AMERICAN MUSEUM Novltates

PUBLISHED BY THE AMERICAN MUSEUM OF NATURAL HISTORY CENTRAL PARK WEST AT 79TH STREET, NEW YORK, N.Y. 10024 Number 2778, pp. 1-48, figure 1 January 31, 1984

Taxonomic and Distributional Notes on Tropical Australian Bats

KARL F. KOOPMAN1

ABSTRACT The bats oftropical Australia are reviewed with the three areas adjacent to tropical Australia, New some 51 species recognized, though a few are very Guinea shares a large number of species, whereas poorly known. A new subspecies, Pipistrellus ten- temperate Australia and the Lesser Sunda Islands uis westralis is described and Rhinolophus me- share relatively few. New Guinea has probably gaphyllus ignifer is synonymized with R. m. me- been an important source area for tropical Aus- gaphyllus. The two previously recognized tralian bats, particularly those confined to the Cape subspecies of Macroderma gigas are also synon- York Peninsula. The low level ofendemism among ymized. A majority oftropical Australian bats are Australian bats strongly implies that there were restricted to mesic areas, but a number are more no bats in Australia prior to the Miocene, when or less arid tolerant. Only one species (Taphozous Australia drifted far enough to the north to be able hilli) actually avoids mesic areas. The Cape York to receive species occurring on the extended Malay Peninsula has the greatest number of species with archipelago. Since then some low level endemism a falling off in numbers to the west and south. Of and adaptive radiation has developed in Australia.

INTRODUCTION More than three decades have passed since portant of these collections are those made Tate (1952a) published on the bats of the by G. Neuhauser (1937-1938), and J. Rob- Cape York Peninsula. During that period, the erts (1949-1969) in northern Queensland; R. American Museum has obtained numerous F. Peterson (1959) in northern Queensland tropical Australian bat specimens, not only and northeastern Northern Territory; Rosen, from already well-represented northern Nelson, and Butler (1969) in northwestern Queensland but also from the Northern Ter- Northern Territory and Western Australia; ritory and Western Australia. The most im- and by W. H. Butler (1963-1966, 1973) in

' Curator, Department of Mammalogy, American Museum of Natural History.

Copyright © American Museum of Natural History 1984 ISSN 0003-0082 / Price $3.20 2 AMERICAN MUSEUM NOVITATES NO. 2778

Western Australia. With recent quickening personal encouragement while he visited here. interest in bat systematics among Australian Mr. John McKean ofthe Conservation Com- mammalogists (see Hall, 1981, and refer- mission of the Northern Territory, Australia ences therein), this seems a propitious time sent me copies of several papers which were to put on record this additional, largely un- useful in this study. I thank Dr. R. M. Timm published material. Specimens from the Mu- ofthe Field Museum ofNatural History, Chi- seum of Comparative Zoology (Cambridge), cago for permission to study specimens in his National Museum ofNatural History (Wash- care. I also thank the entire staff of the Di- ington), Field Museum of Natural History vision of Mammals and the Bird and Mam- (Chicago), and British Museum (Natural His- mal Laboratories, National Museum of Nat- tory-London) are referred to where they pro- ural History, Smithsonian Institution, vide additional taxonomic and distributional Washington, D.C. for access to their collec- information. tions and for numerous courtesies. Ms. M. E. I define tropical Australia in this paper as Rutzmoser of the Museum of Comparative Australia north of the tropic of Capricorn, Zoology, Cambridge, Massachusetts, has giv- thus approximately the northern two-thirds en me information about specimens and was ofQueensland, all but the southernmost por- very helpful during my visit. As always Mr. tion ofthe Northern Territory, and about the John Hill of the British Museum (Natural northern three-fifths of Western Australia. History) has been helpful in numerous ways, About 22 genera and 51 species of bats are both during my visits to London and in re- known from within this area. This is depau- sponse to my inquiries. Likewise, Miss Jean perate compared to a number of other trop- Ingles was extremely helpful during the same ical areas, but it can probably be explained visit. Finally, I acknowledge with thanks the by three factors, the peripheral position of help I received from Mrs. Laurel E. Schiller Australia, the small portion occupied by rain of the Department of Life Sciences, Indiana forest, and the absence ofhigh mountains and State University, Terre Haute. Not only did extensive highland areas, thus contrasting