Enzootic mosquito vector species at equine encephalitis transmission foci in the República de Panamá

The identification of mosquito vector species present at arboviral enzootic transmission foci is important to understand transmission eco-epidemiology and to propose and implement prevention and control strategies that reduce vector-borne equine encephalitis transmission. The goal of this study was to identify mosquito species potentially involved in the transmission of enzootic equine encephalitis, in relation to their abundance and diversity at three endemic regions in the República de Panamá. We sampled adult mosquitoes during the dry and rainy season of Panamá. We employed CDC light traps with octanol, EV traps with CO2 and Trinidad 17 traps baited with live hamsters. Traps were deployed in the peridomicile and extradomicile of houses from 18:00 to 6:00 h. We estimated the abundance and diversity of sampled species. We collected a total of 4868 mosquitoes, belonging to 45 species and 11 genera, over 216 sampling nights. Culex (Melanoconion) pedroi, a major Venezuelan equine encephalitis vector was relatively rare (< 2.0% of all sampled mosquitoes). We also found Cx. (Mel) adamesi, Cx. (Mel) crybda, Cx. (Mel) ocossa, Cx. (Mel) spissipes, Cx. (Mel) taeniopus, Cx. (Mel) vomerifer, Aedes scapularis, Ae. angustivittatus, Coquillettidia venezuelensis, Cx. nigripalpus, Cx. declarator, Mansonia titillans, M. pseudotitillans and Psorophora ferox all species known to be vectorially competent for the transmission of arboviruses. Abundance and diversity of mosquitoes in the sampled locations was high, when compared with similar surveys in temperate areas. Information from previous reports about vectorial competence / capacity of the sampled mosquito species suggest that sampled locations have all the elements to support enzootic outbreaks of Venezuelan and Eastern equine encephalitides.

Introduction in a previous field study elsewhere or which has been experimentally shown as competent to transmit the pathogen in the laboratory [13]. In this study we present results from a series of mosquito surveys in three regions with a history of equine encephalitis transmission, placing an emphasis on the diversity and abundance of potential VEEV and EEEV vectors.

Study site
We designed this study to compare mosquito species composition from three enzootic arbovirus (VEEV and EEEV) transmission foci in the Panamá and Darién provinces and the autonomous indigenous Comarca (territorial political division assigned to indigenous groups) Ngäbe Buglé. In the selected study areas infections in humans, horses and/or wildlife animals have been reported [4,6], or in mosquito pools identified to the genus level [27]. Darién is the easternmost province in Panamá, bordering Colombia. The natural landscape is dominated by tropical rainforest and the climate is tropical with an extended dry season. Total annual rainfall is over 2500 mm, with one or two dry months with less than 60 mm. Temperature ranges between 18 and 23˚C around the year [28]. In this province we selected the following locations for mosquito sampling: Mercadeo with 36 households and 206 inhabitants, Santa Librada with 170 households and 300 inhabitants and Los Pavitos with 30 houses and 95 inhabitants. Western Panamá Province has a warm pre-mountain humid tropical rainforest. Annual rainfall adds to 1571 mm, with a mean annual temperature of 26.5˚C [29]. Here, we selected El Cacao and Ciri Grande as sampling locations. Ngäbe Buglé is also covered by premountain tropical rainforest and has an annual rainfall around 400 mm and mean temperatures around 25˚C year-round. Here, we collected mosquitoes at Pumona. In all the sampling locations it is worth highlighting that the landscape is very homogeneous from the standpoint of ecological disturbances, since at the local scale of our sampling locations there was a similar mix of forest and cattle farming grounds near households. In all studied locations the main economic activities are cattle farming, wood extraction and subsistence agriculture.

