Vol. 43, no. 2 Journal of Vector Ecology 293

Diversity patterns of hematophagous in Atlantic forest fragments and human-modified areas of southern Bahia,

Lilian S. Catenacci1,2,3,4, Joaquim Nunes-Neto2, Sharon L. Deem4, Jamie L. Palmer4, Elizabeth S. Travassos-da Rosa2, and J. Sebastian Tello5,6

1Curso de Medicina Veterinária, Federal University of Piauí State/CPCE, Bom Jesus, PI, Brazil, [email protected] 2Division of Arbovirology and Hemorrhagic Fevers, Evandro Chagas Institute, Anannindeua, PA, Brazil 3Royal Zoological Society of Antwerp, Centre for Research and Conservation, Antwerp, Belgium 4Saint Louis Zoo, Institute for Conservation Medicine, St. Louis, MO, U.S.A. 5Missouri Botanical Garden, Center for Conservation and Sustainable Development, St. Louis, MO, U.S.A. 6Pontificia Universidad Católica del Ecuador, Escuela de Biología, Quito, Ecuador

Received 7 June 2018; Accepted 16 August 2018

ABSTRACT: There have been several important outbreaks of -borne diseases in the Neotropics in recent years, particularly in Brazil. Some taxa are also considered to be indicators of environmental health. Despite the importance of understanding abundance and distribution to the understanding of disease dynamics and design strategies to manage them, very little is known about their ecology in many tropical regions. We studied the abundance and diversity of mosquitoes and sand in the Bahia State of Brazil, a point of origin for arbovirus outbreaks, including Zika and Chikungunya fever. During 2009-2014, 51 mosquito taxa were identified, belonging to three dipteran families, Ceratopogonidae, Culicidae, and Psychodidae. The family Culicidae, including the Sabethini tribe, were the most abundant (81.5%) and most taxa-rich (n=45). While season (winter and summer) was a strong factor determinant of the occurrence of the most abundant taxa, the stratification level in the forest (ground or tree level) had a strong effect and the dominant taxa at ground level were completely different from the dominant species collected at tree level. We suggest that sites with a mix of forest and agroforestry systems support the highest biodiversity of hematophagous insects as compared to highly disturbed landscapes. Journal of Vector Ecology 43 (2): 293-304. 2018.

Keyword Index: Arbovirus, mosquito, Culicidae, abundance, richness, Brazil.

INTRODUCTION (CPCV), and recently, West Nile virus (WNV), chikungunya virus (CHIK), and Zika virus (ZIKV) (Figueiredo et al 2010, Many species of hematophagous insects, particularly Serra et al. 2016). In sylvatic areas, culicid mosquitoes also species in the families Ceratopogonidae, Culicidae, transmit pathogens to wild (Medeiros-Sousa et al. Psychodidae, and Simuliidae, are vectors of important 2013). Thus, rural communities where people, mosquitoes, tropical diseases (Vasconcelos and Calisher 2016, Cardoso and wild animals co-exist in close proximity are at an increased et al. 2017, Harbach 2017). Considering the emerging risk for the emergence of infectious diseases, particularly those arbovirus outbreaks in recent years throughout the world that are vector-borne, such as arboviruses. Understanding the (Figueiredo and Figueiredo 2014, Figueiredo and Figueiredo spatial and temporal variation in abundance and diversity of 2015, Vasconcelos and Calisher 2016), knowledge of disease culicid mosquitoes, as well as the role of climate in shaping vector distribution and biodiversity is essential to identify this variation, is important to assess disease risks in natural potential areas of high risk for pathogen transmission and environments (Pinto et al. 2009). Furthermore, the condition to conduct assessments of environmental health in protected of culicid populations can be used as a bioindicator for and anthropogenically disturbed areas (Navarro et al. 2015). environmental health assessments (Montes 2005, Paula et al. Among vectors, mosquitoes (Diptera: Culicidae) are 2015). particularly prominent. Approximately 3,500 species of Unfortunately, the ecology of mosquitoes in many parts Culicidae have been described worldwide, and Brazil is home of the tropics is poorly understood. Even though recent to about 450 species (Paula et al. 2015, Harbach 2017). The outbreaks of Zika and chikungunya in Brazil can be traced hematophagous Culicidae are divided into two subfamilies, back to rural areas of southern Bahia State (Brazil 2016), little Anophelinae and Culicinae. The Anophelinae are vectors of information exists about mosquito diversity, demography, many arboviruses and Plasmodium species and the subfamily or abundance. Moreover, this region contains remnants of Culicinae are vectors of a great variety of arboviruses, such as Atlantic forest, which are home to a large number of mammals, yellow fever virus (YFV), dengue virus (DENV) serotypes, St. birds, and (Ribeiro et al. 2009, Cassano et al. Louis encephalitis virus (SLEV), Ilheus virus (ILHV), Rocio 2011), some of which may be important for the circulation virus (ROCV), Bussuquara virus (BSQV), Cacipacore virus dynamics of many arboviruses (Souza et al. 2015). The goal 294 Journal of Vector Ecology December 2018 of this study was to investigate the entomofauna in sylvatic from November to April. Winter is the colder, drier part of and rural areas of southern Bahia, Brazil, and to evaluate the the year from May to October (Figure 2). effect of seasonality and the stratification level of the forest Samples were collected at seven different sites in sylvatic on taxa diversity. We also discuss in more detail and rural environments (Table 1). These sites were selected the Culicidae that are potential bioindicators and vectors of based on type of vegetation and proximity to areas where wild human and wildlife diseases. mammals have been monitored as part of other ecological and health studies (Oliveira et al. 2011, Catenacci et al. 2016). MATERIALS AND METHODS The sites with higher amounts of primary or second-growth forests were REBIO-Una, Ecoparque de Una, and Lagoa Study area Encantada, respectively. Three of the seven sites were sampled This study was conducted in the municipalities of Ilhéus in both “summer season” and “winter season,” accounting for and Una, located in the southern Bahia Atlantic Forests ten sampling events (Table 1 and Figure 2). of Brazil from 2006 to 2014 (Figure 1). Across this region, anthropogenic activities have resulted in a mosaic of land- Mosquito sampling use strategies. Native Atlantic forest is highly fragmented The Brazilian Federal and State Entomological and embedded within an agricultural matrix dominated by Services conducted the mosquito collections as part of cacao agroforestry systems, as well as plantations of rubber their surveillance program from 2009 to 2014. At each tree, coconut, banana, and cassava. The region also contains site, sampling lasted between three to ten days. Each day, the Una Biological Reserve and its Buffer Zone, the Una mosquitoes were surveyed during three time intervals: two Biological Wildlife Refuge (Alger and Caldas 1994, Sollberg during daylight (08:00-12:00 and 16:00-18:00) and one at et al. 2014). The area is surrounded by small villages where night (18:00-06:00). During daylight sampling, mosquitoes the majority of people are living by subsistence agriculture on were collected at ground and tree levels. At ground level, small family-based farms. Here, people have limited access three or more team members collected mosquitoes using to basic infrastructure, such as electricity, potable water, hand nets (polyester net bag 30 cm in diameter, attached to a sanitation, and health care (Carneiro Santos et al. 2014, 30 cm aluminum handle) constructed by the research team. Sollberg et al. 2014). With regards to climate, the mean annual At tree level, climbing equipment or a platform were used, temperature is 24° C, and annual rainfall averages 1,500 mm. and a trained person collected mosquitoes with a hand net at Based on the average temperature over the last three decades, 15 m high. Adult mosquitoes were transferred immediately to we define summer as the warmer and wetter part of the year, Eppendorf tubes using a manual suction tube (Castro catcher,

