Occurrence and infestation rates of Streblidae (Diptera, Hippoboscoidea) on bats (Mammalia, Chiroptera) in a semideciduous seasonal forest fragment in western Paraná, Brazil

DOI 10.1590/1678-4766e2020026 ABSTRACT. The objective of this work was to evaluate the occurrence and infestation rates of Streblidae species on bats in a semideciduous seasonal forest fragment in western Paraná, South of Brazil. A total of 289 specimens of eight Streblidae species were collected from six species of bats belonging to the family Phyllostomidae. The most frequent Streblidae species were Megistopoda proxima (Séguy, 1926), Aspidoptera falcata Wenzel, 1976 and Paratrichobius longicrus (Miranda Ribeiro, 1907). The highest prevalence found was Megistopoda aranea (Coquillet, 1899) on Artibeus planirostris (Spinx, 1823). The highest intensity found was M. aranea on Artibeus lituratus (Olfers, 1818). The highest abundance found was M. proxima on Sturnira lilium (E. Geoffroy, 1810). All recorded species were highly specific to their hosts, except M. aranea . Two infracommunities were recorded: M. aranea and Metelasmus pseudopterus (Coquillett, 1907) on A. planirostris , and M. proxima and A. falcata on S. lilium . The results showed a trend of occurrence of ectoparasites on females, considering the most frequently captured bats, the seasons had no pattern in the prevalence and intensity factors in this sampling area to them. Studies evaluating the life cycle of these parasites, and the behavior and morpho-physiology of these hosts are necessary since the parasitism rate can be affected by other factors that were not evaluated in

Bat flies are highly specialized hematophagous ectoparasites connected exclusively with bats; these species are divided into two families, Streblidae and Nycteribiidae (Dick & Patterson, 2006). They present cosmopolitan distribution, but the Streblidae family is more diverse in the Americas; and 96 Streblidae (Diptera) species and 23 genera were found in Brazil (Graciolli, 2020).
Different extrinsic factors may affect the rate of ectoparasites on the host, such as bat body size, gregarious habit and geographical distribution, and the abundance of these parasites may vary depending on the bat sex, reproductive condition, age, and sanity within a same species (Marshall, 1982;Rui & Graciolli, 2005). Few studies on these factors are found in Brazil (Linhares & Komeno 2000;Rui & Graciolli, 2005;Santos et al., 2009) and are mainly focused on taxonomic inventories and descriptive analyzes (Torres et al., 2019).
The state of Paraná presents the greatest richness of bat species (64) and Streblidae species (33) in the South of Brazil (Passos et al., 2010;Lourenço et al., 2016;Graciolli, 2020). Moreover, Paraná is one of the Brazilian states with the highest number of records and studies of bat's ectoparasites Graciolli et al., 2017); however, it lacks surveys of these species in several regions (Prevedello et al., 2005).
Studies of ectoparasites on bats is incipient; thus, studies in new regions can assist in the understanding of the geographic distribution of these parasites and their relations with their hosts, in addition, ectoparasitological analyzes provide us with a better understanding of evolutionary and ecological mechanism of this parasite-host relationship (Prevedello et al., 2005;Barbier et al., 2016;Vasconcelos et al., 2016;Torres et al., 2019).
In this context, the objectives of this work were to evaluate (i) the occurrence and infestation rates of Streblidae species on bats in a semideciduous seasonal forest fragment in western Paraná, South of Brazil, (ii) the main parasitological indexes, and (iii) the effect of the sex of the host and the seasons on the parasitism rate.

Study area.
The work was carried out at the São Camilo State Park (SCSP), in Palotina, western Paraná, south of Brazil (24°08ʼ57ˮS; 53°54ʼ38ˮW). This park covers 0.57% of the Palotina territory; it has an area of 385.34 hectares of semideciduous seasonal forest -a phytophysiognomy of the Atlantic Forest. This area is one of the few forest remnants in the region, which had most of its original vegetation removed for agricultural crops; the surroundings of the SCSP consists of soybean and maize crops.
According to the Köppen classification, the climate of the region is subtropical humid (Cfa), with average temperature lower than 18°C in the coldest months and above 22°C in the warmest (Fig. 1). The regions annual precipitation varies from 1600 mm to 2000 mm, with concentrated rainfall in the summer, and no defined dry season (IAP, 2006).