she make her own collections of bats from markedly with New Guinea to the north. Al- the Kimberley region freely available, but she though most species of Australian bats do also shared her ideas concerning species dis- occur in the tropical portion, extra-tropical tinctions and ecology of bats from that re- species and populations are discussed only as gion. they throw light on those to the north. The most important ecological division of SYSTEMATICS tropical Australia is the Great Dividing a All six bat families known from Australia Range. These mountains separate fairly occur in the tropical north and are discussed narrow strip of relatively mesic habitat (with in turn. All specimen measurements are in patches ofrain forest) along the Pacific Ocean otherwise stated. from the dryer areas farther west. Many millimeters unless species appear to be confined to this strip east of the Great Dividing Range (see maps in FAMILY PTEROPODIDAE Hall and Richards, 1979). West of the Great Five genera ofpteropodids are known from Dividing Range, there are a few relatively Australia and occur north of the Tropic of mesic areas in the extreme north of the Capricorn. With the exception of Pteropus, Northern Territory and Western Australia, each is known by only a single species, at least but conditions become increasingly arid to as represented by American Museum mate- the south. rial. GENUS Pteropus: All the Australian species occur in the tropical north. Four species are ACKNOWLEDGMENTS well known and are treated below. The fifth First of all I thank Dr. John Calaby of (P. brunneus) is of uncertain status. Still CSIRO in Lyneham, ACT, Australia, who known only from the type collected on Percy lent me critical specimens from the CSIRO Island off east central Queensland, it has collections and also provided considerable always constituted a problem. It is clearly 1984 KOOPMAN: AUSTRALIAN BATS 3 very distinct from any other Australian - Islands appear to be the most probable source Pteropus, and, unlike any ofthem, is a mem- area for brunneus (whether regarded as the ber ofthe subnigergroup (called by Andersen, single known representative of a population 1912, the hypomelanus group). Hall and or as based on an individual waif), I am un- Richards (1979, p. 13) suggest that the type willing to synonymize brunneus with admi- ofbrunneus was an accidental windblown waif ralitatum solomonis since it falls outside its from the Louisiades, which would make the known size range. I am particularly reluctant name a senior synonym of P. hypomelanus to do this because brunneus (Dobson, 1878) luteus. However, I have compared the type is an older name than solomonis (Thomas, of brunneus (at the British Museum) with 1 904-or-even admiralitatum (Thomas, 1894) Louisiade specimens of h. luteus and the two and would therefore become the name for taxa are clearly distinct, the characters An- both the subspecies and species. dersen (1912) gives appear to hold. The only Pteropus alecto: This is a widespread other members of the subniger group, which species, going all the way across tropical Aus- occur on islands bordering the Coral Sea (and tralia from eastern Queensland to Western therefore the only members which could rea- Australia. All Australian populations are re- sonably be blown to Percy Island) are ad- ferable to P. a. gouldi, which also barely gets miralitatum solomonis (Solomons), sanctae- into extreme southern New Guinea (Waith- crucis (Santa Cruz Islands), and o. ornatus man, 1979). Other subspecies occur on var- (New Caledonia). Of these, ornatus is much ious islands to the northwest of Australia. too large (forearm 145-152 vs. 118). Judged Tate (1 952a, p. 61 0) recorded specimens from by Troughton's (1930) description, sanctae- five localities on Cape York. The American crucis is the right size (forearm 1 12-121), but Museum also has a single specimen from the dorsal side of the tibia is hairless, unlike Normanton, northwestern Queensland, tak- brunneus, a character which Andersen (1912, en by R. F. Peterson in 1959. There are also pp. 90, 91) emphasized; although it should numerous specimens collected by W. H. But- be mentioned that the species P. admirali- ler (1964-1966, 1973) from five localities in tatum as presently constituted (Laurie and northern Western Australia (Kalumburu; Hill, 1954, p. 33) contains subspecies which Mitchell River; Parry Creek; Tunnel Creek; have tibiae either hairy (solomonis) or rela- Tambrey). The last locality is near the south- tively hairless (admiralitatum, colonus, gow- ern limit of the range in the west as given by ern). Of these members of the subniger group Ride (1970, p. 180), but Kitchener and Vick- around