The phlebotomine fauna (Diptera: Psychodidae) of Guaraí, state of Tocantins, with an emphasis on the putative vectors of American cutaneous leishmaniasis in rural settlement and periurban areas

Phlebotomine sandflies were captured in rural settlement and periurban areas of the municipality of Guaraí in the state of Tocantins (TO), an endemic area of American cutaneous leishmaniasis (ACL). Forty-three phlebotomine species were identified, nine of which have already been recognised as ACL vectors. Eleven species were recorded for the first time in TO. Nyssomyia whitmani was the most abundant species, followed by Evandromyia bourrouli, Nyssomyia antunesi and Psychodopygus complexus. The Shannon-Wiener diversity index and the evenness index were higher in the rural settlement area than in the periurban area. The evaluation of different ecotopes within the rural area showed the highest frequencies of Ev. bourrouli and Ny. antunesi in chicken coops, whereas Ny. whitmani predominated in this ecotope in the periurban area. In the rural settlement area, Ev. bourrouli was the most frequently captured species in automatic light traps and Ps. complexus was the most prevalent in Shannon trap captures. The rural settlement environment exhibited greater phlebotomine biodiversity than the periurban area. Ps. complexus and Psychodopygus ayrozai naturally infected with Leishmania (Viannia) braziliensis were identified. The data identified Ny. whitmani as a potential ACL vector in the periurban area, whereas Ps. complexus was more prevalent in the rural environment associated with settlements.

Sandfly capture and taxonomy -Sandflies were captured from January 2005-June 2008 using CDC light traps (HP model) (Pugedo et al. 2005) and Shannon traps (Sudia & Chamberlain 1962). The light traps were used in peridomestic areas near animal shelters or in the forest every month for three consecutive nights from 06:00 pm-06:00 am. Light traps were placed in each ecotope 1 m above the ground. The captured specimens were fixed in 70% alcohol and identified in accordance with Galati (1995Galati ( , 2003; the abbreviation of the species' names conforms to Marcondes (2007). Shannon traps were used in MS 1 in a forested part of the rural settlement area over a period of 12 h (06:00 pm-06:00 am) monthly from March-June 2008. The anthropophily of the species was also evaluated in these captures. A single capture with a Shannon trap was made in November 2008 on four consecutive nights to search for natural Leishmania spp infection in phlebotomine females using multiplex polymerase chain reaction (PCR). The insects were kept alive in nylon cages (Barraud 1929) and were taken to the laboratory (Entomology Laboratory of the Center for Zoonosis Control of Guaraí). For sandfly identification, males and the last abdominal segments of the females were cleared and mounted on slides. Males were identified by analysing the morphology of the genitalia (gonostyle, gonocoxite, structures of the parameres, pump and ejaculatory ducts) as well as the colour of the thighs and thoracic pleurae; the spermathecae, individual ducts and common duct were also observed in females. After identification, the insects were grouped into pools of 10 females and 10 males by species and locality. Male insects were used in diagnostic assays as negative controls. The pools of sandflies were kept at -20ºC for DNA extraction.
Statistical analysis -The index of species abundance (ISA) and the standardised index of species abundance (SISA) (Roberts & Hsi 1979) were calculated for the phlebotomine species collected in the rural and periurban MSs. Excel 2010 and Diversidade de Espécies v. 2.0 software (DivEs) were used to analyse the data. The data were entered into Tables organised by MS, species and period of capture. Each column was classified separately according to the number of specimens of each species. The values in each column were ranked, with the highest value classified as 1, the second as 2 and so on. ISA was calculated according to this formula: where k is the number of columns in the table (number of months of collection), a is the number of columns in which the species was absent multiplied by c, c is the highest value of the ranking obtained in all columns + 1 and Rj is the sum of the values for each species.
The minimum and maximum limits of the index were determined according to the highest distribution classification, where the limit is different in each series of data. To avoid this variation and to standardise the index we used a scale of values between 0-1 to calculate SISA: Species abundance was considered high when the SISA value was close to the maximum value of 1. The results provide information on the relative abundance of species and on the temporal distribution of the collected individuals.
To analyse and compare the overall number of phlebotomine insects captured in the rural and urban areas, the DivEs software was used to analyse the data using the Shannon-Wiener diversity index (H) and the evenness index (J) (Rodrigues 2005).