Mosquito sampling
At each sampling location we put three kinds of traps over three consecutive nights (18:00 to 6:00) at 1.5 m above the ground in peridomiciliary areas and forests near to the houses, hereafter referred as extradomicile [30,31]. In each locality we employed 10 CDC light traps baited with octanol (Fig 2A), eight modified Trinidad 17 (TT-17) traps (Fig 2B), baited with one live hamster and eight EVS traps ( Fig 2C) baited with CO 2 [25]. Sampling was done during February-March (dry season) and September-October (rainy season) of 2011 and 2012, trying to sample species from both the dry and rainy season. Collected mosquitoes were killed, by flashfreezing, soon after collection and identified at the genus level in the field. Samples were then placed in plastic vials by trap type and sampling date and stored in liquid nitrogen before transportation to the Departamento de Entomología Médica at the Instituto Conmemorativo Gorgas de Estudios de la Salud, where identification at the species level, whenever possible, was performed using taxonomic keys [24,32,33] and the reference collection at the Institute.

Data analysis
We used mosquito species abundance data to estimate mosquito species relative abundance and diversity using the software EstimateS, 8.2.0™ [34]. We specifically estimated the Simpson and Shannon-Wiener diversity indices to compare patterns of diversity at each study site. The choice of these two indices was done given the emphasis of the former on dominant species, as opposed to the latter which focuses on the whole community [35]. We also estimated species richness by counting the number of species at each site and by estimating the Margalef index. Finally, we estimated species similarity between sampling locations using the Sorensen index. We also estimated the median abundance and its SE for females of all collected taxonomic units. For the analyses we used the additive mosquito counts, from all three types of traps, for each mosquito taxonomic unit. In all the analyses we considered taxonomic units identified at the genus level as a distinct species, since they likely included, in all cases, individuals belonging to species not identifiable with morphological keys.

Ethical clearance
No permits were required since humans were not involved in the study. Use of hamsters was approved by the "Comité Institucional de Uso y Cuidado de Animales de Laboratorio" (CIU-CAL) of Instituto Conmemorativo Gorgas de Estudios de la Salud, in accordance with law No. 23 of January 15 1997 (Animal Welfare Assurance) of República de Panamá, as presented within the research protocol of project "Estudio del subtipo ID del complejo de Encefalitis Equina Venezolana en Panamá".

Materials and data availability
All data analyzed in the results section are presented within the text of this article. Voucher specimens of collected mosquito species are available at the Colección de Insectos, Departmento de Entomología Medica, Instituto Conmemorativo Gorgas de Estudios de la Salud.

Results
For each of the six study locations we sampled a total of 36 nights, totaling 216 sampling nights across all study sites. We collected a total of 4868 mosquitoes belonging to the following 11    Table 1. Mosquito species composition and abundance in three regions with equine encephalitis transmission in the Repú blica de Panamá. Data are presented as total by region. The sampling trap-nights effort is indicated by n. Please, note that sampling effort was the same at the location level, and differences in n reflect differences in the number of sampled locations by region.  Table 2 and Table 3). The highest abundance and richness of species was found  (Table 4) showed that Mercadeo and Santa Librada, in Darién, shared 70% of the mosquito species, the highest species similarity observed in this study. The extent of species similarity was also high between Santa Librada and Los Pavitos (65% of species shared), both located in Darién. Although with a lower species richness, El Cacao and Ciri Grande, both in Panamá, had high mosquito species similarity (58% of species shared), similar to what was observed for Los Pavitos and Pumona (58% of species shared). Interestingly, Pumona shared over half of the species with all other sampling locations but Ciri Grande (40%) ( Table 4).