the Coral Sea, solomonis seems most er (1981) record the species from consider- like brunneus, agreeing in hairiness ofthe tib- ably farther south. ia, but was keyed out by Andersen (1912, p. Pteropus conspicillatus: This is primarily a 91) on the basis of shorter forearm (110 vs. New Guinea species. The nominate subspe- 1 8) and shorter maxillary tooth row (less vs. cies (to which Australian populations belong) more than 22). Since 1912, several published occupies at least the eastern half of New (Sanborn, 1931; Sanborn and Beecher, 1947; Guinea, another subspecies occurs in the Hill, 197 1) and unpublished (American Mu- northwest and also in the Moluccas. In Aus- seum series from Malaita) records of solo- tralia, P. conspicillatus is confined to north- monis have appeared and these have consid- eastern Queensland. Two other New Guinea erably extended the range of forearm Pteropus species (P. neohibernicus, and P. measurements, which now stands at 104-1 16, macrotis) occur in Western Province, just its upper end now closely approaching the north of Torres Strait, but do not reach Aus- forearm length (1 18) of the single specimen tralia. Tate (1952a, p. 610 recorded P. con- of brunneus. None of the above cited papers spicillatus from nine localities on the Cape give any maxillary toothrow measurements, York Peninsula, and the American Museum but I have checked the American Museum of Natural History has specimens from four material and all specimens have maxillary other localities in the same region: Babinda tooth row measurements ofconsiderably less Creek (H. C. Raven in 1921), Lake Barrine than 22 mm., much less than that of the type (G. Neuhauser in 1937); Coen (G. Neuhauser of brunneus (23.7). Although the Solomon in 1938), Cooktown (J. Roberts in 1949). 4 AMERICAN MUSEUM NOVITATES NO. 2778

Pteropus poliocephalus: This species is en- islands, Bismarcks and Solomons. Tate demic to eastern Australia, though closely re- (1952a, p. 611) recorded a single specimen lated to P. macrotis and P. pohlei of New from the northern part ofthe Cape York Pen- Guinea. It is chiefly a temperate species only insula (Portland Roads) but the American reaching the southeastern corner of tropical Museum of Natural History now also has Australia (Hall and Richards, 1979, p. 11). specimens from Shipton's Flat (collected by The American Museum of Natural History J. Roberts in 1967 and 1969), Mission Beach has no tropical Australian specimens, the (by J. L. McKean in 1968), and Captain Billy northernmost being a series from Mundub- Creek-"Heathlands" (by F. R. Allison in bera in southeastern Queensland. 1971). The first locality is also on the Cape Pteropus scapulatus: This is a widespread York Peninsula, but I have not been able to species almost endemic to Australia, though find either of the other two localities. there are relatives in the Bismarcks and Sol- Macroglossus minimus: Until recently this omons. However, it is now known from ex- species was called M. lagochilus, but it has treme southern New Guinea (Waithman, been recently shown (Lekagul and McNeely, 1979, p. 321) and has even been recorded as 1977) that the species should be called M. an accidental from New Zealand (Daniel, minimus. Its range extends from the Malay 1975). Tate (1952a, pp. 610, 611) reported Peninsula to the Solomons and south to this species from 10 localities in northeastern northern and northeastern Australia. With the Queensland, from the northern part of the reallocation of the name minimus, there are Cape York Peninsula to the southern edge of apparently two (m. minimus and m. lago- the tropical zone. Specimens were collected chilus) in the western part of the range (see by J. Roberts in 1949 and 1950 from Cook- Hill, 1983), but in the east, where it is the town, Flaggy, and Green Hills in the same only species of Macroglossus, three subspe- region. There are also numerous specimens cies were recognized by Andersen (1912). from Western Australia collected by W. H. These were microtus in the Solomons, nanus Butler in 1963, 1965, and 1973 from the fol- in the Bismarcks and New Guinea, and pyg- lowing localities: Ningbing; Parry Creek; maeus in the Murray Islands ofTorres Straits. Stockade Creek; Frazier Downs. All are in The genus was then unknown from the Aus- the extreme north, though the species is widely tralian mainland. Iredale and Troughton distributed in Western Australia. (1934, p. 92) listed another Australian local- Dobsonia moluccensis: The subspecies D. ity (Sunday Island, Western Australia) for m. magna has an extensive distribution in nanus. Tate (1952a, p. 612) identified spec- New Guinea and other subspecies occur in imens from the Cape York Peninsula