PCR multiplex studies -To detect Leishmania spp infection, molecular analysis was performed on a total of 250 females and 40 males, which were grouped into pools of 10 specimens each. DNA was extracted as previously described (Pita-Pereira et al. 2005). Multiplex PCR was designed to simultaneously amplify the cacophony gene IVS6 region in Neotropical sandflies, which was used as an internal control for the polymerase enzyme activity and DNA extraction and the conserved kinetoplast DNA minicircle from Leishmania spp. The amplified products were further analysed by dot blot hybridisation with a biotinylated Leishmania (Viannia)-specific probe (Pita-Pereira et al. 2005). Rigorous procedures were used to prevent contamination: negative control groups (male sandflies) were included in the DNA extraction step, instrument and working areas were decontaminated with diluted chloride solution and ultraviolet light and artificially infected females were added as positive controls.
The H and the J analyses revealed that the indexes recorded in the rural settlement environment (H = 0.9641 and J = 0.6059) were higher than those of the periurban environment (H = 0.7758 and J = 0.5252) (Table III).
In the rural settlement area, 2,515 specimens representing 39 species were captured, 13 of which were found exclusively in this environment:  (Table III). When evaluating the overall number of specimens captured per ecotope, it was observed that the MS 2 forested area had the most individuals (n = 848), followed by other environments in which animals were present, such as chicken coops: MS 1 (n = 707), MS 2 (n = 695) and MS 4 (n = 581) (Supplementary data).
The H and the J indexes of the evaluated ecotopes showed the highest values in the banana grove ecotope  The analysis of species frequency per ecotope showed that Ev. bourrouli was present in all ecotopes and predominated in the rural settlement area, as well as in the banana grove ecotope at MS 4 in the periurban area. Ny. antunesi was the predominant species in the rural settle-ment area chicken coop at MS 2 (42.01%). Ny. whitmani was the most prevalent in the other ecotopes near domestic animal shelters in the periurban area, such as the chicken coop at MS 1 (44.62%), the pig sty at MS 2 (49.52%) and the chicken coop at MS 3 (56.63%), followed by Lu. longipalpis in all ecotopes (Supplementary data).
The Shannon trap captures at the MS1 rural settlement area yielded 1,096 specimens, including 190 males (18%) and 906 (82%) females, resulting in a female/male ratio of 4.55:1.0. Among the 14 species identified, those from the Psychodopygus genus were the most prevalent, including Ps. complexus, Ps. llanosmartinsi and Ps. ayrozai. All of these species were observed biting humans during the captures. Among these species, Ps. complexus (42.61%) was predominant (Table IV).
A total of 290 specimens were collected and used for PCR and dot blot hybridisation. Sandflies of the same species were analysed in pools of 10; there were 23 Ps. complexus pools (3 male and 20 female pools) and six Ps. ayrozai pools (1 male and 5 female pools) ( Table V). The results showed that four of 25 (16%) female pools were positive for Leishmania (Viannia) sp. infection. In the Ps. complexus pools, only three of 20 were found to be positive (15%). In the Ps. ayrozai pools (n = 5), one pool (20%) was positive for infection (Supplementary data).

DISCUSSION
Initial studies of phlebotomine fauna in TO focused on the description of species (Barreto 1946, Martins et al. 1962, 1964, 1975; further entomological studies identified 32 sandfly species in the state (Lustosa et al. 1968, Andrade Filho et al. 2001), including the new species Micropygomyia (Silvamyia) echinatopharynx and Martinsmyia reginae (Andrade Filho et al. 2004, Carvalho et al. 2010. These findings indicate a great diversity of sandfly species in TO. Recent investigations of ACL and AVL vectors in rhe municipality of Porto Nacional recorded 48 sandfly species, 22 of which were the first records of these species in the state (Vilela et al. 2011).