Figure 1. Study sites of hematophagous arthropod surveys: Ilhéus city (top) and Una city (bottom). Vol. 43, no. 2 Journal of Vector Ecology 295

Table 1. Study sites, including the dominant landscape, the municipality, and the geographic coordinates.

Study site Predominant type of vegetation* Cities Coordinates - Agroforestry with second-growth forest b,c 14º39’49.4’’S Almada Farm (site 1) - cacao farm Ilhéus - Neighbor of Bonfin and Santa Rita Farm 39º11’36.8’’W Bonfim Farm (site 2) - rural settlement -Agroforestry with second-growth forest b,c 14º39’31.24’’S Ilhéus with cacao farm - Neighbor of Almada and Santa Rita Farm 39º11’37.6’’W - Agroforestry with second-growth forest b,c 14º40’35.6’’S Santa Rita Farm (site 3) - cacao farm Ilhéus - Neighbor of Bonfim and Almada Farm 39º11’10.8’’W Colônia de Una Farm (site 4) - - Agriculture with second-growth forest b,d 15º17’11.1’’S Una Familiar agriculture farm - Close to REBIO-UNA 39º08’25.0’’W -Second-growth forest b with few agroforestry c, Lagoa Encantada (site 5) – 14º37’10.5’’S close to a big lake Ilhéus conservation area 39º08’14.6’’W - Close to Bonfim, Santa Rita and Almada Farm -Mostly old-growth foresta in the East side and Una Biological Reserve (Rebio-Una) Second-growth b forest with old-growth foresta 15º10’55.3’’S (site 6) – conservation area Una in the West side 39º04’20.7’’W -Neighbor of Ecoparque de Una Ecoparque de Una (site 7) – -Second-growth forest b 15º10’11.7’’S Una conservation área -Neighbor of Ecoparque de Una 39º03’16.4’’W *Vegetation types: aold-growth forest: forest with little or no sign of past human disturbance, a closed canopy, trees in general at least 20 m high with large diameters, bromeliads in a wide range of sizes and an extensive layer of vines; bSecond-growth forest: forest with visible signs of previous human disturbance, which has been subjected to either ‘general’ (recovering from complete deforestation) or ‘selective’ logging (recovering from the cutting of selected species). Bromeliads were present, but not as often as we found in the primary forest; cAgroforestry: forest in which the undergrowth has been cut and replaced by cacao trees and other agroforestry systems, such as banana, cupuaçu, and rubber trees; dAgriculture: plantation areas, including coconut, cassava tree, passiflora, and rubber trees. glass material with 50 length per 25 mm diameter). In the was calculated of the lowest taxonomic rank available for each lab, these samples were stored at appropriate humidity and insect collected. It is important to note that for the analyses temperature conditions (Serra et al. 2016). During nocturnal in this manuscript, the term “taxonomic unit” or “taxa” is sampling, two non-baited automatic CDC light traps used rather than “species” and “taxonomic diversity” rather (BioQuip Products®, CA, U.S.A.) were used and mosquitoes than “species diversity” because not all individuals could be were collected the following morning. These CDC light traps identified to species level. Thus, each taxonomic name was were set only at ground level, about 1.5 m above the ground. used as a unit of analysis, but these taxonomic units may At the field laboratory, mosquito samples were correspond to species, genera, or families. transferred to Eppendorf cryovials and frozen in liquid The structure of arthropod assemblages is described nitrogen. All samples were kept at -70º C in the Bahia State’s using the relative abundances of each family: Culicidae, Central Laboratory of Public Health until shipped by air Ceratopogonidae, and Psychodidae (Schmidt and Barcellos on dry ice to the Division of Arbovirology and Hemorragic 2007). This was done for all data combined, as well as for Fevers at Evandro Chagas Institute (SAARB-IEC) in Belém, each site, strata (ground-level or tree level), and season PA, Brazil. At the SAARB-IEC laboratory, mosquitoes were (summer and winter), separately. For the family Culicidae, identified using dichotomous keys (Consoli and Oliveira calculations were performed for the relative abundances of 1994). They were then grouped into pools of 1-91 specimens each taxonomic unit. These data were then used to construct by date, collection site, sex, and taxonomic identification. All abundance distribution curves, which describe how common pools were frozen and stored at -70° C in the SAARB-IEC for each taxonomic unit is relative to rare ones (Melo 2008). future arbovirus studies. Next, comparisons were done of levels of taxonomic diversity in Culicidae among sites, habitats, and seasons. Data analyses Because diversity within a site (or habitat or season) can During taxonomic work, we assigned each individual be high solely owing to a larger sampling effort, we used to the lowest possible identification level possible. Not all rarefaction curves to make our measures of taxonomic individuals, however, could be identified to species, and some richness comparable. Rarefaction is a very common statistical were identified to sub-, genus, or family level. To describe technique applied to biodiversity data. Rarefaction curves are the taxonomic resolution of the dataset used, the frequency produced by repeatedly re-sampling N number of individuals 296 Journal of Vector Ecology December 2018