as nan- the Moluccas and Bismarcks. Australian us but gave no reasons for ignoring pyg- populations, which are confined to the Cape maeus. McKean (1972a), noting a great deal York Peninsula, also belong to D. m. magna. of variation in this species in various eastern Tate (1952a, p. 611) recorded it from three parts ofits range, synonymized both microtus localities, all in the northern part ofthe Cape and pygmaeus with nanus. I have previously York Peninsula. The American Museum of (Koopman, 1982, p. 8) discussed this prob- Natural History has no other Australian ma- lem in New Guinea and am inclined to agree terial. with McKean's decision, though I am not at Nyctimene robinsoni: Ride (1970, pp. 182, all certain that only one subspecies is present 205) has recorded N. albiventer from three in the region. As I pointed out in (1982), some localities in eastern Australia. However, ofthe variation seems to be sexual (male skulls McKean (1972b) has shown that the New larger than females) though the small number South Wales record is in error and Winter ofskulls ofsexed adults makes this somewhat and Allison (1980) have shown that this is uncertain. In any case, ifwe recognize a single also almost certainly true for the Queensland subspecies in Australia, in Queensland this records. Ifso, then the only Australian species subspecies is confined to the Cape York Pen- of Nyctimene is N. robinsoni. This bat is en- insula east of the Great Dividing Range and demic to eastern Queensland, though the west of Queensland; it is only known from species has close relatives in the East Papuan very restricted areas in the far north ofNorth- 1 984 KOOPMAN: AUSTRALIAN BATS 5 ern Territory and Western Australia. Hall and georgianus as a new species (T. troughtoni). Richards (1979, p. 19) shows the range ofthe I shared this view until, after reading Mc- species extending south to Mackay, but the Kean and Price (1967), I reviewed all the Mackay specimen is actually referable to Sy- Australian Taphozous in the American Mu- conycteris australis (see Andersen, 1912, p. seum of Natural History with both Trough- 781). The only eastern Australian specimens ton (1925) and McKean and Price (1967) in of Macroglossus minimus that the American hand. I now agree with McKean and Price Museum of Natural History has are from the that georgianus is indeed conspecific with single locality (Seagren's Farm) on the Cape troughtoni rather than with australis. I be- York Peninsula on which Tate (1952, p. 612) lieve that the reason why Tate (and initially reported. However, from Western Australia, I) went wrong was because in northern there are three specimens from Mitchell Riv- Queensland where the two species are in close er collected by W. H. Butler in 1973. proximity and where virtually all of Tate's Syconycteris australis: This species, as now material came from, the two species are most constituted (see Koopman, 1982, pp. 8-10) readily distinguished by size, particularly if is centered in New Guinea but reaches the skins and skulls are studied. Taphozous Moluccas, Bismarcks, and Australia east of troughtoni was based on these large northern the Great Dividing Range. The nominate Queensland specimens. However, as Mc- subspecies, the only one in Australia, though Kean and Price (1967) pointed out, the species restricted to a narrow band along the coast becomes smaller as one goes west and south extends well beyond the southern edge of the and, as a result, the types of georgianus from tropics. Tate (1952a) did not record this southern Western Australia are much smaller species, but previously (1942b, p. 346) Tate than northern Queensland troughtoni, but recorded six skulls only from "northern agree well in size with australis. Recently, the Queensland." These were collected by G. problem has become much more complicat- Neuhauser, probably in 1938. The American ed with the description of two new species, Museum of Natural History has since ac- T. kapalgensis (McKean and Friend, 1979) quired two additional Syconycteris from from Northern Territory and T. hilli (Kitch- Queensland, one obtained by J. Nelson in ener, 1980) from Western Australia and 1962 at Nambour, another by J. Roberts at Northern Territory. Taphozous hilli was Shipton's Flat in 1969. The first locality is in compared with all previously described Aus- southeastern Queensland, a little to the south tralian species, but T. kapalgensis, though of the Tropic of Capricorn; the second in the compared with georgianus and the Lesser southern part of the Cape York Peninsula. Sunda leucopleurus (a subspecies of T. Ion- gimanus) was not compared with what seems to be the more closely related T. australis. FAMILY