Checklist of bat flies (Diptera: Nycteribiidae and Streblidae) and their associated bat hosts in Malaysia

: The number of publications on bat flies of Malaysia suggests that this group of parasitic dipterans is understudied. From April 2011 to September 2013, we surveyed 10 localities from seven states in Malaysia with the main objective to compile a checklist of bat flies. As a result, a total of 15 species of bat flies were recorded from 24 species of hosts. Our surveys indicated that there is a correlation between the number of bat species and the number of bat flies species recorded due to the host specific nature of bat flies.


INTRODUCTION
Bat flies are parasitic dipterans that are only found on bats. Bat flies from the families Streblidae and Nycteribiidae are part of the superfamily Hippoboscoidea, which includes other obligate sanguivorous families such as Glossonidae and Hippoboscidae (Dittmar et al. 2006). Generally, nycteribiids and streblids can be distinguished easily by the presence (streblids) or absence (nycteribiids) of wings.
Based on a compilation from various sources, there are approximately 30 species of bat flies recorded from Malaysia (Theodor 1967;Maa 1968Maa , 1971. Despite these studies, knowledge of the taxonomy and ecology of bat flies in Malaysia is scarce and there are no recently published literatures on this group from Malaysia since the late 19th century. To remedy this, surveys were conducted to explore the diversity and to compile a checklist of the bat flies and their incidental hosts. Various surveys were carried out between April 2011 and September 2013 to record the host and parasite association of bats and bat flies at several sites in Malaysia. Herein, the results from our surveys are reported.

MATERIALS AND METHODS Study sites
Fieldwork was carried out at 10 different localities in Malaysia. The study sites in Sarawak, Malaysian Borneo include Batang Ai National Park, Gunung Gading National Park, Mulu World Heritage Area and Wind Cave Nature Reserve. The sites in mainland Malaysia include Air Terjun Batu Hampar (Kedah), Gua Musang district (Kelantan), Wang Kelian State Park (Perlis), Pulau Tioman Wildlife Reserve (Pahang), Sekayu Recreational Forest (Terengganu) and Tambun (Perak) (Figure 1).
These study sites have varied habitats, including mixed dipterocarp forest, hill dipterocarp forest, alluvial forest, secondary forest, heath forest, forest near limestone cave areas, and agricultural plots. The details of each trapping site are shown in Table 1.

Data collection
Fieldwork was conducted from April 2011 to September 2013. Different types of sampling methods were used to maximize the representativeness of the bat host species. Two four-bank-harp traps (Francis 1989) and 10 polyester nets were positioned in forest understories, across small streams, and at forest edges. Additionally, hand nets were used to collect roosting bats.
During each sampling day, mist nets and harp traps were set up before the emergent time of bats and were closed at 7:00 h the following day. Mist nets were checked every 15-30 minutes for the first three hours and then every two to three hours before they were closed the next morning. Meanwhile, harp traps were checked three to five times during each sampling night.
Captured bats were placed inside cloth bags. Each bat specimen was identified following Payne et al. (1985), Kingston et al. (2006) andFrancis (2008). Each host was scanned with an intensive light source for bat flies

Lists of species
Azhar et al. | Checklist of bat flies in Malaysia and bat flies were identified following the publications of Jobling (1930), Theodor (1967), and Maa (1968Maa ( , 1971. The voucher specimens were deposited at the Parasitology Laboratory at Universiti Malaysia Sarawak.

Statistical analysis
To evaluate the relationship between the number of host species collected and their associated bat flies species, a simple linear regression analysis was performed using the statistical package available in Microsoft Excel® 2010. The number of bat species with bat flies (host) captured from each site (10 localities) were used as the independent variable while number of bat fly species collected at each site from different hosts were used as the dependent variable. A scatterplot with at the fur area of the body, around the wing, and the uropatagium. Following this, bat flies from each host were stored inside labeled cryovial tubes in 70% ethanol. Hosts were then released after the inspection process. In order to minimize cross-contamination between hosts and bat flies, each cloth bag was cleaned before re-used.

Identification of bat flies
Collected bat flies were first sorted based on morphological characteristics under dissecting microscope. Sorted bat fly specimens were then cleaned using lactophenol solution and mounted on slides with Hoyer's medium. Cover slips were ringed with nail polish to prevent desiccation of medium during storage. Prepared specimens were examined using compound microscope  a linear trendline (regression) was produced. Trendline was enforced to intercept on Y-axis at zero value, indicating that no bat flies were recorded when there was no host.

