A new genus and species of false vampire (Chiroptera: Megadermatidae) from peninsular Thailand Soisook, Pipat Prajakjitr, Amron Karapan, Sunate Francis, Charles M. Bates, Paul J. J. Zootaxa 2015 2015-03-16 3931 4 528 550 7NDGS Soisook et al. 2015 Soisook et al. [151,482,1460,1486] Mammalia Megadermatidae Eudiscoderma Animalia Chiroptera 4 532 Chordata species thongareeae sp. nov.  Thongaree’s disc-nosed bat   Figures 1–11; Tables 1–2     Holotype. PSUZC-MM. 2014.157 (formerly BL130), adult male, body in ethanol, skull and baculum extracted, collected by A. Prajakjitrand Siriporn Thongareeon  11 April 2001.    Typelocality. Sirindhorn Waterfall, Bala Forest, Halabala WS, Wang District, Narathiwat Province, S. Thailand, 5°48'08" N, 101°49'37"E,  135 masl.    Paratypes.PSUZC-MM. 2014.158 (formerly BL189), field number AP060517-08, adult female, body in alcohol, skull extracted, Halabala Wildlife Research Station, Bala Forest, WS, Wang District, Narathiwat Province, S. Thailand, 5°48'01" N, 101°49'55"E,  140 masl, collected by A. Prajakjitron  17 May 2006;   PSUZC-MM.2014.164, field number SK141007.1, post lactating female, dry skin, skull extracted, Halabala Wildlife Research Station, Bala Forest, WS, Wang District, Narathiwat Province, S. Thailand, 5°48'04" N, 101°49'56"E,  150 masl, collected by S. Karapanand Phuchita Kongon  7 October 2014.   Measurements of the holotype(mm).FA: 66.41, E: 36.05, Tragus: 16.33, HB: 77.41, TIB: 35.46, HF: 18.12, 3MET: 49.44, 4MET: 53.45, 5MET: 58.27, 3D1P: 24.72, 3D2P: 42.56, GTL: 29.35, SL: 28.36, CCL: 26.22, ZB: 17.96, BB: 12.03, MW: 12.93, BCH: 11.98, PC: 4.01, RW: 8.24, C–M3: 10.96, C1–C1: 6.81, M3–M3: 9.45, C–M3: 12.37, M: 20.48, BL: 2.1.   Etymology.The species  thongareeaeis named in honour of Ms Siriporn Thongaree, the former head of Halabala Wildlife Research Station, who dedicated her life to researching the diversity and ecology of wildlife in the southernmost part of Thailand and to promoting its conservation. The proposed English name is “Thongaree’s disc-nosed bat”.   Diagnosis.Discriminating characters are as in the generic diagnosis but with the following additional features. This is a large megadermatid with a FA of 64.36–66.51 mm. The skull is large with a GTL of 29.22–29.44 mmand CCL of 26.01–26.22 mm( Table 2).   Description. (characters additional to those included in the diagnoses).Based on two females, the body mass is 30.0 and 36.2 g. The ears are very large ( Table 2); they join over the forehead at about 30% of the height of the inner margins; these margins are covered with short orange-brown hairs on both the dorsal and ventral sides ( Fig. 1b). The tragus is bifurcate; the posterior process, with a length of 16.33-19.42 mm, is blade-shaped, and with a pointed tip; it is about half the height of the ear. The anterior process of the tragus is much shorter. The dorsal pelage is mid-brown basally and dark brown on the hair tips. The ventral pelage is paler, with whitish brown bases and greyish brown tips. Each wing is broad; the second phalanx of the third digit (3D2P; 41.75–42.56 mm) is distinctly longer than the first (3D1P; 23.52–24.74 mm). The ears, wing and tail membranes are dark brown ( Fig. 1a).   TABLE 1.Matrix summarising selected discriminating characters (dental, cranial, bacular and) of the five genera/six species of Megadermatidaecurrently recognised.    