EMBALLONURIDAE The American Museum of Natural History A single genus, Taphozous, has in recent has no specimens of T. kapalgensis, but Dr. years been recognized from Australia, but I Calaby has kindly lent me three topotypes agree with Barghoorn (1977) in treating Sac- (including the holotype) from the CSIRO col- colaimus as a separate genus, based on dis- lections, which I have been able to compare tinctive skull characters and even a distinc- with American Museum specimens of the tive external one. Each genus has several three other species. I have found four char- Australian species and has its share of taxo- acters that are useful for distinguishing these nomic problems. four species and I here give the character states GENUS Taphozous: Until recently (e.g., in what could be regarded as a random access Ride, 1970, p. 170), two Australian species key. Gular sac, present (australis, kapalgen- were recognized in this genus (sensu stricto), sis, hilli) vs. absent (georgianus); skull size, T. australis and T. georgianus. Tate (1952a, small (hilli) vs. medium (australis, kapalgen- p. 607), however, had regarded the types of sis) vs. medium to large (georgianus); sphen- the two species in the British Museum as in- oid pits, narrow (hilli) vs. medium (georgi- distinguishable and described what earlier anus, kapalgensis) vs. broad (australis); authors (e.g., Troughton, 1925) had called anterior mandibular emargination, strong 6 AMERICAN MUSEUM NOVITATES NO. 2778

(australis, georgianus, hilli) vs. weak (kapal- tralis) and dry interior (georgianus) habitats. gensis). I believe that all four species are dis- From Northern Territory, the American Mu- tinct from one another and consider them so seum ofNatural History has specimens from in the following accounts. 18 mi. w. Wollogorang (collected by R. F. Taphozous australis: Except for one old Peterson in 1959) and from 3 mi. n. Kath- record from southeastern New Guinea, which erine (collected by P. Spaulding and W. Hos- is probably either accidental or erroneous, mer in 1960 and by A. J. Coventry in 1963). this species is endemic to Australia. Although The American Museum of Natural History the map in Hall and Richards (1979, p. 21) also has a great deal of material collected by shows T. australis extending across the Great Rosen, Nelson, and Butler in 1969 and by Dividing Range into northwestern Queens- W. H. Butler in 1963-1965 and 1973 from land (and apparently on into Northern Ter- the following localities in Western Australia: ritory), all American Museum material is from Kalumburu; King Edward River; Ningbing; northeastern Queensland (east of the Great Parry Creek; Manning Creek; Inglis Gap; Dividing Range) and I suspect that the more Tunnel Creek; Napier Downs; Mt. Anderson; western records represent another species Black Elvire River; 20 mi. n. Callowa; Pea- (probably georgianus). Tate (1952a, p. 609) wah Mundabullangana; Whim Creek; Bar- listed nine localities for australis, but, using row Island; Tambrey; Woodstock; 6 mi. ne. the characters that Troughton (1925, pp. 331, Yardie Homestead; Wittenoom Gorge; Wil- 332) gives, I would allocate the specimens lie Wollie Spring (ca. 20 mi. wnw. Poonda). from Quamby and Chillagoe to T. georgi- McKean and Price (1967) tentatively recog- anus. The American Museum ofNatural His- nized two subspecies, though they were evi- tory also has specimens of T. australis from dently uncertain as to where to draw the Cowie Bay (locality not found) collected by boundary. The type locality of T. g. trough- J. Roberts in 1950. toni is Mt. Isa in northwestern Queensland Taphozous georgianus: As outlined by and that of T. g. georgianus is King George McKean and Price (1967), Ride (1970), Par- Sound in southwestern Western Australia (but ker (1973), and Hall and Richards (1979) this I know of no recent records from anywhere species has an extensive distribution across near the alleged type locality). Although there northern Australia and extends some dis- is considerable variability, specimens in the tance south ofthe Tropic ofCapricorn. How- American Museum from Queensland and ever, the records on which this distribution Northern Territory tend to be large, thus was based were all published before the de- agreeing with troughtoni and those from scription of T. kapalgensis and T. hilli, so Western Australia, regardless oflatitude, tend many ofthese records should be reexamined. to be smaller, thus agreeing with typical geor- Tate (1952a, pp. 605, 608) recorded this gianus; I allocate specimens to the two sub- species (as troughtoni) from three localities species accordingly. It should be pointed out, in northern and eastern Queensland south of however, that