Ectoparasitic flies (Diptera: Streblidae) on bats (Mammalia: Chiroptera) from a Private Natural Heritage Reserve in southeastern Brazil

Due to the small number of records of Streblidae on bats, despite extensive study on these mammals in the state of Rio de Janeiro, a survey was carried out in an area of the Atlantic Forest. The present study was carried out at Bom Retiro Farm Natural Heritage Private Reserve. We captured 401 bats of 17 species, 13 genera, and four families; bat flies infested 221 bats of only four species. Carollia perspicillata (Linnaeus, 1758) has the highest fly diversity, with seven fly species: Trichobius joblingi (Wenzel, 1966) (n = 23), Megistopoda proxima (Séguy, 1926) (n = 15), Strebla guajiro (García & Casal, 1965) (n = 15), Aspidoptera falcata (Wenzel, 1976) (n = 6) Paratrichobius longicrus (Miranda Ribeiro, 1907) (n = 8), Paraeuctenodes similis (Wenzel 1976) (n = 3), and Trichobius anducei (Guerrero, 1998) (n = 1). Two species infested Platyrrhinus lineatus (É. Geoffroy, 1810): Aspidoptera falcata (n  =  1) and Anastrebla caudiferae (Wenzel, 1996) (n  =  1). Paradyschiria parvula (Falcoz, 1931) (n = 11) infested Noctilio leporinus (Linnaeus, 1758) and M. proxima (n = 12) and Trichobius uniformis (Curran, 1935) (n = 1) infested Sturnira lilium (É.  Geoffroy, 1842). Sturnira lilium has the highest infestation rate, with ten out of 46 captured individuals parasitized, followed by Carollia perspicillata, with 33 out of 164 captured parasitized, and by P. lineatus with only two parasitized individuals out of ten. Among 97 streblid flies captured, M. proxima was the most abundant (27.83%), followed by T. joblingi (23.71%), and S. guajiro (15.46%). All remaining bat fly species represented 33%. Paradichyria parvula has the first record for Rio de Janeiro State. Key-Words. Flying mammals; Parasitism; Megistopoda proxima; Atlantic Forest.


INTRODUCTION
Bats are the group of mammals with the second highest species diversity in the Neotropics (Findley, 1993;Nowak, 1994;Voss & Emmons, 1996;Burgin et al., 2018), only behind rodents. It has 1.386 species world wide (Burgin et al., 2018) and 180 recorded in Brazil (Reis et al., 2017), with seven species included on Brazilian Fauna Red List (ICMBIo/MMA, 2018). Bats maintain parasitic relationships, such as endoparasitism and ectoparasitism. In endoparasitism, the relationships occur with protozoa, helminths, flatworms, and nematodes (Santos & Gibson, 2015). In the case of ectoparasitism, somearthropods use bats as a means of transport and feed on their hair follicles (Kunz, 1982). Despite the importance of investigating the ectoparasitic insect community, infestation patterns, and the factors that regulate this infestation, these topics are still poorly studied (Rui & Graciolli, 2005).
The distribution of ectoparasites in bats seems to follow a certain proportionality. The abundance of ectoparasites could differ among hosts in the same population according to different age, sex, and reproductive categories (Marshall, 1982). Cases of ectoparasite over population in bats are rare, as they would result in significant damage to the host, comprising from minor problems to severe injuries, such as blood loss, malnutrition, and skin and fur damage (Kunz, 1982). Only two families of hematophagous dipterans, Streblidae and Nycteribiidae, are known to parasitize bats (Silva & Ortêncio-Filho, 2011). Together with Glossinidae and Hippoboscidae, these families form the superfamily Hippoboscoid within the Calyptrate, infraorder Schizophora, suborder Brachycera (Petersen et al., 2007). The relationships between the families in the superfamily are still in debate, altought two clades have been accepted: Nycteribiidae and Streblidae, and Glossinidae and Hippoboscidae (Petersen et al., 2007).
In Brazil, there are few studies on bat ectoparasites (Komeno & Linhares, 1999;Bertola et al., 2005) as well as in the state of Rio de Janeiro State (Almeida et al., 2010;França et al., 2013). So, it is possible and expected that new records of the relationship between bat and ectoparasites will be found in areas with few studies. Thus, the present study aims to fill in the knowledge gapabout ectoparasites found on bats captured at Bom Retiro Farm Private Natural Heritage Reserve, inserted in an Atlantic Forest fragment. We want to contribute to the knowledge of the ectofauna streblidae of the state of Rio de Janeiro, as the first record of the ectoparasite Paradichyria parvula (Falcoz, 1931) for the state.