Species First upper premolar(P 2) Anterolingual cingular cusp of uppercanine (C 1) Mesostyle of first upper molars (M 1) ( Fig. 7) Crown area of first lower premolar (P 2) Rostrum of skull ( Fig. 8) Preorbital and posterior orbitalprocesses (Fig. 8) Coronoid process of each half mandible (Fig. 6) Bacular morphology ( Fig. 9)  Principal blood vessels from hind feet merge ( Fig. 1a)    Eudiscoderma  thongareeae absent absent well developed, situated labially equal to or exceeds second premolar (P4) without expanded frontal shield but with deep and broad depression not developed well developed, considerably exceeds tip of C1 long shaft with two short prongs in the interfemoral membrane    Macroderma  gigas absent (but present in fossil  Macroderma) large greatly reduced, situated lingually about 70% of second premolar (P4) with an incipient frontal shield and without depression moderately developed weakly developed, lower than tip of C1 comprising two separate bones, each shaped like a 'bill-hook' in the interfemoral membrane    Megaderma  lyra minute and intruded small to medium greatly reduced, situated lingually about 60–80% of second premolar (P4) without expanded frontal shield or depression preorbital and postorbital processes incipient poorly developed, about equal or lower than tip of C1 comprising two separate bones, each very small and peg-like in the body below the anus    Megaderma  spasma minute and intruded very small not reduced, situated towards labial side 90% of second premolar (P4) without expanded frontal shield but with shallow and small depression preorbital process absent, postorbital process incipient moderately developed, equal to, or exceeds, tip of C1 short shaft with two prongs in the body below the anus    Cardioderma  cor absent absent well developed, situated labially about 50–60% of second premolar (P4) with expanded frontal shield and deep, rounded and broad depression very well developed, especially postorbital well developed, greatly exceeds tip of C1 flattened shaft, with expanded head and base in the body below the anus    Lavia frons absent very small well developed, situated labially about 60–95% of second premolar (P4) with expanded frontal shield and shallow and narrow depression very well developed, especially postorbital moderately developed, exceeds tip of C1 medium shaft with two prongs in the interfemoral membrane    TABLE2.Selected external and craniοdental measurements (in mm) οf  Macroderma gigas,  E. thongareeaegen. sp. nοv.,  Megaderma lyra,  M. spasma,  C. corand  L. frons. Mean, ± SD, minimum (Min) and maximum values (Max) and sample size οf specimens examined are given (sample sizes, which differ frοm thοse indicated, are in brackets). List οf specimens examined definitiοns οf measurements are in the methοd sectiοn.     Characters  M. gigas   E. thongareeae   M. lyra(n=15)   M. spasma(n=25)   C. cor   L. frons  (n=1) ♂ hοlοtype 2♀ paratypes Mean ± SD Min-Max Mean ± SD Min-Max (n=2) (n=2)  FA 102.80 66.41 64.36, 66.51 66.63 ± 2.23 64.00-70.00 [12] 57.88 ± 2.02 53.80-62.50 56.00 [1] 61.80, 60.5  E 52.00 36.05 32.06, 37.49 35.19 ± 3.03 29.20-40.00 [13] 35.27 ± 2.68 30.00-42.69 41.00 [1] 42.20 [1]  ΤRG 22.90 16.33 16.71, 19.42 16.93 ± 2.12 14.92-21.23 [7] 18.38 ± 1.60 14.10-20.55 [21] - -  ΤIB 46.33 35.46 35.47, 36.26 35.06 ± 2.53 32.77-38.01 [5] 32.81 ± 1.02 31.40-34.31 [10] - -  HF 19.14 18.12 16.00, 16.41 16.68 ± 1.67 13.78-19.00 [11] 14.55 ± 1.47 9.30-16.90 [23] 20.00 [1] 14.2 [1]  3MEΤ 68.23 49.44 48.75, 49.78 49.07 ± 3.60 45.34-53.42 [4] 42.65 ± 1.41 40.54-44.30 [6] - -  4MEΤ 74.55 53.45 52.31, 53.77 54.33 ± 3.66 51.09-58.64 [4] 46.61 ± 1.72 43.77-48.48 [6] - -  5MEΤ 79.54 58.27 57.71, 57.75 58.32 ± 4.42 54.51-63.26 [4] 51.28 ± 2.15 48.65-54.39 [6] - -  GΤL 38.01 29.35 29.22, 29.44 29.89 ± 1.22 27.84-31.45 25.78 ± 0.65 24.74-26.63 [23] 26.84, 26.86 26.24, 25.18  CCL 44.44 26.22 26.01, 26.15 26.62 ± 1.05 24.93-27.99 22.97 ± 0.55 21.91-23.95 [23] 23.59, 23.46 23.00, 22.18  ΖB 23.47 17.96 17.42, 17.76 16.94 ± 0.56 16.23-18.41 14.34 ± 0.36 13.61-14.88 [23] 15.68 [1] 15.55, 14.80  BB 15.24 12.03 11.81, 11.99 12.47 ± 0.46 11.99-13.48 10.72 ± 0.27 10.10-11.28 [23] 11.00, 11.12 11.49, 10.81  BCH 14.84 11.98 10.91, 11.25 12.42 ± 0.62 11.85-13.30 [4] 9.95 ± 0.48 9.17-10.99 [22] 9.07, 9.63 9.99, 9.36  MW 16.79 12.93 12.27, 12.33 12.98 ± 0.60 12.19-14.30 11.35 ± 0.26 10.83-11.92 [23] 11.51, 11.84 12.53, 10.90  PC 4.35 4.01 3.87, 3.88 5.24 ± 0.46 4.63-6.15 3.85 ± 0.19 3.54-4.22 [23] 2.56, 3.50 3.99, 3.49  RW 9.70 8.24 7.50, 8.23 7.77 ± 0.20 7.52-8.30 