Sabethini, and Uranotaeniini), and 14 genera (Coquillettidia, Culex, Mansonia, Psorophora, Aedes, Haemagogus, Limatus, Sabethes, Johnbelkinia, Runchomyia, Trichoprosopon, , Uranotaenia, and Wyeomyia). Sabethini was the dominant tribe, corresponding to 81.5% of individuals collected, followed by Aedini (6.8%), Mansoniini (2.9%), and Uranotaeniini (0.13%). Anophelinae represented 7.7% of samples collected. When considering all sites together, three taxonomic units showed the highest abundance: one genera Wyeomyia spp. (20.1%), one subgenera Wyeomyia (Phoniomyia) spp. (6.7%), and one species, the Limatus pseudomethysticus (39.8%) (Figure 4a). Table 2 shows the three most dominant taxa in each combination of season and habitat, as well as each site. Our results show no major difference in the most abundant genera identified during the dry, winter season (Table 2). However, considering the stratification level of the forest, the dominant taxa at ground level was completely different from the dominant species collected at tree level, independent of the season (Table 2). Hg. (Hag.) janthinomys, Sa. (Sbo.) chloropterus, and Wyeomyia (Phoniomyia) spp. were the dominant culicids at tree level, while Wyeomyia spp., Anopheles (Ano.) spp., Li. Pseudomethysticus, and Limatus spp. were dominant at ground level (Table 2). For all sites, seasons, and habitat types, the abundance distribution curves (Figure 4) show the classical ecological patterns where few taxa are Figure 2. Annual variation in mean temperature and total very common, while most of the taxa in natural assemblages precipitation by month for each sample site. White circles are rare (Melo 2008). represent values that were higher than the mean for each The genus Aedes represented 0.2% of individuals at site. Grey circles represent values that were at least one all sites together. Aedes serratus was the only Aedes species standard deviation higher than the mean for each site. Season present at all sites, except at Santa Rita farm, while Ae. specification determined by this data. Summer is defined as albopictus and Ae. fluviatilis were collected only at the November through April. Winter is defined as May through Bonfim and the Colonia de Una farms, respectively. TheAe. October. scapularis was captured on Almada, Santa Rita Farm, and Lagoa Encantada, and the Ae. fulvus were captured only at Almada and Santa Rita Farms. The taxaAnopheles spp. was from the full pool of samples. Then, the average diversity is present at all sites, while Hg. (Hag.) janthinomys occurred at calculated for the all samples of N individuals. Finally, average six sites (except Santa Rita Farm) and the genus Sabethes spp. expected diversity is plotted against an increasing value of and Culex spp. were present at five sites (Table 3). During the N (Melo 2008, Distler et al. 2009). Rarefaction curves were summer at ground level, the Hg. (Hag.) janthinomys was one calculated to describe the observed taxonomic richness for of the most abundant species collected in Lagoa Encantada all Culicidae in the dataset, as well as separately for each site, and REBIO-UNA (Table 2). and each habitat by season combination. We used R (R Core Ten taxa were exclusive to one of the more pristine team) to build the rarefaction curves. areas: REBIO-Una, Ecoparque and Lagoa Encantada. Cq. (Rhy.) nigricans, Cq (Rhy.) spp., and Sa. (Sab.) cyaneus were RESULTS common for all three sites. An. (Nys.) triannulatus, Ma. (Man) spp., and Sa (Sab.) spp. were unique to REBIO and the Ur. A total of 7,699 individuals comprised of 49 taxa was (Ura) calosamata, Ur (Ura) spp., Sa (Sab.) belisariori, and Ur. collected, all within the families Culicidae, Psychodidae, and (Ura) geometrica were only found at Ecoparque (Table 3). Ceratopogonidae (Figure 3a). These three insect families When using rarefaction curves to compare taxonomic occurred at the same time only in the most pristine site, richness, we found that the ground level assemblages during REBIO-Una (Site 6; Figure 3d). Culicidae was the dominant the summer had higher richness than any other habitat/ family, being the most abundant at all sites, strata (ground season combination (Figure 5b; Table 2). This was followed in level and tree level) (Figure 3b), and seasons (winter and diversity by the winter ground level group, the summer tree summer) (Figure 3c and d). level group and the winter tree level group (Figure 5b). Within the Culicidae, 46 taxonomic units were identified In across-site comparisons during the summer, Santa from the 7,534 culicids collected (Table 2). These were Rita Farm clearly had the lowest taxonomic richness (Figure distributed among five tribes (Aedini, Cullicini, Mansoniini, 5c and Table 2), while Almada Farm had the highest number Vol. 43, no. 2 Journal of Vector Ecology 297

Figure 3. Relative abundance of mosquito families. (A) all data combined, (B) each combination of habitat and season, (C) ground-level summer data for each site, and (D) ground-level winter data for each site.

Figure 4. Abundance distributions for Culicidae taxa. (A) all Culicidae data combined, (B) each combination of habitat and season, (C) ground-level data for each site. 298 Journal of Vector Ecology December 2018

Table 2. Most abundant culicids among study sites, habitat, and season.