the smallest T. georgianus I the Cape York Peninsula. However, I believe have seen are three specimens in the National the specimens he allocated to australis from Museum of Natural History from Oenpelli two localities (Quamby and Chillagoe) to be in the Northern Territory. These have fore- T. georgianus, just as McKean and Price arm lengths of 58-69 and condylocanine (1967, p. 107) suspected. The American Mu- lengths of 18.5-19.4. Condylocanine lengths seum of Natural History also has specimens of the five Northern Territory skulls of g. from two other localities in Queensland (Lap- troughtoni in the American Museum of Nat- pa Junction and Mungana) collected by R. F. ural History are 21.2-21.7. At first I thought Peterson in 1959. Specimens from Lappa that these Oenpelli skulls belonged to one of Junction, Chillagoe, and Mungana are inter- the other species of Australian Taphozous, esting since they lie only a little to the west but except for their small size, they agree with ofCooktown and Cardwell where Troughton T. georgianus in the characters enumerated (1925, p. 336) recorded T. australis. The two above. Obviously more work needs to be done species here appear to have parapatric dis- on geographical variation of T. georgianus, tributions corresponding to wet coastal (aus- particularly in the Northern Territory. 1984 KOOPMAN: AUSTRALIAN BATS 7

Taphozous kapalgensis: McKean and only from northeastern Queensland, east of Friend (1979) made no mention of T. aus- the Great Dividing Range, but McKean, tralis when describing this species, and the Friend, and Hertog (1980) have recorded it American Museum has no specimens. How- from the same limited area of the Northern ever, as indicated above; after studying bor- Territory from which Taphozous kapalgensis rowed specimens of this species, it seems to was described. Tate (1952a) recorded a single me to be perfectly distinct. specimen from Babinda Creek in northern However, it appears to be confined to what Queensland. There are no other Australian seems to be an unusually mesic area ofnorth- specimens in the American Museum of Nat- western northern Territory. The dryer area ural History. between this limited area and the mesic range Saccotaimus flaviventris: This is a wide- ofaustralis is occupied only by the more arid- spread endemic Australian bat, occurring both adapted T. georgianus. in the tropics and well to the south of the Taphozous hilli: This species was de- tropic of Capricorn. The map in Hall and scribed by Kitchener (1980) from a number Richards (1979, p. 21) indicates the species' oflocalities in Western Australia but was also occurrence throughout the Cape York Pen- recorded from Tennant Creek in Northern insula, but they show no actual records from Territory from which the American Museum the wet eastern coast and there are no spec- of Natural History also has a specimen col- imens in the American Museum of Natural lected by W. Hosmer in 1960. Taphozous History from there. (A specimen was referred hilli appears to be a perfectly good species, to by Winter and Allison, 1980, p. 34.) Tate particularly well differentiated from the sym- (1952a, p. 606) recorded S. flaviventris from patric T. georgianus. The species is much less two localities in northern Queensland south common than georgianus, since among all of the Cape York Peninsula. The only other the Western Australian specimens in the specimens of this species in the American American Museum of Natural History none Museum are from northernmost Western are T. hilli. From the distribution given by Australia (Ningbing and Inglis Gap), collect- Kitchener it is the most arid adapted of the ed by W.-H. Butler in 1965. Australian Taphozous and straddles the Tropic of Capricorn. GENUS Saccolaimus: There are three Aus- FAMILY MEGADERMATIDAE tralian species of this genus, all occurring in There is only one Australian member of the tropical portion and two of which are this small, widespread Old World tropical confined to it. There appear to be no taxo- family. The genus is endemic to Australia. nomic problems involving this genus within Macroderma gigas: This species occurs Australia. across the whole of tropical Australia and Saccolaimus mixtus: This is chiefly a New also somewhat south of the Tropic of Capri- Guinea species, but Tate (1952a, p. 606) re- corn, particularly in Western Australia. Tate corded it from a single locality in the northern (1952a) recorded this species from one cen- part ofthe Cape York Peninsula. There seem tral Queensland locality. The American Mu- to be no other published Australian records. seum also has a specimen from the north- Saccolaimus saccolaimus: In Australia un- eastern Northern