6.89 ± 0.29 6.38-7.48 [23] 6.35, 6.88 7.39, 6.90  C-M3 15.26 10.96 10.81, 10.82 11.56 ± 0.37 10.78-12.21 9.84 ± 0.28 9.39-10.28 [23] 9.85, 9.87 9.61, 9.26  C1-C1 8.19 6.81 6.48, 6.54 6.09 ± 0.35 5.65-6.64 [14] 5.33 ± 0.23 4.75-5.93 [23] 5.80, 6.15 6.14, 5,90  M3-M3 13.57 9.45 9.67, 9.73 10.20 ± 0.27 9.83-10.69 8.54 ± 0.24 8.12-8.94 [23] 8.81, 9.34 9.04, 8.96  C-M3 16.61 12.37 12.19, 12.28 12.65 ± 0.46 11.73-13.30 10.89 ± 0.36 10.19-11.61 [23] 11.12, 11.06 10.71, 10.43  M 27.02 20.48 20.93, 21.05 20.90 ± 0.90 19.39-22.41 [14] 17.79 ± 0.53 16.95-18.70 [23] 18.73, 18.56 17.39, 17.19   FIGURE 5.Occlusal views of the skulls (left) and mandibles (right) of a)  M. gigas(♂HZM.1.20168) from Northern Territory, Australia; b)  E. thongareeae gen. sp. nov.(♂PSUZC-MM2014.157, holotype) from Narathiwat, S. Thailand; c)  M. lyra(♀PSUZC-MM.2014.160) from Satun, S. Thailand; d)  M. spasma(♂PSUZC-MM.2014.163) from Narathiwat, S. Thailand; e)  C. cor(♂HZM.12.2641) from Kenya; f)  L. frons(sex not known, HZM.26.7791) from Democratic Republic of Congo. Scale = 10 mm.   FIGURE 6.Lateral views of the skulls and mandibles of a)  M. gigas(♂HZM.1.20168) from Northern Territory, Australia; b)  E. thongareeae gen. sp. nov.(♂PSUZC-MM2014.157, holotype) from Narathiwat, S. Thailand; c)  M. lyra(♀PSUZC- MM.2014.160) from Satun, S. Thailand; d)  M. spasma(♂PSUZC-MM.2014.163) from Narathiwat, S. Thailand; e)  C. cor(♂HZM.12.2641) from Kenya; f)  L. frons(sex not known, HZM.26.7791) from Democratic Republic of Congo. Scale = 10 mm.   FIGURE 7.Right upper (on left) and right lower (on right) first molars of a)  M. gigas(♂HZM.1.20168) from Northern Territory, Australia; b)  E. thongareeae gen. sp. nov.(♂PSUZC-MM2014.157, holotype) from Narathiwat, S. Thailand; c)  M. lyra(♀HZM.11.39934) from from Madurai, Tamil Nadu, India; d)  M. spasma(♂HZM.5.28163) from Sirsi, Karnataka, India; e)  C. cor(♂HZM.12.2641) from Kenya; f)  L. frons(sex not known, HZM.26.7791) from Democratic Republic of Congo. Scale = 1 mm; ant, anterior; lab, labial. Drawings by PJJB and PS. The skull is large ( Table 2), with a robust rostrum. The sagittal crest extends almost horizontally from the anterior part of the interorbital region to the condyle ( Fig. 6b). The prenasal notch is relatively broad and rounded posteriorly ( Fig. 5b) and extends to the level of the first upper molar (M1). The posterior margin of the bony palate is adjacent to the midpoint of the third upper molar ( M3). Each zygoma is thick and without a dorsal process ( Fig. 6b). The maxillary toothrow (C– M3) is slightly convergent anteriorly. The upper premolar ( P4) is about two-thirds the crown area of the first upper molar ( M1) ( Fig. 4a). The first ( M1) and second ( M2) upper molars are almost square. The heels of P4, M1 and M2 are characteristically rounded, well developed but not elongated ( Fig. 4a). The mesostyle of M2 is situated medially, and in consequence, the labial margin of the tooth is broadly V-shaped. M1 is about two-thirds the crown area of M2; the third molar ( M3) is small and without a metastyle ( Fig. 5b). In the lower dentition, the first ( I2) and second ( I3) incisors are tricuspidate and overlapping. The lower canine (C1) is robust and has a well defined cingulum but is without cingular cusps; it is about twice the height of the first ( P2) and second ( P4) lower premolars; it also exceeds both P2 and P 4 incrown area ( Fig. 4b). The second lower molar ( M2) is the largest of the lower cheekteeth; in both the first and second molars ( M1 and M2) the cristid obliqua is shortened in comparison to the protocristid and metacristid, in consequence the size of the talonid is reduced; in the third molar ( M3), the cristid obliqua is noticeably elongated. In all three lower molars, the protoconid exceeds the hypoconid in height when viewed laterally.   FIGURE 8.Dorsal view of the skulls of a)  M. gigas(♂HZM.1.20168) from Northern Territory, Australia; b)  E. thongareeae gen. sp. nov.