Procedures.
Bat specimens were collected monthly in four nights of third quarter or new moon between September 2016 and August 2017. Eight mist nets (12 x 2.5 m) were distributed in tracks inside the forest, in the forest fragment edges, and near water bodies. The nets were opened after sunset and checked every 20 minutes for six hours. The sampling effort of this study was 69,120 m 2 h -1 (Straube & Bianconi, 2002).
The captured bats were removed from the net, placed in cotton bags, and taken to the field base, where they were identified regarding species and sex. The bats were identified based on the keys developed by Gardner (2008), Gregorin & Taddei (2002) and Miranda et al. (2011). The ectoparasites found were collected manually or with tweezers and fixed in 70% ethyl alcohol in individual containers. The bats were tagged with numbered metal rings (code: PESC) in order to identify the recaptured individuals. After the screening they were released at the same spot where they were captured.
At least one specimen of each bat species was selected as material testimony, under the licence of the Instituto Chico Mendes de Conservação da Biodiversidade (ICMBio Licence #43560-2). The procedures were approved by the Ethical Committee on the Use of Animals of Universidade Federal do Paraná, Palotina Sector (CEUA n° 39/2014). The specimens were deposited in the Museu de História Natural Capão da Imbuia, in Curitiba PR, Brazil (Appendix 1).
The Streblidae species were identified according to the pictorial key developed by Graciolli & Carvalho (2001) and deposited in the Padre Jesus Santiago Moure Collection of the Department of Zoology of the Universidade Federal do Paraná, Curitiba, PR, Brazil.
Data analysis. The prevalence rates of species of captured bats and ectoparasites found on them (infested bats/ examined bats), mean infestation intensity (amount of bat flies/infested bats), mean infestation abundance (amount of bat flies/examined bats) and parasite-host specificity index (percentage of total bat flies of a single species found on the primary host) were evaluated (Bush et al., 1997;Dick & Gettinger, 2005).
Statistical tests were performed for the three most frequent bat species in the study area with Streblidae infestation of higher specificity for these species. The effect of the sex of the host, and seasons on the prevalence was evaluated using the chi-square test. The differences in intensity due to host sex was evaluated using the Bootstrap-T Test with randomization with 2000 replications (Rózca et al., 2000) and the differences in intensity due to the seasons was evaluated using the Kruskal-Wallis test. The Spearman's correlation test was used to analyze the abundance of parasites in the most frequent infracommunities. The analyses were performed in the Quantitative Parasitology 3.0 (Rózca et al., 2000) and BioEstat 5.3 (Ayres et al., 2007) programs.
The season analyzed in this study were: Summer (December, January and February), Autumn (March, April and May), Winter (June, July and August) and Spring (September, October and November).
The parasites collected on recaptured bats throughout the study were not statistically evaluated; they were discussed separately in order to understand possible patterns of infestation or reinfestation of the bat flies on the hosts.

RESULTS
A total of 380 bat specimens were captured, which consisted of 12 species of three families -Molossidae: Twenty-six bats were recaptured in different months, 50% of them were not infested with Streblidae species neither in the capture nor in the recapture, 34% were infested only at recapture, 8% only at capture, and 8% were sampled with bat flies in both capture and recapture. One individual of A. lituratus was parasitized with one specimen of P. longicrus and was recaptured three days later with two specimens of P. longicrus, even though the Streblidae specimen was removed after the first capture. Higher prevalence of P. longicrus on A. lituratus (x²=13,602, df=3, p= 0.003) was found in the winter. Higher prevalence of A. falcata on S. lilium (x²= 1,046, df= 3, p= 0.790), and M. proxima on S. lilium (x²= 7,434, df= 3, p= 0.059) was found in the summer, but with no statistical differences. Megistopoda aranea on A. fimbriatus had higher prevalence in the autumn and presented no significant differences (x²= 2,638, df= 3, p= 0.451). The mean intensities due to the seasons had no significant differences for any host-parasite combination (Tab. IV).