Territory (18 mi. w. til recently this species was called S. nudiclu- Wollogorang) collected by R. F. Peterson in niatus. I agree with Goodwin (1979), how- 1959. There are a number of specimens from ever, in treating nudicluniatus as a subspecies northern Western Australia collected by W. of S. saccolaimus. The species has a wide H. Butler in 1964, 1965, and 1973 (Kalum- range from India to the Solomon Islands. buru; Koolan Island; Tunnel Creek; 20 mi. While the pattern of geographical variation s. Marble Bar). Two subspecies have been in this species is poorly understood, New recognized, M. g. gigas originally described Guinea and Australian populations, together from Mt. Margaret on Wilson's River in with the single Solomon Island record, may southwestern Queensland and M. g. saturata be included in S. s. nudicluniatus. Until re- from Kalumburu in northern Western Aus- cently, the species in Australia was known tralia. The two subspecies were distinguished 8 AMERICAN MUSEUM NOVITATES NO. 2778

(Douglas, 1962) entirely on the basis ofcolor, to the south of it, and I have assumed that among American Museum specimens only all are in that area. In 1958 and 1959, J. L. the specimens which Tate reported (from near Harrison collected specimens from Innisfail. Rockhampton) being of the dark (saturata) Finally, in 1959, R. F. Peterson collected this type. One of the light-colored specimens is a species from Chillagoe and Lappa Junction topotype of saturata. It has been in alcohol at the base of the Cape York Peninsula and since 1964, but is much lighter than Rock- from Lyndhurst Station to the south of it. hampton specimens which have been in al- This last locality is unusually far inland, ac- cohol since 1948 and agrees well with a Koo- tually being on the west side of the Great lan Island specimen that has been in alcohol Dividing Range. Two Australian subspecies since 1973. Although it is indeed likely that have, since 1933, been recognized in this desert populations tend to be lighter in color species, R. m. megaphyllus described from than those from more mesic areas, I cannot southern New South Wales, and R. m. ignifer see these two color phases as distinct sub- from Coen in northern Queensland. The orig- species. Since there seem to be no other char- inal description of ignifer (Allen, 1933) dis- acters that correlate with the color difference tinguished it entirely on the basis of color. I cannot see the utility of recognizing sub- McKean and Price (1967) have shown quite species. convincingly that the color characters do not hold but felt that they could distinguish the two subspecies on the basis of size (as deter- FAMILY RHINOLOPHIDAE mined by forearm length). In the course of Since I include the Hipposiderinae in this determining the pattern ofgeographical vari- family, there are three Australian genera. ation in the New Guinea region (Koopman, Rhinonycteris is monotypic, there are two 1982), I compared the two alleged subspecies well-marked species ofRhinolophus, but there in Australia and found that ifanything R. m. are five Australian species of Hipposideros. megaphyllus was smaller (based on skull size) Rhinolophus megaphyllus: This, by far the than R. m. ignifer (contra McKean and Price). commoner of the two Australian species, oc- My problem, however, was uncertainty as to curs from the Bismarck archipelago through whether I had any genuine R. m. megaphyl- eastern New Guinea and its islands and East- lus. McKean and Price give the range of R. ern Australia (for the most part east of the m. ignifer as Bramston Beach north and of Great Dividing Range) to Victoria. There are m. megaphyllus as Brookfield south. Bram- three quite distinct subspecies in the New ston Beach is near Innisfail at the base of the Guinea region (Koopman, 1982), but Aus- Cape York Peninsula but I have been unable tralian populations are much more uniform to find Brookfield. Hall and Richards (1979), and (as shown below) probably only a single however, place the boundary at about subspecies can be recognized. Tate (1952a) Townsville. I have therefore done a more recorded the species from 19 localities in thorough analysis using, besides the speci- northern and central Queensland, and the mens from north of the Tropic of Capricorn American Museum of Natural History has which I have listed, also a few from southern obtained a great deal more material. With a Queensland, northern New South Wales, and few exceptions it likewise comes from north Victoria. These temperate zone specimens are of the Tropic of Capricorn. First, there is a all alcoholics from which I have extracted great deal of material collected by J. Roberts some representative skulls. I