(♂PSUZC-MM2014.157, holotype) from Narathiwat, S. Thailand; c)  M. lyra(♀PSUZC-MM.2014.160) from Satun, S. Thailand; d)  M. spasma(♂PSUZC-MM.2014.163) from Narathiwat, S. Thailand; e)  C. cor(♂HZM.12.2641) from Kenya; f)  L. frons(sex not known, HZM.26.7791) from Democratic Republic of Congo. Scale = 10 mm.  Echolocation.  Eudiscoderma thongareeaeemits multi-harmonic, low intensity Frequency modulated (FM) signals ( Fig. 10). Up to six strong harmonics were observed, with a very low frequency first harmonic. The start frequency ( sf) of the sixth (top) harmonic and the terminal frequency ( tf) of the first (bottom) harmonic are 107.6–129.6 kHz (mean± SD:113.9±7.9 kHz) and 11.2–14.0 kHz (mean±  SD:12.7±1.0 kHz), respectively. The maximum energy is usually found in the third harmonic with a peak frequency ( fmaxe) of 53.1–55.1 kHz (mean± SD:54.0±0.6 kHz) ( Fig. 10b, c). The call duration ( d) and inter-pulse interval ( ipi) are 1.8–2.6 ms (mean±  SD:2.1±0.2 ms) and 22.9–91.0 ms (mean± SD: 55.8±23.5 ms), respectively.   FIGURE 9.The dorsal (left) and lateral view (right) of the baculum of a)  M. gigas*(♂HZM.1.20168) from Northern Territory, Australia; b)  E. thongareeae gen. sp. nov.(♂PSUZC-MM2014.157, holotype) from Narathiwat, S. Thailand (note that right prong is partially broken); c)  M. lyra*(♂HZM.7.32146) from Vietnam; d)  M. spasma(♂HZM.3.27330) from Sri Lanka; e)  C. cor(♂HZM.61.31851) from Djibouti; f)  L. frons(♂HZM.9.9664) from Nigeria. Scale = 1 mm. Drawings by PJJB and PS. *: the bacula of  M. gigasand  M. lyracomprise two separate bones, the lateral view of  M. lyrawas not drawn.   FIGURE 10.Spectrograms (a, b) and power spectrum (c) of the echolocation call of  E. thongareeae gen. sp. nov., ♀PSUZC- MM.2014.164 (paratype).  Genetics.DNA barcoding based on the mitochondrial gene COI clearly support genetic divergence between  E. thongareeaeand the sympatric  M. spasmaand  M. lyra( Fig. 11) with a genetic distance of about 20.9% and 19.8%, respectively. It differs from  C. corby 23.7%. Although comparative genetic data for  M. gigasand  L. fronsare not available for comparison in this study, morphological differences between them and  E. thongareeae, as described above, are compelling. The phylogenetic tree lacks strong bootstrap support for the higher level nodes, due to the high divergences within COI and its consequent limited resolution. Nevertheless, this suggests that not only is  M. lyrahighly divergent from  M. spasma(>21% genetic distance), but it is not a sister species to  M. spasma. It should therefore likely be considered in its own genus,  Lyroderma(see discussion, below).   Distribution.Currently,  Eudiscoderma thongareeaeis only known from a small area of Bala Forest in the southernmost rainforest complex of peninsular Thailand ( Fig. 2).  Ecological and conservation notes.Very little is known about the ecology of  E. thongareeae.It was observed roosting in a hollow tree in dense evergreen rainforest near a stream (A. Prajakjitr, pers. comm.). The male specimen ( holotype) was seen hunting a large beetle (order Coleoptera) nearby, before being caught in the middle of a mist net (about 2 mabove the ground) set along a forest trail. Beetle remains were also found under the treeroost, suggesting that this species feeds on large coleopterans (A. Prajakjitr, unpublished data). Unfortunately, this roost was subsequently destroyed when strong winds severely damaged the tree. A female specimen (PSUZC-MM.2014.158) was caught on a rainy night with a hand net, while the bat was hanging in a building of the HBWRS, taking shelter from the weather. A recently collected female (PSUZC-MM.2014.164) was caught in a mist net (about 3 mabove the ground), which was set in an open space in front of a building and close to the forest edge. This species appears to be very rare. Despite extensive bat surveys in southern peninsular Thailand over the last 15 years, it is still only known from three specimens, all of which were collected from a small area of Bala Forest. This low capture rate is similar to the case of two other rare, recently discovered, bat species in the Bala Forest,  Murina balaensis( Soisook et al.2013)and  Kerivoula krauensis( Douangboubpha et al.2014). Much less research has been undertaken on the Malaysian side of the border, so it is possible that it also occurs there, but the rapid loss of lowland rainforest in peninsular Thailand and Malaysia suggests that it is very likely at risk and should be considered as a top priority for conservation.   