Total Culicidae Count of Richness Richness Site Season Habitat Three Most Abundant Culicidae Taxa (N) (N) Taxa (n=47) (n=200) Li. pseudomethysticus (39.8%), All All All 7,534 5,734 45 11.84 20.73 Wyeomyia sp. (20.1%), Wyeomyia (Pho.) sp. (6.7%) Wyeomyia sp. (28.4%), An. (Ano.) sp. Summer Ground 3,308 1,907 33 13.01 21.04 (15.8%), Li. pseudomethysticus (15.8%) Li. pseudomethysticus (55.5%), Winter Ground 3,910 3,561 30 8.06 14.07 Wyeomyia sp. (16.5%), Limatus sp. (9.8%) All Hg. (Hag.) janthinomys (33.3%), Sa. Summer Tree 65 60 6 5.77 NA (Sbo.) chloropterus (31.7%), Wyeomyia (Pho.) sp. (16.7%) Wyeomyia (Pho.) sp. (41.7%), Hg. Winter Tree 251 206 17 9.97 16.85 (Hag.) janthinomys (20.4%), Sa. (Sbo.) chloropterus (13.6%) Wyeomyia sp. (42.5%), Li. Summer Ground 1,155 1,145 25 10.52 17.31 pseudomethysticus (23.4%), Wyeomyia Fazenda (Pho.) sp. (6%) Almada (Site 1) Hg. (Hag.) janthinomys (33.3%), Sa. Summer Tree 65 60 6 5.77 NA (Sbo.) chloropterus (31.7%), Wyeomyia (Pho.) sp. (16.7%) Fazenda Wyeomyia sp. (29.5%), Li. Bomfim (Site Summer Ground 139 139 13 8.97 NA pseudomethysticus (22.3%), Li. durhamii 2) (21.6%) Fazenda An. (Ano.) sp. (72.9%), An. (Ano.) Santa Rita Summer Ground 454 406 10 4.72 7.83 mediopunctatus (16.5%), Li. durhamii (Site 3) (5.9%) Wyeomyia sp. (46.7%), Limatus sp. Summer Ground 15 15 4 NA NA (40%), Ps. (Gra.) cingulata (6.7%) Wyeomyia sp. (36.6%), Limatus sp. Colônia de Winter Ground 1,152 1,072 16 7.26 10.53 (32.6%), Li. durhamii (9.8%) Una (Site 4) Wyeomyia (Pho.) sp. (31.8%), Wyeomyia Winter Tree 22 22 7 NA NA sp. (22.7%), Sa. (Sbo.) chloropterus (18.2%) Cq. (Rhy.) nigricans (28.8%), Ps. (Jan.) Summer Ground 163 163 14 9.35 NA ferox (25.2%), Hg. (Hag.) janthinomys Lagoa (23.3%) Encantada (Site 5) Li. pseudomethysticus (34.4%), Winter Ground 128 128 11 7.33 NA Wyeomyia sp. (32%), Wyeomyia (Pho.) sp. (17.2%) Hg. (Hag.) janthinomys (61.5%), An. Summer Ground 1,382 39 4 NA NA (Nys.) triannulatus (20.5%), An. (Ano.) mediopunctatus (10.3%) Li. pseudomethysticus (61.9%), REBIO-Una Winter Ground 562 417 14 7.22 12.09 Wyeomyia (Pho.) sp. (12.7%), Wyeomyia (Site 6) sp. (10.8%) Wyeomyia (Pho.) sp. (58.8%), Hg. Winter Tree 142 97 7 5.32 NA (Hag.) janthinomys (20.6%), Sa. (Sbo.) chloropterus (14.4%) Li. pseudomethysticus (86.2%), Winter Ground 2,068 1,944 19 5.01 9.57 Wyeomyia sp. (5.6%), Wyeomyia (Pho.) Ecoparque de sp. (1.7%) Una (Site 7) Wyeomyia (Pho.) sp. (25.3%), Hg. Winter Tree 87 87 15 12.11 NA (Hag.) janthinomys (23%), Sa. (Sbo.) chloropterus (11.5%) Vol. 43, no. 2 Journal of Vector Ecology 299

Table 3. Culicid mosquitoes collected during 2009-2014 in Bahia, Brazil.

Tribe Genus SubGenus Species fulvus Ochlerotatus scapularis Aedes serratus Georgecraigius fluviatilis Aedini Stegomyia albopictus Haemagogus Haemagogus janthinomys Grabhamia cingulata Psorophora albipes Janthinosoma ferox Carrollia sp. Culicini Culex Culex sp. Melanoconion sp. nigricans Coquillettidia Rhynchotaenia sp. Mansoniini venezuelensis sp. Mansonia Mansonia titillans longipes Johnbelkinia sp. durhamii Limatus pseudomethysticus sp. Runchomyia Runchomyia sp. albiprivus belisarioi cyaneus Sabethini Sabethes quasicyaneus Sabethes sp. tarsopus Sabethinus sp. Sabethoides chloropterus Peytonulus soperi Trichoprosopon digitatum Phoniomyia sp. Wyeomyia sp. calosomata Uranotaeniini Uranotaenia Uranotaenia geometrica sp. intermedius Anopheles mediopunctatus sp. Kerteszia sp. Anopheles albitarsis Nyssorhynchus sp. triannulatus Stethomyia nimbus 300 Journal of Vector Ecology December 2018

Figure 5. Rarefaction curves comparing Culicidae richness across varying sample sizes. (A) all Culicidae data combined, (B) each combination of habitat and season, (C) ground-level summer data for each site, and (D) ground-level winter data for each site. Lines around each point represent one standard error away from the mean estimate of richness. of taxonomic units. These differences among sites almost humid habitats and dense vegetation, they are also capable completely disappear during the winter (Figure 5d). However, of tolerating areas with less humidity and vegetation cover richness in winter was lower when compared with the same than Ceratopogonidae (Müller et al. 2011, Sawalha et al. sites during the summer (Figures 5c and 5d). 2017). Given that individuals from the Psychodidae family are present in four of the seven surveyed sites, further studies DISCUSSION should investigate the occurrence of leishmaniasis and arboviruses in insects and humans that live near those sites to Abundance of hematophagous insects and the importance evaluate disease transmission and exposure risks. of culicids for public health Almost 10% of the Brazilian culicid taxa are distributed As expected, the three families of insects (Psychodidae, in southern Bahia, with Sabethini as the most dominant Ceratopogonidae, and Culicidae) were present in the tribe. This tribe is almost entirely Neotropical (Cardoso et al. most pristine study site given that better conserved areas 2010, Gomes et al. 2010, Navarro et al. 2015) and most taxa will support increased insect richness due to the variety of seem to be zoophilic but will opportunistically bite humans resources, hosts, breeding sites, and microhabitats ( Taipe- both in the forested and non-forested areas. However, these Lagos and Natal 2003, Pinto et al. 2009, Nnko et al. 2017, mosquitoes are found more commonly in areas of old-growth Sawalha et al. 2017). Psychodidae, the sandflies family, are or secondary-growth forest with little human alteration vectors for leishmaniasis (Dias-Lima et al. 2003, Sawalha et and may be considered environmental health bioindicators al. 2017) and more than 21 species of arboviruses, including (Paula et al. 2015). Culicids should be considered one of the Phlebovirus . Ceratopogonidae is similarly known to transmit most important emerging and re-emerging disease vectors some arboviruses, including the Oropouche virus (Mellor et (Navarro et al. 2015) to be investigated because of their high al. 2000). One of the reasons to explain the lower abundance abundances and ongoing arbovirus outbreaks in humans and of the Ceratopogonidae in cacao agroforestry systems is wild animals (Vasconcelos and Calisher 2016). their inability to breed in cacao peels, the main substrate left by the farmers after extracting seeds from the cacao fruit. Even though many species in the Psychodidae family prefer Vol. 43, no. 2 Journal of Vector Ecology 301

