).
The pluviometric data and temperature were obtained from the meteorological station closest to Bandeirantes (50 km away), which was situated in the municipality of Ourinhos, in the state of São Paulo, from the Agrometeorology Information Center (ciiagro. sp.gov.br/dados/entrada.htm).
In the urban area, 264 collections with automatic light traps were performed, but only 22 specimens (0.083 sandflies/trap) belonging to six species were captured: Br. brumpti, Ev. cortelezzii,Ny. neivai,Ny. whitmani,Pi. pessoai and Sc. sordellii. Ny. neivai predominated (65%) (Table I). Of the 11 sites sampled, no sandflies were obtained at two locations (the henhouse on the Souto smallholding and jaboticaba tree in the Humberto Teixeira neighbourhood). The sites with the largest number of sandflies (three specimens each) were the "rural village" suburb, the Itapeva smallholding, the house in front of the cemetery and FALM.
A total of 3,843 specimens were captured with automatic light traps (144 traps installed) in the rural area, yielding an average of 26.69 sandflies/trap. The most prevalent species were: Ny. neivai (42.83%), Ny. whitmani (37.34%), Br. brumpti (8.07%) and Pi. pessoai (6.12%). Ny neivai was the most abundant (SISA = 0.897), followed by Ny. whitmani (0.862), Pi. pessoai (0.710), Pi. fischeri (0.399) and Mg. migonei (0.339). Of the total of sandflies captured in the rural areas, 65.31% were collected in Jacutinga (96.2% of which were collected at the   (14), followed by Perobinha (9), Cabiúna (7) and Tabuleta (4). The highest diversity index values were observed in Jacutinga at the forest edge under a guava tree (1.45) and in the pigsty (1.42). This index was zero in a tree close to the pigsty in the Tabuleta neighbourhood and this index was very low (0.48) in the jaboticaba tree close to the home at the Perobinha. In the ecotopes of the Tabuleta neighbourhood, the smallest numbers of insects were captured, but these specimens had the highest diversity indices (0.91-0.96). Two ecotopes in Perobinha (an orange tree close the pigsty and a bamboo grove) yielded the second and third greatest frequencies of sandflies captured (13.2% and 10.4%, respectively); however their diversity indices were the lowest of all the samples sites (0.45 and 0.43) and had a predominance of the two Nyssomyia species (Table II).
Amongst the most abundant species, Ny. neivai and Ny. whitmani predominated throughout the study period. Ny. whitmani showed higher monthly averages than Ny. neivai from January-May and in August. Pi. pessoai was the third most frequent species, with the highest averages captured in February and March. The highest average captures of Ny. neivai ranged from 4.28-2.31 insects/ month and the highest average captures of Ny. whitmani ranged from 4.34-2.67 insects/month. The lowest average captures for both species occurred in July (Fig. 1).
The seasonal distribution by site and ecotope of Ny. neivai and Ny. whitmani is presented in Table III. For the Ny. neivai captured in Jacutiga, the lowest Williams' averages were observed in all of the ecotopes during the fall and the highest Williams' averages were obtained in the spring. In Perobinha, the averages were remarkably higher in the summer on the edge of the forest and in the pigsty. In Cabiuna, the averages were generally low, with the exception of the pigsty and peridomicile during the summer. Concerning Ny. whitmani captured in Jacutinga, the lowest Williams' averages occurred in the fall for all the ecotopes and the highest occurred during the summer at the forest edge. In Perobinha, the averages were much higher in the summer for all of the ecotopes. In Cabiuna, the highest average occurred in the fall along the edge of the forest and in the summer the highest averages were found in the pigsty and the peridomicile.
Ten species were captured using the Shannon traps. The most frequently captured species were Ny. neivai (56.9%) and Ny. whitmani (32.8%) and the white trap captured significantly more specimens of these two taxa than the black trap (Table IV). In April 2010, the capture was conducted from 06:00 pm-06:00 am, but the other two captures were interrupted before 06:00 am due to heavy rain and rough winds. During the common capture period (06:00 pm-03:00 am), the highest average capture of Ny. neivai and Ny. whitmani, for both sexes, occurred using the white trap between 11:00 pm-12:00 am. Using the black trap, the females of these two species were most frequently captured between 09:00 pm-11:00 pm (Fig. 2).

DISCUSSION
The frequencies of the specimens of the subtribes Psychodopygina (83.2%), Lutzomyiina (10.5%), Brumptomyiina (6.1%) and Sergentomyiina (0.2%) found in this present study are in agreement with those found in this same municipality by Massafera et al. (2005). However, after capturing 38,662 sandfly specimens in 37 municipalities, da Silva et al. (2008) found that 86.1% of the sandflies belonged to Psychodopygina, 12.7% to Lutzomyiina, 0.8% to Sergentomyiina and 0.4% to Brumptomyiina; the two last subtribes inverted their positions in the ranking. However, within the subtribe Psychodopygina, the da Silva et al. (2008) study found that Ny. neivai (47.4%) predominated over Ny. whitmani (35.8%) and the frequency of Pi. pessoai (6.5%) was also close to that observed in the present study. These three species are considered vectors of ACL (Pessôa & Coutinho 1941, de Queiroz et al. 1994, Rangel & Lainson 2003, Pita-Pereira et al. 2005, Andrade Filho et al. 2007, Cruz et al. 2012. The predominance of Ny. neivai and its greater abundance have also been observed in the municipality of Itambaracá, which borders Bandeirantes (Cruz et al. 2012). In other studies, different results have been found for PR, where Ny. whitmani prevailed in the northwestern region (67.82%) (Teodoro et al. 1991), in the northern region (76.3%, 70%) (Teodoro et al. 1993a, Muniz et al. 2006) and even in Bandeirantes (58.9%) (Massafera et al. 2005). Ny. neivai and Ny. whitmani were also found to be the predominate species in 10 municipalities in the northwestern region of PR .