Comparisons.  Eudiscoderma thongareeaeshares with the African  Cardioderma corthe characters of a deep rostral depression and an enlarged coronoid process in each half mandible. In the dentition, like  C. cor, the upper and lower canines are without discrete anterolingual cingular cusps; the small upper premolar (P2) is absent and the mesostyle of the first upper molar (M1) is situated labially and not reduced. Externally there are similarities in the morphology of the noseleaf. However, as can be seen from Tables 1and 2, there are also many differences including the larger size of  E. thongareeae; marked differences in the shape of the baculum ( Fig. 9) and differences in the structure of the blood vessels in the interfemoral membrane. In the skull, there are differences in the shape, noticeably the absence in  C. corof an abrupt, step-like discontinuity between the rostrum and the anterior braincase ( Fig. 6) and the absence in  E. thongareeaeof well developed preorbital and postorbital processes and an associated supraorbital flange ( Fig. 8). There are also differences in the teeth. For instance, the upper and lower canines of  C. corare much less robust; the preparacrista is shortened and the postmetacrista elongated in the first upper molar (M1) of  C. corbut about equal in  E. thongareeae( Fig. 7); and the first lower premolar (P2) is enlarged in  E. thongareeaebut is relatively small in  C. cor.   Eudiscodermahas some similarities with the other African taxon of megadermatid,  Lavia frons.Of the three subspecies recognised for  Lavia frons, two subspecies,  affinis Andersen & Wroughton, 1907and  rexMiller, 1905, are described on the basis of size differences but they are considered poorly defined by Happold ( 2013). In both  E. thongareeaeand  L. frons, the principal veins from the hind feet merge in the interfemoral membrane prior to entering the body (although in contrast to  E. thongareeae,there does not appear to be a well defined blood vessel in  L. fronsrunning down the centre line of the membrane). The baculum shows some similarity in form, although that of  E. thongareeaehas a relatively longer and more robust shaft and shorter prongs ( Fig. 9). The upper cheek teeth also show a similarity in structure, noticeably in the development of the mesostyles in M1 and M2 and in the relative sizes of the preparacrista and postmetacrista ( Fig. 7). The small upper premolar ( P2) is absent in both taxa. However, there are also many significant differences. For instance, the noseleaf of  L. fronsis large, broad and with convex-sides and a truncated top; it has a longitudinal ridge which broadens at the base and is distinctively yellow in colouration (see Monadjem et al. 2010, figure 139; Happold 2013, figure 76). The pelage colour is also distinctly different, with that of  L. fronscharacteristically tinged with yellow. In the skull, the rostrum of  L. fronshas only a shallow and narrow rostral depression but has well developed preorbital and postorbital processes and an expanded supraorbital flange ( Fig. 8) and the coronoid process of each half mandible is not enlarged ( Fig. 6); the upper and lower canines (C1 and C1) are less robust in  L. fronsand have anterolingual cingular cusps.   