DISCUSSION
The results obtained in this paper demonstrate important descriptive data on the parasitic pattern of Streblidae on bats in a Semideciduous Seasonal Forest fragment in Western Paraná. The study evaluated the prevalence of parasitized bats and parasitological indexes founded over the course of a sample year and presented interesting findings regarding host sex and seasons. The data obtained were coincident in terms of host-ectoparasite associations, as well as regarding the bats sampled, corroborating with the scientific literature (Graciolli & Carvalho, 2001;Miretzki, 2003;Bianconi et al., 2004;Prevedello et al., 2005;Anderson & Ortêncio-Filho, 2006;Ortêncio-Filho & Reis, 2009).
Neotropical regions present positive correlation of richness of Phyllostomidae and Streblidae species (Dick & Gettinger, 2005;Patterson et al., 2008;Dornelles et al., 2017). Studies of semideciduous seasonal forest in the northwest of Paraná (Silva & Ortêncio-Filho, 2011), Rio de Janeiro (França et al., 2013) and in an area influenced by the Cerrado biome in the northwest of Paraná (Anderson & Ortêncio-Filho, 2006) found five, nine and six Streblidae species on five, seven and four Phyllostomidae species, respectively. Thus, the richness of Streblidae species found in São Camilo State Park (SCSP) can be explained by the richness and abundance of Phyllostomidae species found when compared to the other captured families (Molossidae and Vespertilionidae).
Megistopoda aranea presents a connection with Artibeus bats, especially A. fimbriatus (Graciolli & Carvalho, 2001). The occurrence of M. aranea on A. lituratus may be due to the proximity of this bat species with the M. aranea preferred host (Rui & Graciolli, 2005). A higher prevalence of M. aranea on A. planirostris and A. fimbriatus was found in the SCSP.
The recaptures showed 34% of infestation and 8% of reinfestation of bat specimens by ectoparasites. Several factors may explain the infestation or reinfestation of bats by ectoparasites, such as the gregarious habit of these hosts that allows the transference of ectoparasites, the number of hosts in their shelters and the infracommunities of parasites on other hosts that share the same shelters (Patterson et al., 2007;McCoy, 2009). Bat recapture data from bat flies' studies are interesting, since it can demonstrate infestation and reinfestation patterns. Our data showed that even bats without parasites captured at first, or bats cleaned during the sampling, later presented Streblidae specimens on their body, corroborating the importance of gregarious habit and the shelter as a dispersing factor for these ectoparasites.
According to Lourenço & Esbérard (2011), reinfestation may occur rapidly in the first few hours of contact between hosts; this explains the rapid reinfestation of the recaptured A. lituratus, within three days.
The M. proxima found on a recaptured A. lituratus was probably an accidental infestation. According to Graciolli & Carvalho (2001), this ectoparasite commonly leaves the host at capturing, thus, when the primary host (S. lilium) falls in the net or at manipulation, they can infest other bats. As a suggestion for future studies, considering the specificity of Streblidae, we believe that is interesting to analyze the impact of the accidental dispersion on the host or to the ectoparasite.
The infracommunity of A. falcata and M. proxima on S. lilium is well documented, with high correlation between these two species of parasite and the genus Sturnira (Linhares & Komeno, 2000;Bertola et al., 2005;Dick & Gettinger, 2005;Anderson & Ortêncio-Filho, 2006;Dornelles et al., 2017). A higher frequency of A. falcata associated with M. proxima was found compared to the A. falcata found alone on the hosts.
Linhares & Komeno (2000) report a correlation between simultaneous parasitism of A. falcata and M. proxima on S. lilium, in which the high the abundance of one parasite the lower the abundance of the other. This pattern was not observed in the present study. The competition between parasite species does not seemed to affect the infracommunities of bat flies, but probably the ecology, behavior, and morphology of the host can affect the occurrence of these infracommunities (Tello et al., 2008;Presley, 2011;Patrício et al., 2016).
The higher prevalence of ectoparasites found on female bats can be explained both by the behavior of the host and the biological cycle of the parasite. Female bats tend to care for their progeny, staying longer in the shelters, where Streblidae species have their larviposition, and pupa and adult development (Graciolli et al., 2008;Esbérard et al., 2012).
Several studies confirm that humidity and temperature are abiotic factors that affect the biological cycle of Streblidae species, and a trend of mortality of ectoparasites due to the reduction of temperature and humidity (Marshall, 1982). Higher prevalence and mean intensity of P. longicrus on A. lituratus, and M. proxima on S. lilium during the warmer season were found in the Atlantic Forest in the state of Rio Grande do Sul, Brazil (Rui & Graciolli, 2005). Similar results were found in the Cinturão Verde Municipal Park in the northwest of Paraná (Anderson & Ortêncio-Filho, 2006). Different results were found in the SCSP, with higher prevalence of P. longicrus on A. lituratus in the winter. Although there is no dry season in the region, winter presents minimum temperatures of 11°C, thus, no higher rates of parasitism in this period was expected. This denotes that changes in temperature and precipitation are not affecting the parasitism rates in the SCSP.
Studies on the host behavior, body size, nesting, cooccurrences, health conditions, and the biological cycles of the parasites are necessary, since these factors may affect the ectoparasitic rates of the studied bats, and these studies can provide a better understanding of the relationship between these parasites and hosts.
Tab. IV. Prevalence indices (P (%)), mean intensity (MI) and number of M. aranea on A. fimbriatus; P. longicrus on A. lituratus; A. falcata on S. lilium; and M. proxima on S. lilium as a function of the seasons during the sample period (NE, number of ectoparasites collected; EH, number of examined hosts).