Evidence of forest stratification in the composition of the Although we did not study the demography of mosquito most common Culicidae mosquitoes species, we compared the richness and abundance of taxa Vertical stratification, a marked preference for different across sites with different management histories. The richness heights within the vegetation, was expected for the Sabethini of culicids and the high abundance of the Sabethini tribe tribe. This is so because most taxa in this tribe breed in suggests that sites with both fragments of Atlantic Forest tree holes, bromeliads, and palm axils (Forattini 1995). and cacao agroforestry systems (capable of maintaining However, as found in other studies, some Sabethine , such characteristics of sylvatic environments) support biodiversity as Li. pseudomethysticus and Wyeomyia spp., prefer to use better than highly disturbed environments. Despite the fact the ground level of the forest, depositing eggs in natural that Sabethini tribe mosquitoes are found in areas of old- substrates (e.g., bamboo, bromeliads, and fruit peel) or growth or secondary-growth forest with no or little human in artificial receptacles left by humans on the ground alteration (Paula et al. 2015), some species from this tribe, (Guimarães et al. 1985). This breeding behavior could explain such as Li. durhami, are associated with areas inhabited their higher abundance at the ground level at the sites studied. by humans because eggs and larvae of this species are Arboviruses have been isolated for these two taxa, including frequently found in artificial breeding sites such as plastic Eastern equine encephalitis virus; however, little information bottles (Guimarães and Arlé 1984). Our study corroborates about their role as arbovirus vectors has been reported. the findings by both Guimarães and Victório (1986) and The high occurrence of Hg. (Hag.) janthinomys, Sa. Guimaraes and Arlé (1984) who report Limatus species (Sbo.) chloropterus, and Wyeomyia (Pho.) spp. at tree level present in sylvatic environments and human modified was expected since they show affinity for forest canopies environments, respectively. (species termed “acrodendrophilic”). These taxa lay eggs Furthermore, the exclusivity of some culicids species almost exclusively in bromeliads (Guimarães et al. 1985, in the most pristine forest fragments, the lower relative Forattini 1995, Pinto et al. 2009, Gomes et al. 2010), which abundance of the genus Aedes in these sites, and the are more frequently located high on the trees (Gomes et al. presence of anophelines in all sites corroborate the idea 2010, Guimarães and Arlé 1984). Furthermore, Hg. (Hag.) that fragments of Bahia Atlantic Forest close to managed janthinomys and Sa. (Sbo.) chloropterus are primatophilic, agroforestry systems support biodiversity better than highly i.e., most often obtain blood meals from monkeys, which disturbed environments. For example, Ae. serratus, found in typically live at tree level (Guimarães et al. 1985). Previous almost every study site, has been preferentially captured in studies showed that the circadian activity of mosquito sylvatic environments in other studies (Forattini 1995). In vectors may be related to the resting periods of their hosts natural conditions, this species can be infected with STLV, (Galindo 1958). Guimarães and Victório (1986), and Pinto et Oropouche, Aura, and Trocara virus in Amazon forests. It al. (2009) observed a more aggressive behavior of Hg. (Hag.) is also considered the vector responsible for the sylvatic janthinomys and Sa. (Sbo.) chloropterus females towards their cycle of ILHV in other regions of Brazil (Cardoso et al. hosts between the late morning hours and the early afternoon, 2010). Additionally, Ae. scapularis is commonly captured a period of the day that corresponds to resting time for in environments modified by humans, while reports on Ae. monkeys, the preferred hosts for these species. However, as fulvus occurrence report them in forested areas with reduced found in other studies, these mosquitoes are also capable of human activity (Forattini 1995), as we found in the present using the forest floor, which was observed at the REBIO-Una study. In Brazil, anophelines are generally common in rural and Lagoa Encantada sites (Guimarães et al. 1985, Gomes et and wild environments, but they are rare in urban areas. This al. 2007, 2010, Pinto et al. 2009). Hg. (Hag.) janthinomys are an is likely due to species requirements for special environmental important vector of yellow-fever and Mayaro virus in sylvatic conditions, such as bromeliads as breeding sites (Pittendrigh environments (Davis 1930, Thoisy et al. 2003, Almeida et 1950, Guimarães et al. 1985, Cardoso et al. 2012, Paula et al. 2014), while species of Sabethes are considered potential al. 2015). The relationship between anophelines and the secondary vectors of yellow fever virus in Central and South presence of favorable human-made habitats or habitats that America along with other arbovirus such as SLEV and ILHV keep bromeliads and water storage containers may contribute (Forattini 1995). As described by Gomes et al. (2010), species to the maintenance of the population of these species in all from the genus Sabethes were present in almost all of the same study sites. sites as Haemagogus sp. and Sabethes chloropterus. The high abundance of large bromeliads in the region (Catenacci et al. 2009, Fontoura et al. 2010), which serve as Culicidae diversity in natural and anthropogenically breeding sites, the tall trees that make up the canopies, and modified habitats the abundance of wild fauna that still populate these areas Culicids are of interest not only from an epidemiological (Cassano et al. 2009) all likely contribute to the biodiversity of perspective, an aspect that has been amply discussed in vectors in the study sites. According to Nnko et al. (2017), the the literature, but also because some taxa may be used as distribution and abundance of vectors is determined by the bioindicators of environmental health. The usefulness of interplay of three factors: suitable climatic conditions, habitat mosquitoes as bioindicators is readily apparent if one assesses for development, and the availability of hosts for food. the extent of change that has occurred in a particular region Even though forest fragments and agroforestry systems and, at the same time, observes the increase or decrease in southern Bahia are able to maintain high diversity of or disappearance of a particular taxon (Paula et al. 2015). mosquitoes and the complex interactions among vectors- 302 Journal of Vector Ecology December 2018 pathogens-wild hosts in nature, continued deforestation have both the competency and capacity to transmit important and an increase in the extent of cattle ranching has resulted pathogens to wild animals, and many of them have not been in drastic changes in the landscape matrix. These changes studied. Haemagogus janthynomys, and others, have shown can directly impact vector populations, which can lead to a a capacity for living in rural and sylvatic environments, as high prevalence of certain species with a higher capacity for have species in the Sabethini tribe, increasing contact with arbovirus transmission resulting in outbreaks of arboviruses humans and the risk of emerging arbovirus infections. The both in people and wild animals, as has been reported in limited literature on entomofauna in the state of Bahia, the other regions of Brazil (Lima et al. 2010, Pauvolid-Corrêa fact that some of the hematophagous arthropods identified et al. 2010). Because of this, carrying out an entomological in this study are considered significant vectors, and the investigation on the species composition of Culicidae considerable human and wildlife interactions in the region assemblages in sylvatic and rural areas can help highlight warrant continued studies on the diversity and abundance of the importance of keeping natural areas in an anthropic hematophagous insects in other areas of the state. environment matrix by showing the biological richness of these arthropods (Paula et al. 2015). Acknowledgments