Eudiscodermashares with the Australian  Macroderma gigasthe absence of the small upper premolar (P2), although this tooth is present in fossil species referred to  Macrodermaby Hand (1985). However, in most other dental features there are noticeable differences, including the structure of the upper and lower canines, premolars and molars ( Table 1). In the skull, there are differences in the shape of the rostrum, braincase and coronoid process of each half mandible. Externally, there are differences in the shape of the noseleaf, the posterior leaf of which is elongate in  M. gigas.In size,  M. gigasgreatly exceeds  E. thongareeae( Table 2). In bacular morphology,  M. gigasdiffers markedly, possessing a structure that comprises two separate bones, each shaped like a ‘bill-hook’ and with a saddle-like expanded base ( Fig. 9a).   Eudiscodermadiffers from  Megaderma lyrain all the characters listed in Table 1, noticeably including its baculum, which in  M. lyracomprises two separate, short, thin bones (based on two specimens from Thailand and Vietnam respectively) ( Fig. 9c). Based on available descriptions,  E. thongareeaealso differs from all subspecies and synonyms of  M. lyra(mapped in Fig. 2). Two of these, namely caurina Andersen & Wroughton, 1907and  sinensis Andersen & Wroughton, 1907, differ from the nominate form only in minor size differences. On the basis of figures included in the typedescriptions and comments of Andersen and Wroughton (1907), the forms spectrum(Wagner, 1844) and  schistacea( Hodgson, 1847)appear to be synonymous with typical  lyra.The holotypeof carnatica( Elliot, 1839) from western India was described as having only three teeth behind the canine in the upper toothrow. However Elliot (1839)himself referred the specimen to ‘  M. lyra?’ and its status remains unclear.   Eudiscodermaoccurs sympatrically with the smaller  M. spasma( Fig. 2). There are differences between the two taxa in external, bacular, cranial and dental characters ( Table 1). Namely in the dentition,  E. thongareeaehas only one upper premolar (P4) since the minute, intruded premolar (P2) found in  M. spasmais absent. Unlike  M. spasma,the upper canine (C1) of  Eudiscodermais noticeably robust, especially in crown area; it is without an anterolingual cingular cusp and has a much larger posterior accessory cusp, which unlike that of  M. spasmais not well-separated from the main cusp. In the upper molars, the structure shows some similarity between the two taxa, although in  M. spasma, the postmetacrista in M1 is more elongated in comparison to the preparacrista than in  Eudiscoderma( Fig. 7). The anterior braincase of  E. thongareeaeis more elevated than that of  M. spasmaand in consequence the profile of the upper braincase is almost straight whereas it is angular in  M. spasma( Fig. 6). The rostrum of  M. spasmais also less robust and without a deep rostral depression; however, the lack of development of preorbital and postorbital processes is comparable between the two taxa. Externally, size apart ( Table 2), the morphology of the noseleaf ( Fig. 1b, d) is similar, although the posterior leaf of  M. spasmais sometimes more angular than the rounded  E. thongareeae. The vein pattern in the interfemoral membrane differs between the two taxa ( Table 1). The bacula are of a similar general form but the relative proportions of shaft to prongs differ, with that of  E. thongareeaehaving a more robust and longer shaft and shorter prongs ( Fig. 9).   Megaderma spasmahas 16 synonyms, all of which were variously accepted as valid subspecies by Simmons (2005). Geographically, they extend from the typelocality of the nominate form in the Moluccas, Indonesia to those of  horsfieldiiBlyth, 1863and ceylonense Andersen, 1918described from western India and Sri Lanka respectively ( Fig. 2). Of the synonyms which are most adjacent to the typelocality of  E. thongareeae,the forms abditumChasen, 1940 and medium Andersen, 1918from peninsular Malaysia and Singapore respectively, differ only in minor size differences from typical  spasma. Of the Indonesian forms,  siumatisLyon, 