The sandflies captured in urban areas in Brazil have frequently demonstrated the ability of some species to adapt to anthropogenic environments (Teodoro et al. 1998). These authors also demonstrated the presence of nine sandfly species in the residual forests within the urban perimeter of Maringá and Ny. whitmani was overwhelmingly dominant. In the present study Ny. neivai was observed, although at a much lower frequency, to occupy urban environments. This species is also prevalent in the city of Piçarras in the state of Santa Catarina, near the homes of patients with ACL, where female sandflies naturally infected with Leishmania (Viannia) sp. have been found (Marcondes et al. 2009); this sandfly species has also been found in a periurban area of Porto Alegre in the state of Rio Grande do Sul (Pita-Pereira et al. 2009).
The high prevalence of Ny. whitmani in altered residual forests in urban environments in PR (Teodoro et al. 2001, Zanzarine et al. 2005 and in many rural regions of northern PR (da Silva et al. 2008) demonstrate that this species is adapting to the anthropic environment.
In Bandeirantes, the vector species Ny. neivai, Ny. whitmani, Pi. pessoai, Pi. fischeri and Mg. migonei have been frequently found in domestic animal shelters, woods and dwellings. Pi. fischeri naturally infected with Leishmania braziliensis has only recently been found in the Southeast Region of Brazil (Margonari et al. 2010, Rocha et al. 2010), but the other species have all been found naturally infected by Leishmania (Viannia) sp. in PR and in other areas of South America (de Queiroz et al. 1994, da Silva & Grunewald 1999, Luz et al. 2000, Pita-Pereira et al. 2005, 2009 The greatest species richness and highest diversity index were observed in a rural district in Jacutinga, near the forest edge and a pigsty, possibly due to the following characteristics: the presence of dense forest near the domestic animal shelters and the presence of orchards and the accumulation of organic matter. These factors are considered favourable for sandfly breeding sites. The high species balance in Tabuleta can be explained by the similar frequencies amongst the species. The low frequency of specimens and low level of species richness at this site may be attributed to the absence of dense forest and the constant presence of strong winds. However, sandfly vectors are still present, which explains the occurrence of cutaneous leishmaniasis in a resident of this site. The proximity of domestic animal shelters to dwellings, the accumulation of organic matter (leaves and fallen fruit, agricultural waste, animal droppings and household food scraps), wet soil due to the discharge of water from domestic use and the shade of trees are favourable for the formation of outdoor sandfly breeding sites (Massafera et al. 2005). These conditions are observed in rural areas of PR and were also present in the area studied, resulting in large numbers of sandflies collected around the dwellings and in domestic animal shelters.
The increased attractiveness of the Shannon traps (310 sandflies/trap/person) compared with the automatic light traps (26.6 sandflies/trap) may be an indication of the high anthropophily of the two predominant species. The significantly increased attractiveness of the white Shannon trap compared with that of the black trap was observed for both Ny. neivai and Ny. whitmani. A similar result was found in the Ribeira Valley in SP for Ny. neivai (Galati et al. 2010); however, the black trap attached significantly more Ny. whitmani than the white trap on the Bodoquena Range in the state of Mato Grosso do Sul (MS) (Galati et al. 2001). The monthly variation of Ny. neivai and Ny. whitmani show that they have great plasticity; although they predominated during the rainy periods, they also peaked during the dry season. This trend was also observed for the Ny. neivai by Galati et al. (2010), but not for the either species by Massafera et al. (2005), whose peaks occurred only during the rainy season. On the other hand, Ny. whitmani only peaked during the dry periods in MS (Galati et al. 1996).
Using the white Shannon traps, both sexes of Ny. neivai were captured at the greatest frequencies from 10:00 pm-02:00 am. In the Ribeira Valley, SP (Galati et al. 2010) and in Bandeirantes (Massafera et al. 2005), the peak frequencies of Ny. neivai capture occurred from 07:00 pm-11:00 am. For Ny. whitmani captured using the white trap, there were two peaks for the females and the last peak was at a time close to that found in Maringá, PR (12:00 am-01:00 am) (Teodoro et al. 2003), but differed from that in Bandeirantes (08:00 pm-10:00 pm) (Massafera et al. 2005) and in the municipality of Corguinho, MS (06:00 pm-07:00 pm) (Galati et al. 1996).  Although seasonality studies of sandflies are still needed, the results obtained in this present study suggest that the spring and summer are the seasons of greatest risk of human infection by Leishmania. A high frequency of sandflies was found in domiciliary areas and the observation that the highest frequencies occur late at night when people are resting suggests that the infection occurs at home. Some people may have been infected during leisure or occupational activities associated with the forest, especially in the Jacutinga neighbourhood; for example, people may have been infected at the fishing area, where the highest vector density occurred. Epidemiological research to investigate the probable sites where people acquire Leishmania infection is required to guide the entomological surveillance and control measures to diminish the sandfly population and educational activities to alert the public to the risks of infection should be implemented in Bandeirantes.