MATERIAL AND METHODS
This study was carried out at Bom Retiro Farm Private Natural Heritage Reserve (RPPN), here after Bom Retiro Farm (22°27′S, 42°18′W), wich is located at approximately 140 km from the city of Rio de Janeiro, southeastern Brazil. It comprises the municipalcities of Casimiro de Abreu and Silva Jardim, with an area of 494.3 ha (Fig. 1), mostly covered by typical Atlantic Forest vegetation, which remains under the influence of the Tropical Atlantic Mass originated from the Tropical Atlantic Antictone for most of the year. The studied area is with in São João River Basin, in an intertropical zone (low latitudes), with intense solar radiation, under high influence of the Atlantic Ocean, with average annual temperatures of 18° to 24° (Cunha, 1995), and a predominantly humid tropical climate (Takizawa, 1995). Private Natural Heritage Reserves (RPPN) are private areas chosen to be protected byan initiative of owners, upon recognition by the Government (Federal Decree #1922/1996). They are relevant due to their biodiversity, landscape aspect and/or environmental characteristics that justify recovery actions. The regulation of RPPNs began in 1990, by the request of some landowners who wanted to turn part of their real estate into private reserves that allowed touristic and research activities (IBAMA, 2005).
Nineteen field expeditions were carried out between January 2012 and May 2016. Bats were captured using mist nets measuring 9 m × 2,5 m. In each campaign, five networks that were opened for 12 hours a night were used, with an capture effort of 25.650 k.m², following Straube & Bianconi (2002). The bats which were captured were put in individual cloth bags, previously identified in the field, using the identification keys by Emmons & Feer (1997) and descriptions provided by Simmons & Voss (1998), Dias et al., (2002), and Dias & Peracchi (2008). The first specimen of each species captured and the specimens that raised doubts about identification were sacrificed and preserved in 70% alcohol. Voucher specimens were deposited in the Adriano Lúcio Peracchi Collection (ALP) of the Mastozoology Laboratory, Institute of Biology, Federal Rural University of Rio de Janeiro (UFRRJ). We declare that all actions performed with the animals have followed and respected the international rules of animals well being, through all parameters required from Animal Ethics Committee (AEC).
The bats analyzed were marked using the capture-mark-recapture method with numbered plastic collars, following Esbérard & Daemon (1999), and then released.
Flies on the body surface of bats were collected with tweezers. Captured ectoparasites were fixed in 70% alcohol and stored in Eppendorf tubes with a label identifying the bat species on which the fly was collected, as well as the site and date of collection. Dr. Gustavo Graciolli, professor at the Biology Institute, Federal University of Mato Grosso do Sul, identified streblid flies.
Associations between streblid flies and bats were analyzed by counting the number of bats captured and the number of infested bats. We also calculated the average prevalence by counting the number of ectoparasites on each bat species in relation to the number of ectoparasites collected and calculated the prevalence of ectoparasites (Margolis et al., 1982). To test for differences in the number of ectoparasites in different bat species, regardless of sex, parasite or weight, the Mann-Whitney test was applied. Species with a few number of individuals were not included in the stastical test. We used the cluster analysis to compare the present study with five other studies carried out in different locations: Distrito Federal (Graciolli & Coelho, 2001), Maranhão (Santos et al., 2009), São Paulo (Bertola et al., 2005), Minas Gerais (Komeno & Linhares, 1999), Paraná (Graciolli & Bianconi, 2007) and Rio de Janeiro II (França et al., 2013).
These ten species correspond to 97 specimens: 48 males (49.49%) and 49 females (50.51%). The male-female ratio was 1:1.02. Sex-ratio was proportional in all fly species captured, with a larger number of females in six of ten species: A. falcata, M. proxima, P. parvula, P. similis, S. guajiro, and T. joblingi. Paratrichobius longicrus showed the same number of males and females. We captured only one male of A. caudiferae and T. uniformis, and only one female of T. anducei (Table 1).