RESULTS
There were 15 species of bat flies recorded from 24 species of bats. The bat flies were represented by subfamilies Nycteribiinae (three genera and four species), Archinycteribiinae (one genus and one species), Cyclopodiinae (two genera and five species) and Brachytarsininae (three genera and five species). The hosts were represented by five families, which are Pteropodidae (six genera and ten species), Rhinolophidae (one genus and four species), Hipposideridae (one genus and six species), Vespertilionidae (three genera and four species) and Miniopteridae (one genus and one species) ( Table 2). In addition, the correlation graph indicated that there was a moderate correlation between the numbers of host species recorded with the number of bat flies species encountered ( Figure 2). The equation for the linear regression is given by y = 1.0719x, R 2 = 0.4804.
Host -Myotis horsfieldii, Murina suilla. Remarks -Eyes are absent. The genitalia have conical phallobase without a dorsal bulge and the aedeagus is slightly curved, with broadly rounded end as wide at the apex as at the base (Theodor 1967). The only previous record   of this particular nycteribiid bat fly was from Selangor, collected from an unidentified host (Theodor 1967).
Remarks -The individuals of this form resemble the characters presented by Stylida caudata. Only female individuals were recorded during sampling. Eyes are absent and labella are short. There is a double genital plate with a basal elliptical cover protruding out from the surface, with three minute hairs on each side at the base and a shield-shaped plate on the surface (Theodor 1967). Host -Hipposideros larvatus.
Remarks -The morphology presented by these individuals are similar to Stylidia euxesta. This species is found in Burma and it has been reported that different subspecies can be found in Thailand and Malaya, and possibly in India as well (Theodor 1967). Eyes are absent. The darkly pigmented claspers are straight, with long dorsal setae, which become shorter the lower they are on the tip. The basal arc is large and rounded with a short anterior process (Theodor 1967).
Remarks -This species can be found in abundance in West Malaysia and is possibly host specific to the bamboo bats, Tylonycteris pachypus and T. robustula (Marshall 1971). The eyes usually have either two or three unequal-sized lenses with a pigmented base. The triangular shaped basal arc has a long and narrow anterior process. The claspers are slightly curved, with setae present on the dorsal surface (Theodor 1967).
Remarks -This species has the highest number of lenses among the members of Nycteribiidae. The eyes are a pair of lenses at each side of the two elliptical frames. The short claspers are tapered to a blunt point with a short ventral hook. The basal arc is triangular (Theodor 1967  Host -Rousettus amplexicaudatus, R. leschenaultii, Eonycteris spelaea. Remarks -The eyes are of single elliptical lenses. The head is compressed and the thorax is pentagonal. The claspers are short and tapered to a point with two rows of pegs in the middle half and hair laterally (Theodor 1967).
Remarks -Only two individuals were recorded from a single host, Dyacopterus spadiceus. The eyes are incomplete divided lenses and the thorax is slightly wider. The clasper is unique in this species because it is strongly spined, with a scantly setose phallobase and a prominent basal process on the inner margin of the anal segment (Maa 1968). The weakly sclerotized dorsal ridges end in two points (Theodor 1967). Leptocyclopodia obliqua Theodor, 1959 Specimen examined ( Figure 12) -(2) Batang Ai National Park, Sri Aman, Sarawak (BTGA003; BTGA027); (2) Gunung Mulu World Heritage Area, Sarawak (GMWHA006; GMWHA017).
Host -Balionycteris maculata. Remarks -The eyes are similar with L. brachytrinax.
The thorax is triangular with about 15 blunt spines. The slender claspers are tapered to a blunt tip in the apical half (Theodor 1967). The spines at the apex of claspers in L. ferrari is absent.
Remarks -The distribution of this species is restricted to the Oriental region (Maa 1971), although there are   several subspecies. Megastrebla gigantea has a very robust body and the wings are very long. The vein R 1 is bent near the base and almost straight at the apical section, and the apex of vein R 2+3 is more distant than that of vein R 4+3 (Maa 1971). (Maa, 1971) Specimen examined ( Figure 14) -(1) Gunung Reng, Jeli, Kelantan (GRK011).

Megastrebla nigriceps
Host -Penthetor lucasi. Remarks -This species is the sole representative of a distinct species group. The body is slender compared to M. gigantea and M. limbooliati. The wings are long, consistent with other members of genus. The vein R 1 has a weak bend near the base, with vein R 2+3 apically curved to C, the apex is nearly equidistant to R 1 and vein R 4+5 apices and the first abscissa of vein R 4+5 is distinctly oblique to vein R 2+3 (Maa 1971   Host -Rhinolophus creaghi, R. lepidus, R. affinis, R. stheno, Miniopterus medius. Remarks -All individuals of this species were collected from the wing membranes of bats of the family Rhinolophidae (216 individuals) except for a single one, which was collected from Miniopterus medius. This medium-sized unidentified species resembles the morphological description for members of the genus Brachytarsina. The head is triangularly rounded with non-prominent eyes. The wing veins M 1+2 with the first abscissa are much longer than the second abscissa while vein M 3+4 is very slightly curved and does not subangulate near the base (Maa 1971). This genus contains about 23 species in the Oriental region and is reported to be parasitic mostly on bats from the families Emballonuridae, Hipposideridae and Rhinolophidae (Maa 1966). Remarks -All individuals of this species were collected from the fur area of bats from family Hipposideridae. This small-sized bat fly possibly belongs to the genus Raymondia. Eyes are absent. The head is narrower than the interdistance between the two anterior most major humeral setae and it has a roundish labial theca almost as long as it is broad (Maa 1962). Wings are short and broad, with five longitudinal and two cross veins and a very setose costa (Jobling 1930).

DISCUSSION
There is little published on the ecology, taxonomy and distribution of Malaysian bat flies. Until now, there were at least 30 species of bat flies recorded in Malaysia (Theodor 1967;Maa 1968Maa , 1971). Our study recorded approximately half of the total number of bat fly species previously recorded from Malaysia. Unlike their hosts, the number of bat flies species encountered per survey is highly dependent on the number of host species recorded (Dick and Gettinger 2005). There was a moderate correlation between the numbers of host species recorded with the number of bat flies species encountered in our study. This compliments the statement by Marshall (1980) where the Malaysian nycteribiid and streblid bat flies are highly specialized; more than 50% were limited to a single host species. Thus, we speculate that higher number of host species documented would yield more bat flies species per survey.