Seasonality affects mosquito diversity We are much indebted to the technical staff of the Previous studies have suggested that the structural Department of the Health of Bahia State (6a DIRES) who complexity of an environment has an important effect on helped the authors to collect the material in the field and the the insect diversity and composition (Pittendrigh 1950, Evandro Chagas Institute team, especially Bruna Laís Sena Pinto et al. 2009, Braga et al. 2010). The high abundance of do Nascimento, Hélio Augusto Cardoso Saraiva, José Wilson large bromeliads in the study region (Catenacci et al. 2009, Rosa Junior, Francisco Correa de Castro, Durval Bertran Fontoura et al. 2010), which serve as breeding sites, the tall Rodrigues Vieira, and Hamilton Monteiro for assistance with trees that make up the canopies, the abundance of wild fauna the identification of the arthropods. Additional thanks go to that still populate these areas (Cassano et al. 2009) likely also Kristel M. de Vleeschouwer from the Project BioBrasil/Centre contribute to the observed biodiversity of vectors. However, for Research and Conservation Antwerpia Zoo, ICMBIO, even for sites that contain the same key characteristics, Kathleen Apakupakul at the Saint Louis Zoo Institute for small variations in environmental conditions can affect the Conservation Medicine, and Leonardo Oliveira from the diversity of insects (Pittendrigh 1950, Pinto et al. 2009). As Bicho do Mato Instituto de Pesquisa for their logistical support. described in other studies, the high humidity and temperature We sincerely acknowledge the owner of the areas. This study of summer allows for an increase the relative abundance of was funded by Saint Louis Zoo WildCare Institute (U.S.A.), insects because these conditions promote the occurrence of The American Association of Zoological Veterinarians Wild breeding sites ( Pittendrigh 1950, Paterno and Marcondes Fund (AAZV, U.S.A.), CNPq (Brasil), the Centre for 2004, Pinto et al. 2009, Sawalha et al. 2017). Usually, in the Research and Conservation of the Royal Zoological Society of wet tropics, epizootics and epidemics of mosquito-borne Antwerp (Belgium), Lion Tamarins of Brazil Fund, National diseases are associated with the onset of the rainy season, Lottery of Belgium, Primate Action Fund, Zoological Society when the population densities of vectors are higher. Pinto et of London. The Flemish Ministry of Science (Belgium) al. (2009) also observed in a tropical forest that the number provided structural support to the Centre for Research and of culicid mosquitoes was higher during the rainy season Conservation of the Royal Zoological Society of Antwerp. and lower at the start of the dry season, both at canopy and ground level. Additionally, a wide range of mosquitoes have REFERENCES CITED the capacity to maintain their eggs in a quiescent stage under poor conditions until the environmental situation improves, Alger, K. and M. Caldas. 1994. The declining cocoa which usually coincides with the summer/rainy season. The economy and the Atlantic Forest of Southern Bahia, importance of habitat and season to explain the diversity Brazil: Conservation attitudes of cocoa planters. of entomofauna might increase if we consider microhabitat Environmentalist 14: 107–119. variations among sites that could influence the development Almeida, M.A.B., J.C. Cardoso, E.dos Santos, D.F da Fonseca, of immature insect stages, such as water salinity, water L.L. Cruz, F.J.C. Faraco, M.A. Bercini, K.C. Vettorello, movement, water pollutants, etc. (Pittendrigh 1950). These M.A. Porto, R. Mohrdieck, T.M.S. Ranieri, M.T. complexities could help explain the difference in richness Schermann, A.F. Sperb, F.Z. Paz, Z.M.A. Nunes, A.P.M. between seasons, habitat, and sites in the present study, even Romano, Z.G. Costa, S.L. Gomes, and B. Flannery. for sites with similar vegetation. 2014. Surveillance for yellow fever virus in non-human In summary, our study suggests that fragmented forest primates in southern Brazil, 2001–2011: a tool for areas, followed by agroforestry systems in southern Bahia prioritizing human populations for vaccination. PLoS host high sylvatic mosquito diversity and low abundance of Negl. Trop. Dis. 8: 1-7. the main genus of culicid that is important for public health, Braga, D.L., J.N.C. Louzada, R.Zanetti, and J. Delabie. 2010. Aedes spp. However, the importance of native fauna that is Rapid evaluation of ant diversity in land use systems in established in forested and managed areas should not be southern Bahia, Brazil. Neotrop. Entomol. 39: 464–469. underestimated, because various taxa in these areas already Brazil. Ministério da Saúde. 2016. Carta anofélica da Bahia, Vol. 43, no. 2 Journal of Vector Ecology 303