Association of flies and bats
Streblids infested four bat species: Carollia perspicillata, Noctilio leporinus, Sturnira lilium, and Platyrrhinus lineatus. We collected 38 females and 33 males on C. perspicillata, seven females and six males on S. lilium, six females and five males on N. leporinus, and one female and one male on P. lineatus.
The specie with the highest number of streblid species was C. perspicillata, with seven species. Secondly, Platyrrhinus lineatus and S. lilium, both with two species captured on each. Noctilio leporinus was parasitized only by one species. Sturnira lilium showed the highest infestation rate: ten out of 46 captured individuals showed ectoparasitic flies; followed by Carollia perspicillata with 33 out of 164 captured specimens parasitized, and P. lineatus with two out of ten individuals parasitized (Table 1). Megistopoda proxima and A. falcata were the only species on two different bat species; the former parasitized C. perspicillata and S. lilium and the latter parasitized C. perspicillata and P. lineatus.
The Mann-Whitney test used to compare the number of ectoparasites on C. perspicillata and S. lilium, regardless of sex, parasite or weight was significant (U = 70, Z = 2.4, and p = 0.01). Carollia perspicillata was more parasitized than S. lilium. Other species were not included in the analysis due to the small number of individuals captured. According to the cluster analysis, there is a similarity between Rio de Janeiro, São Paulo, Minas Gerais and Paraná (Fig. 2).
The Streblidae richness at Bom Retiro Farm did not differ from the average observed in other studies (Table 2). When compared, our results regarding the number of bats analyzed, both the number of host species (4) and the number of fly species (10) were lower than recorded in other studies (e.g., Bertola et al., 2005, carried out in São Paulo). When comparing the present study with the other six carried out in different locations (Distrito Federal, Maranhão, São Paulo, Minas Gerais, Paraná, and Rio de Janeiro II), using a cluster tree, an important similarity in streblid composition related to the distance between research islands was observed (Fig. 2). There is a similarity between the results obtained and the distance, as closer states, such as São Paulo, Minas Gerais, and Paraná, have more species in common than Distrito Federal and Maranhão, which are far from each other. The similarity in the faunal composition of streblids in four of six studies (Rio de Janeiro, São Paulo, Minas Gerais and Paraná), may be related to the fact that these works also had a large amount of capture of C. perspicillata; more over, these have also been carried out in the same ecossystem, the Atlantic Forest, with the exception of Minas Gerais, which was in the Cerrado.
The high species richness of the family Streblidae found in Maranhão can be related to the higher abundance and distribution of this group in equatorial regions (Guerrero, 1993(Guerrero, , 1994(Guerrero, , 1995aSantos et al., 2009).
Regarding the male-female ratio, the present study showed a slight dominance of females, as reported only by Santos et al. (2009). Usually, the number of male is higher, probably because they move more than females (Wenzel, 1976), which are more restricted to the roost for breeding purposes.

Association of flies and bats
A total of 45 out of 221 bats had ectoparasites, which represents an infestation rate of 20.4%. Bats of the family Phyllostomidae were the most parasitized by ectoparasites, and C. perspicillata obtained a higher number of species than that observed in studies carried out in Minas Gerais (Komeno & Linhares, 1999), São Paulo (Graciolli et al., 2006) and Rio de Janeiro II (França et al., 2013). Carollia perspicillata had seven species, a number higher than those found in the study of Bertola et al. (2005; 5 species), Santos et al. (2009;4 species), and França et al. (2013;4 species). Simmons & Voss (1998) and Carvalho et al. (2013) pointed out that the height of mist nets between 2.5 and 3 meter sallow shigh capture of phyllostomines, in particular frugivores and nectarivores, as they prefer to fly at the understory level. The high number of C. perspicillata may be related to the predominance of plants of the family Piperacea at Bom Retiro Farm, as species of this family comprises the most visited plants by C. perspicillata during feeding (Pereira et al., 2018;Muller & Reis, 1992). Among the ectoparasites found on C. perspicillata, T. joblingi was also observed in Minas Gerais (Komeno & Linhares, 1999), Maranhão (Santos et al., 2009), São Paulo (Bertola et al., 2005), Distrito Federal (Graciolli & Coelho, 2001), and Rio de Janeiro II (França et al., 2013); in all of these studies, this species was the most captured on C. perspicillata. Strebla guajiro was also observed in those studies, except for Distrito Federal (Graciolli & Coelho, 2001).
In the present study, we captured two ectoparasites of the Streblidae family on P. lineatus (A. falcata and A. caudiferae), but none of the other studies compared here reported the presence of ectoparasites on this bat species. The presence of P. parvula parasitizing N. leporinus corroborates previous findings, as this bat is considered one of the host types of this parasite (Graciolli & Carvalho, 2001).
The records of parasite-host associations and the occurrence of streblids carried out in the study represent a significant contribution to deepening the knowledge about the distribution of fly species and their relationship with bats. The few existing studies, presented here in this work, show how this group and the relationship of ectoparasitism with bats is neglected. Further more, understanding these ecological relationships can help in the management and conservation processes of these taxonomic groups, especially in environments such as fragments of the Atlantic Forest, a biodiversity hotspot. Our results show the need to expand these studies, not only to other areas but also in the same region of the present study, thus improving the understanding of ectoparasitism relationships between these groups and how the processes of fragmentation and loss of habitat can interfere in this.

ACKNOWLEDGMENTS
We thank Maycon Contildes, Emerson Martins Pereira and Carlos Alberto Soares Raimundo for their help with fieldwork. Professor Dr. Gustavo Graciolli helped with streblid identification. To teacher Adriano Lúcio Peracchi for the orientation. We also thank Luiz Nelson, owner of Bom Retiro Farm for his friendly welcome to his property. Chico Mendes Institute for Biodiversity Conservation -ICMBio (SISBIO) granted us the research permit (34216-4).