1916,  lasiaeLyon, 1916, and niasenseLyon, 1916 are island races, located off the west coast of Sumatra, that exhibit some small variations in external, cranial and dental dimensions. In Java, the form  trifolium Geoffroy St. Hilaire, 1810is considered to represent typical  M. spasmaon the basis of descriptions included in Andersen and Wroughton (1907). Specimens referred to the taxon pangandaranaSody, 1936 average slightly larger than typical  trifolium. The taxon  carimataeMiller, 1906from Karimata Island, west of Borneo is a typical  spasmabut with slightly larger ears. The taxon  natunae Andersen & Wroughton, 1907was described on the basis of one specimen collected from Bungaran Island, off the north-west coast of Borneo. It is very large. However, the cranial and dental morphology clearly shows that it belongs to  spasma;this agrees with the previous findings of Tate (1941). On mainland Borneo, the form kinabaluChasen, 1940 differs only from typical  spasmain minor size differences; and the form  celebensisShamel, 1940from Sulawesi was described on minor differences in size and colour. According to Andersen and Wroughton (1907), specimens referable to the taxon  philippinensisWaterhouse, 1843from the Philippines are ‘in every respect indistinguishable from Celebes individuals.’ On mainland Southeast Asia, minus Andersen, 1918from Cambodia, as the name suggests, is small. In Myanmar, the taxon majus Andersen, 1918, despite its larger size is clearly referable to  spasmaon the basis of its cranial and dental morphology; it was considered indistinguishable from the Indian taxon  horsfieldiiBlyth, 1863by Sinha (1977). The Sri Lankan ceylonense Andersen, 1918is a small race of  spasma. In terms of echolocation, the call characters of  E. thongareeaeare generally similar to those of the sympatric  M. spasmaand  M. lyra,which is also known to occur in peninsular Thailand. The call records of individuals of  M. lyra(n=2) and  M. spasmafrom Thailand (n=1) and India (n=1) also have multi-harmonic FM signals with four to six strong harmonics (PS unpublished data). However, the fmaxeof the third harmonic of both species is slightly higher than in  E. thongareeae.In  M. lyra, the maximum energy sometimes lies in either the second harmonic, with a fmaxeof 41.4–49.9 kHz (mean±SD: 44.1±2.3 kHz) or third harmonic with an fmaxeof 58.3–68.5 kHz (mean±SD: 64.2±3.2 kHz). In  M. spasma, the maximum energy is usually found in the third harmonic, with an fmaxeof 56.5–65.9 kHz (mean±SD: 59.8±3.9 kHz). In the Australian  M. gigas, Nelson (1989)summarised the call characters of this species as having a low intensity and frequency of the first harmonic with an fmaxof about 20 kHz. The maximum energy is in the second and third harmonic with an fmaxeduring the prey catching flight of about 60 kHz. The African species,  Cardioderma cor, uses low intensity broadband and short duration FM sweeps with an fmaxeof 56.7±11 kHz ( Taylor et al., 2005). For  Lavia frons, detailed data are not available, although it is known to use low intensity broadband FM signals ( Happold, 2013). 2979705302 2001-04-11 PSUZC A. Prajakjitr & Siriporn Thongaree Thailand 135 5.8022223 Sirindhorn Waterfall, Bala Forest, Halabala WS, Wang District 21 101.82694 4 532 MM.2014.157 (formerly BL130) 1 1 Narathiwat Province holotype 2979705301 2001-04-11 PSUZC A. Prajakjitr & Siriporn Thongaree Thailand 135 5.8022223 Sirindhorn Waterfall, Bala Forest, Halabala WS, Wang District 21 101.82694 4 532 MM.2014.157 (formerly BL130) 1 1 Narathiwat Province holotype 2979678302 2006-05-17 PSUZC A. Prajakjitr Thailand 140 5.800278 Halabala Wildlife Research Station, Bala Forest, WS, Wang District 21 101.83194 4 532 MM.2014.158 (formerly BL189) 1 1 Narathiwat Province paratype 2979678303 2014-10-07 PSUZC S. Karapan & Phuchita Kong Thailand 150 5.801111 Halabala Wildlife Research Station, Bala Forest, WS, Wang District 21 101.83222 4 532 MM.2014.164 1 1 Narathiwat Province paratype