Brasil (Diptera: Culicidae). Bol. Entomológico Secr. Rev. Soc. Bras. Med. Trop. 47: 677–683. Saúde Estado Bahia 3: 1–2. Figueiredo, L.T.M. and L.T.M. Figueiredo. 2015. The recent Cardoso, C.W., M. Kikuti, A.P.P.B. Prates, I.A.D. Paploski, arbovirus disease epidemic in Brazil. Rev. Soc. Bras. L.B. Tauro, M.M.O. Silva, P. Santana,. M.F.S. Rego, M.G. Med. Trop. 48: 233–234. Reis, U. Kitron, and G.S. Ribeiro. 2017. Unrecognized Figueiredo, M.L.G, A.C. Gomes, A.A. Amarilla, A.S. Leandro, emergence of chikungunya virus during a Zika virus R.F. Orrico, S.M. Araujo, J. Castro, E.L. Durigon, V.H. outbreak in Salvador, Brazil. PLoS Negl. Trop. Dis. 11: Aquino, and L.T.M. Figueiredo. 2010. Mosquitoes 1-8. infected with dengue viruses in Brazil. Virol. J. 7: 152. Cardoso, J. da C. M.B. Paula, A. Fernandes, E. Santos, M.A.B. Fontoura, T., E. Cazetta, W. Nascimento, L. Catenacci, K.M. Almeida, D.F. Fonseca, and M.A.M. Sallum. 2010. New De Vleeschouwer, and B. Raboy. 2010. Diurnal frugivores records and epidemiological potential of certain species on the Bromeliaceae Aechmea depressa L.B. Sm. from of mosquito (Diptera, Culicidae) in the State of Rio Northeastern Brazil: the prominent role taken by a small Grande do Sul, Brazil. Rev. Soc. Bras. Med. Trop. 43: forest primate. Biota Neotropica 10: 351–354. 552–556. Forattini, O.P. 1995. Principais de importância Cardoso, J.D.C., E.S. Bergo, T.M.P. Oliveira, D.C. Sant’ana, sanitária no Brasil. Cad. Saúde Pública 11: 157–158. M.T. Motoki, and M.A.M. Sallum. 2012. New records Galindo, P. 1958. Bionomics of Sabethes chloropterus of Anopheles homunculus in central and Serra Do Mar Humboldt, a vector of sylvan yellow fever in Middle biodiversity corridors of the Atlantic Forest, Brazil. J. America. Am. J. Trop. Med. Hyg. 7: 429–440. Am. Mosq. Contr. Assoc. 28: 1–5. Gomes, A.C., M.A.N. Torres, L. Ferri, F.R. Costa, and A.M. Carneiro-Santos, H.L., L.A. F. Martins, R.H.S. Peralta, J.M. Silva. 2007. Finding of Haemagogus (Conopostegus) Peralta, and H. Werneck de Macedo. 2014. Frequency leucocelaenus (Diptera: Culicidae), in the municipality of of amoebiasis and other intestinal parasitoses in a Porto Alegre, State of Rio Grande do Sul. Rev. Soc. Bras. settlement in Ilhéus City, State of Bahia, Brazil. Rev. Soc. Med. Trop. 40: 487–488. Bras. Med. Trop. 47: 101–104. Gomes, A.C., M.A.N. Torres, M.B. Paula, A. Fernandes, A.M. Cassano, C.R., G. Schroth, D. Faria, J.H.C. Delabie, and L. Marassá, C.A. Consales, and D.F. Fonseca. 2010. Ecologia Bede. 2009. Landscape and farm scale management de Haemagogus e Sabethes (Diptera: Culicidae) em áreas to enhance biodiversity conservation in the cocoa epizoóticas do vírus da febre amarela, Rio Grande do Sul, producing region of southern Bahia, Brazil. Biodivers. Brasil. Epidemiol. Serviços Saúde 19: 101–113. Conserv. 18: 577–603. Guimarães, A.E. and M. Arlé. 1984. Mosquitos no Parque Cassano, C.R., M.C.M. Kierulff, and A.G. Chiarello. 2011. Nacional da Serra dos Órgãos, estado do Rio de Janeiro, The cacao agroforests of the Brazilian Atlantic forest Brasil: I- distribuição estacional. Mem. Inst. Oswaldo as habitat for the endangered maned sloth Bradypus Cruz 79: 309–323. torquatus. Mamm. Biol. 76: 243–250. Guimarães, A.E., M. Arlé, and R.N.M. Machado. 1985. Catenacci, L.S., K.M. De Vleeschouwer, and S.L.G. Nogueira- Mosquitoes in the National Park of Serra dos Orgãos, Filho. 2009. Seed dispersal by golden-headed lion State of Rio de Janeiro, Brazil: II. Vertical distribution. tamarins Leontopithecus chrysomelas in southern Bahian Mem. Inst. Oswaldo Cruz 80: 171–185. Atlantic Forest, Brazil. Biotropica 41: 744–750. Guimarães, A.E. and V.M.N. Victório. 1986. Mosquitoes in Catenacci, L.S., M.S. Pessoa, S.L.G. Nogueira-Filho, and the Parque Nacional da Serra dos Orgãos, State of Rio de K.M.D. Vleeschouwer. 2016. Diet and feeding behavior of Janeiro, Brazil: III. Hour preference for blood sucking. Leontopithecus chrysomelas (Callitrichidae) in degraded Mem. Inst. Oswaldo Cruz 81: 93–103. areas of the Atlantic Forest of South-Bahia, Brazil. Int. J. Harbach, R.E. 2017. Mosquito Taxonomic Inventory. Valid Primatol. 37: 136–157. Species List. Available on: http://mosquito-taxonomic- Consoli, R.A.G.B. and R.L. Oliveira. 1994. Principais inventory.info/valid-species-list mosquitos de importância sanitária no Brasil, Editora Lima, M.A., N.S. Romano-Lieber, and A.M.R.C. Duarte. Fiocurz, RJ. 228 pp. 2010. Circulation of antibodies against yellow fever virus Davis, N.C. 1930. The susceptibility of marmosets to yellow in a simian population in the area of Porto Primavera fever virus. J. Exp. Med. 52: 405–416. Hydroelectric Plant, São Paulo, Brazil. Rev. Inst. Med. Dias-Lima, A.G., M.L.S. Guedes, and I.A. Sherlock. 2003. Trop. São Paulo 52: 11–16. Horizontal stratification of the sand fauna (Diptera: Medeiros-Sousa, A.R., W. Ceretti-Junior, P.R. Urbinatti, D. Psychodidae) in a transitional vegetation between Natal, G.C. Carvalho, M.B. Paula, A. Fernandes, M.H.S.H. caatinga and tropical rain forest, state of Bahia, Brazil. Mello, R.C. Oliveira, L.D. Orico, E.F.B. Gonçalves, and Mem. Inst. Oswaldo Cruz 98: 733–737. M.T. Marrelli. 2013. Mosquito (Diptera: Culicidae) Distler, T., P.M. Jørgensen, A. Graham, G. Davidse, and I. survey in parks of São Paulo City I. Biota Neotropica 13: Jiménez. 2009. Determinants and prediction of broad- 317–321. scale plant richness across the western neotropics. Ann. Mellor, P.S., J. Boorman, and M. Baylis. 2000. Culicoides Mo. Bot. Gard. 96: 470–491. biting midges: their role as arbovirus vectors. Annu. Rev. Figueiredo, M.L.G. and L.T.M. Figueiredo. 2014. Emerging Entomol. 45: 307–340. alphaviruses in the Americas: Chikungunya and Mayaro. Melo, A.S. 2008. O que ganhamos “confundindo” riqueza de 304 Journal of Vector Ecology December 2018