The present study identified 43 phlebotomine species in Guaraí, 11 of which were newly recorded species, underscoring the rich biodiversity of the Brazilian Cerrado. Among the identified sandfly species, Ev. bourrouli was predominant in the rural settlement area and presented the second highest percentage in the periurban area. This species was found in various environments and ecotopes, such as preserved forest and animal shelters (chicken coops and pig sties), which underscores its eclectic behaviour. In previous studies carried out in TO, this species was the most frequent at Porto Nacional along with   Evandromyia sallesi (Andrade Filho et al. 2001). However, it is important to emphasise that this species has not been reported to transmit pathogens to humans. Of the species identified in this study, 11 were recorded in TO for the first time, including Ps. complexus, Ps. davisi, Ps. hirsutus hirsutus, Ps. ayrozai, Ps. paraensis and T. ubiquitalis, which are putative ACL vectors in Brazil; other sandfly vectors of leishmaniasis that had already been recorded in TO were also found in Guaraí, including Ny. antunesi, Nyssomyia flaviscutellata, Ny. whitmani, Ny. intermedia, Mig. (Mig.) migonei and Lu. longipalpis.
Previous studies developed in TO identified Ny. whitmani as the main vector of L. (V.) braziliensis, especially in areas affected by hydroelectric construction and agricultural activities, where this sandfly has been found inside and outside residences, near animal shelters and even within forest boundaries (Carvalho 2008, Vilela et al. 2008, 2011. Ny. whitmani is associated with the transmission of L. (V.) braziliensis in most Brazilian endemic areas, occupying different types of plant cover and modified environments (da Costa et al. 2007, Rangel 2010. Ny. whitmani was the most abundant species among the species collected in the present study. Furthermore, it was the most abundant ACL vector in periurban areas and one of the most represented species in the rural settlement areas. This species was found in all ecotopes, with a particularly high density in animal shelters (chicken coops and pig stys) from residences in the periurban environment. Studies performed in Porto Nacional discussed the capacity of Ny. whitmani to adapt to environmental changes by expanding its transmission cycle (Vilela et al. 2011).
Of the ACL vectors observed in Guaraí, it is important to highlight the identification of Ps. complexus, an important L. (V.) braziliensis vector in the low-altitude forests of the state of Pará (PA) in the Amazon Region of northern Brazil (Souza et al. 1996, Lainson & Shaw 2005, Garcez et al. 2009, Rangel & Lainson 2009). This sandfly species has been recorded outside of the Amazon Region, but there has been no evidence to suggest its involvement in L. (V.) braziliensis transmission in the states of Pernambuco (Andrade et al. 2005) and Mato Grosso (Azevedo et al. 2002, Ribeiro et al. 2007).
The present study provides evidence for the presence of Ps. complexus in all evaluated ecotopes, including animal shelters in the periurban area. Although Ny. whitmani is thought of as the most important ACL vector in TO (Rangel & Lainson 2009, Vilela et al. 2011, the discovery of naturally L. (V.) braziliensis-infected Ps. complexus, the spatial distribution of this sandfly in forest and peridomiciliary ecotopes, its predominance in Shannon traps and its relatively high abundance in settlements and periurban areas, together with epidemiological evidence reported in the literature, suggest that Ps. complexus may also play a role in the transmission cycle of ACL in the rural settlement areas of Guaraí.
Ps. ayrozai, another sandfly species that is naturally infected by L. (V.) braziliensis in Guaraí, has been suggested as a putative vector of Leishmania (Viannia) naiffi in PA (Lainson et al. 1994, Lainson & Shaw 2005. Ps. ayrozai is an anthropophilic species (de Aguiar et al. 1985, Gomes & Galati 1989 and experimental infection studies have revealed its susceptibility to Leishmania (Leishmania) forattinii (Barretto et al. 1986, Lainson & Shaw 2005. Despite reports showing the vector competence of Ps. ayrozai, this species was not found frequently in this study and no further evidence was found of its participation in local transmission cycles. Ny. antunesi, which was found in the rural settlement area, has been proven to be a Leishmania (Leishmania) lindenbergi vector in the northern PA and has been identified as anthropophilic in previous studies conducted in TO by Andrade Filho et al. (2001); however, cases of human infection by L. (V.) lindenbergi have not been reported in this study area thus far.
The ecoepidemiology of ACL in Brazil is strongly related to the transmission cycle of L. (V.) braziliensis and the vector Ny. whitmani, particularly in environmentally impacted areas, rural environments and the periphery of cities from all geographical regions. This new epidemiological profile has resulted from drastic environmental changes and the capacity of this sandfly to adapt to new ecological niches.
The Cerrado biome has experienced significant environmental impacts caused by deforestation. In this context, the establishment of highly populated, poor rural settlement groups and periurban areas without adequate infrastructure typically results in close contact between people and pathogen vectors.