Discussion
Knowledge about mosquito species diversity in transmission areas is fundamental to understand the entomological risk of vector-borne disease transmission, given that slight bionomic differences between species can lead to significant differences in transmission, the persistence of a disease, or the ability of a vector-borne disease to spread into new host species [36][37][38][39][40]. The mosquito diversity patterns we observed are within what is normally expected for ecological communities of mosquitoes and other diptera species, where local environmental factors are  (Fig 1) had more similar faunas, as inferred from the higher Soresen similarity ( Table 4). The most species rich region was Darién, followed by Ngäbe Buglé and then by Panamá. Here, it is important to highlight this result likely not only reflects the larger sampling effort at Darién, but also that individual Darién sampling locations had a higher species richness when compared to locations in the other two studied regions. Mercadeo was the sampling site with the highest mosquito abundance and species richness, including most of the Culex (Melanoconion) spp, which include many major equine encephalitides vector species [9,44], in contrast with sites from Panamá and Ngäbe Buglé where species from this subgenus were either absent or not identifiable at the species level. This result is very important since it implies a potentially higher entomological risk for enzootic VEEV transmission in Darién, something that could explain the common occurrence of VEE and EEE outbreaks in this region over recent years [4,45]. By contrast Ciri Grande had the lowest species richness, a high abundance of Culex spp. (64.2%).
A detailed examination of the species we collected reveals that from the 45 species (out of a total of 52 taxonomic units) we collected, at least 22 species have been reported as VEEV vectors in Panamá or elsewhere in the New World [3,9,26]. The species previously identified as VEEV vectors include 10 Culex spp., eight belonging to the Melanoconium subgenus, Spissipes  [12,[58][59][60]. C. venezuelensis is associated with permanent water bodies with floating vegetation [61]. It is a vector of Mayaro, Oropuche, VEE and SLE viruses [33, 62,63] and West Nile virus [64]. VEEV has also been isolated from Ps. ferox and Ps. albipes [65]. VEEV subtypes IC and IAB have been isolated from M. indubitans, M. titillans, M. dyari, Ps. confinnis and An. pseudopunctipennis [13,66,67]. Ps. albipes, Ae. serratus and Ae. fulvus are susceptible to the infection with VEE [13,68]. Ae. angustivittatus has been found infected with Ilheus virus in Panamá and VEEV in Colombia [69][70][71]. Ae. scapularis has been incriminated as VEEV vector in epizootic and enzootic outbreaks [70,72,73]. Cx. nigripalpus was collected at the three sites in Darién. This species is able to colonize urban and rural landscapes and exhibits a catholic bloodfeeding [74,75]. This species is a major SLE virus vector in the USA [76], but also in Central America, Ecuador and Trinidad and Tobago [77]. Cx. coronator, also collected in this study at Darién, has an ecology similar to that of Cx. nigripalpus and has been found infected with SLE virus [71] and Mucambo virus in the Brazilian Amazon [78].
A major limitation of our study was our inability to identify a large proportion of Culex spp. mosquitoes and other specimens that we were only able to identify at the genus level (29.8%; 1453/4868). This was mostly due to poor specimen conditions, but also to some mosquitoes having distinctive features from those of species described in taxomic keys for mosquito species of the New World. In that sense it would be desirable to develop a barcoding library to molecularly identify all mosquitoes present in the República de Panamá, as has been done elsewhere [83]. This can help to both aid the description of new species and with the identification of morphologically damaged specimens. A second limitation was the sampling during night time, which could have limited the possibility of sampling Haemagogus spp, of which we only found one mosquito, and Sabethes spp, which we did not collect. Both Haemagogus and Sabethes are genera with species known to be active during daytime, and which include some species that are medically important, given their role in the transmission of yellow fever virus, another major arbovirus [84,85]. Similarly, the study would have greatly benefited by sampling mosquitoes in areas where no alphavirus transmission has been detected, in order to better understand the role of dominant vector species on disease transmission [38,86] or mosquito diversity on infection [41], while also looking at domestic and wildlife reservoirs, as done for other zoonotic vector borne diseases, for example Leishmaniasis [87][88][89][90], in the República de Panamá and for alphaviruses in other regions of Latin America [9].
Finally, we would like to highlight this report is the first to describe the mosquito fauna of locations that have frequently reported VEE outbreaks in the República de Panamá. We were able to identify 22 species that are vectorially competent for VEEV transmission, and other species that also transmit medically important arboviruses and parasites across the New World [9,23,44]. This result is very important as it is a first step for further research looking at the ecology of VEEV-mosquito interactions in order to better understand the enzootic transmission of this and related viruses, especially the invasion of new areas by VEEV [51,91], as well as, transmission during the inter-epizootic periods in the República de Panamá. Further research is needed to better understand why, even though all our study sites had a similar environment, where primary and secondary forest were mixed with cattle farming and agricultural land, in places like the sites in Panamá province there were very few Culex (Melanoconion) spp, even though they have been found previously in this region [10,92], and they were common in the two other study regions. elimination and prevention of cattle screwworm (COPEG) and the Regional Directorate of Animal Health of the Ministry of Agricultural Development of the Darién.

Author Contributions
Conceptualization: Scott Weaver.