espécies e equabilidade em um índice de diversidade? the Caxiuanã National Forest, Pará, Brazil. Mem. Inst. Biota Neotropica 8: 21-27. Oswaldo Cruz 104: 592–598. Montes, J. 2005. Culicidae fauna of Serra da Cantareira, Sao Pittendrigh, C.S. 1950. The ecoclimatic divergence of Paulo, Brazil. Rev. Saúde Pública 39: 578–584. Anopheles bellator and A. homunculus. Evolution 4: 43– Müller, G.C., V.D. Kravchenko, L. Rybalov, and Y. Schlein. 63. 2011. Characteristics of resting and breeding habitats of Ribeiro, M.C., J.P. Metzger, A.C. Martensen, F.J. Ponzoni, and adult sand flies in the Judean Desert. J. Vector Ecol. 36: M.M. Hirota. 2009. The Brazilian Atlantic Forest: How 195–205. much is left, and how is the remaining forest distributed? Navarro, S., P.D. Enríquez, Y. Campaña, and W. Benítez- Implications for conservation. Biol. Conserv. 142: 1141– Ortiz. 2015. New Sabethes (Diptera: Culicidae) species 1153. records for Ecuador, from Colonso-Chalupas biological Sawalha, S.S., A. Ramlawi, R.M. Sansur, I.M. Salem, and reserve, province of Napo (Amazon). J. Entomol. Zool. Z.S. Amr. 2017. Diversity, ecology, and seasonality of Stud. 3: 169–172. sand flies (Diptera: Psychodidae) of the Jenin District Nnko, H.J., A. Ngonyoka, L. Salekwa, A.B. Estes, P.J. Hudson, (Palestinian Territories). J. Vector Ecol. 42: 120–129. P.S. Gwakisa, and I.M. Cattadori. 2017. Seasonal Schmidt, L.S. and A. Barcellos. 2007. Abundance and species variation of tsetse fly species abundance and prevalence richness of Heteroptera (Hemiptera) from Parque of trypanosomes in the Maasai Steppe, Tanzania. J. Estadual do Turvo, southern Brazil: Pentatomoidea. Vector Ecol. 42: 24–33. Iheringia Sér. Zool. 97: 73–79. Oliveira, L.C., L.G. Neves, B.E. Raboy, and J.M. Dietz. 2011. Serra, O.P., B.F. Cardoso, A.L.M. Ribeiro, F.A.L. Santos, and Abundance of jackfruit (Artocarpus heterophyllus) affects R.D. Slhessarenko. 2016. Mayaro virus and Dengue virus group characteristics and use of space by golden-headed 1 and 4 natural infection in culicids from Cuiabá, state lion tamarins (Leontopithecus chrysomelas) in Cabruca of Mato Grosso, Brazil. Mem. Inst. Oswaldo Cruz 111: Agroforest. Environ. Manag. 48: 248–262. 20–29. Paterno, U. and C.B. Marcondes. 2004. Mosquitoes with Sollberg, I., A. Schiavetti, and M.E.B. Moraes. 2014. morning biting activity in the Atlantic forest, Brazil Agricultural management in the Una’s wildlife refuge: a (Diptera, Culicidae). Rev. Saúde Públ. 38: 133–135. perspective of conservation by agroforestry. Rev. Árvore Paula, M.B., A. Fernandes, A.R. Medeiros-Sousa, W. Ceretti- 38: 241–250. Júnior, R. Christe, R.C. Stroebel, L. Pedrosa, R.M.M.S. Souza, R.L., R.F. Santos, J. Nunes Neto, H.A. Monteiro, and Almeida, G.C. Carvalho, U.D. Pereira, M.C.O. Jacintho, P.F.Vasconcelos. 2015. Monitoramento Entomológico em D. Natal, and M.T. Marreli et al. 2015. Mosquito (Diptera: área de ocorrência de febre amarela silvestre no oeste da Culicidae) fauna in parks in greater São Paulo, Brazil. Bahia. Saúde Públ. 39: 136-140. Biota Neotropica 15: 1-9. Taipe-Lagos, C.B. and D. Natal. 2003. Culicidae mosquito Pauvolid-Corrêa, A., F.N. Tavares, J. Alencar, J.S. Silva, M. abundance in a preserved metropolitan area and its Murta, N.M. Serra-Freire, A.O. Pellegrin, H. Gil-Santana, epidemiological implications. Rev. Saúde Públ. 37: 275– A.E. Guimarães, and E.E. Silva. 2010. Preliminary 279. investigation of Culicidae species in South Pantanal, Thoisy, B., J. Gardon, R.A. Salas, J. Morvan, and M. Kazanji. Brazil and their potential importance in arbovirus 2003. Mayaro virus in wild mammals, . transmission. Rev. Inst. Med. Trop. São Paulo 52: 17–24. Emerg. Infect. Dis. 9: 1326–1329. Pinto, C.S., U.E. Confalonieri, and B.M. Mascarenhas. Vasconcelos, P.F.C. and C.H. Calisher. 2016. Emergence of 2009. Ecology of Haemagogus sp. and Sabethes sp. human arboviral diseases in the Americas, 2000-2016. (Diptera: Culicidae) in relation to the microclimates of Vector Borne Zoonot. Dis. 16: 295–301.