have grouped from 1948 to 1951 which add the following the specimens geographically as follows from localities: Collingwood; Home Rule; Grass north to south: Area 1 (Portland Roads, Iron Tree; Helenvale; Boiling Springs, Mt. Pov- Range, Pascoe River, Wenlock); Area 2 (Coen, erty; Stucky's Gap; Wyalla; Green Hills; Ay- Nesbit River, Peach River); Area 3 (Cook- ton; Middle Normanby; Mt. Amos, Phoeni- town area, Laura, Mt. Finnigan, Shipton's cian; Endeavor Bridge; Mc Ivor; Mt. Cook. Flat); Area 4 (Cairns, Jullattan-Mossman I have not been able to find most of these Road, Mt. Carbine); Area 5 (Walter Hill localities but those that I have are all in the Range, Irvine Bank, Lappa Junction, Chil- general vicinity ofCooktown or immediately lagoe, Innisfail); Area 6 (Mt. Etna, Cromarty, 1 984 KOOPMAN: AUSTRALIAN BATS 9

Lindhurst Station); Area 7 (southern Queens- confined to the tropical portion. Nine species land, New South Wales, Victoria). Areas are known from New Guinea, four being 1-5 are regarded as typical ignifer, Area 6 as shared. megaphyllus but perhaps showing intergra- Hipposideros ater: This is a widespread dation with ignifer, Area 7 as typical mega- species of the Indo-Malayan and Australian phyllus. Unfortunately, the vast majority of regions. Two subspecies, albanensis (from the specimens are from Areas 1-3; the areas Cape York) and gilberti (from the northern farther south (some ofwhich ofnecessity cov- part of the Northern Territory) have been er great distances) are much more poorly rep- described from Australia. Both have been resented. In the following, the ranges of fore- synonymized with H. a. aruensis, the New arm and condylocanine lengths are given for Guinea subspecies (Hill, 1963, p. 33; Mc- each area, forearm first and condylocanine Kean and Price, 1967, pp. 110, 111). I have second: Area 1 (44-48, 16.9-17.8); Area 2 seen no specimens from the Aru Islands (the (44-48, 16.8-17.5); Area 3 (44-48, 16.8- type locality ofaruensis), but specimens from 17.4); Area 4 (44-48, 17.3-17.8); Area 5 (45- the Western Province of Papua should be 47, 16.9-17.6); Area 6 (43-47, 16.6-17.0); quite similar. Specimens from this latter area Area 7 (47-49, 17.1-18.0). The discrepancy have been compared with those from Cape between my earlier (Koopman, 1982) and York and I can see no consistent differences. McKean and Price's (1967) conclusions are I would agree that albanensis should be re- explained. My "typical megaphyllus" con- tained in the synonymy of aruensis, but I am sisted only of Area 6 and it is apparent that not at all certain whether I can agree with this central Queensland sample tends to be McKean and Price in synonymizing gilberti smaller than populations either to the north with aruensis. The western subspecies was or to the south, particularly in condylocanine distinguished on both size and color. Hill length. Most if not all of McKean and Price's (1963) believed that the two subspecies could material of m. megaphyllus is from my Area not be separated on size and his histograms 7 and does tend to run somewhat larger than offorearm measurements (p. 32) support this. their ignifer from Areas 1-5. Evidently there McKean and Price (1967) believe that the is some geographical variation along the al- variation of their northern Queensland ma- most 2000 mi. range of the species in Aus- terial overlapped gilberti in both size and col- tralia. However, the magnitude of this inter- or and again, at least as far as size is con- populational variation is small in relation to cerned, their histograms of forearm intrapopulational variation, and morpholog- measurements bear this out. I find this puz- ical ranges broadly overlap. I therefore see zling since the specimens I have seen do not little utility in recognizing two subspecies and show this. A series of 11 Queensland fore- would synonymize ignifer with R. m. mega- arms (from three localities, including Bram- phyllus. ston Beach where McKean and Price's spec- Rhinolophus philippinensis: This species is imens came from) have lengths of38-4 1. Four known from several areas from the northern skulls, including one from the smallest ofthese Phillipines to northeastern Australia. In the have condylocanine lengths of 13.5-14.2. In latter country, its known distribution is re- contrast, seven forearms from the Northern stricted to a rather small area in the south- Territory (including specimens in the Na- eastern Cape York Peninsula, where it is rep- tional Museum of Natural History and resented by the endemic R. p. robertsi. This CSIRO, one ofthem the type ofgilberti) mea- was as yet unknown to Tate (1952a), but was sure 35-38 and available skulls have described a few months later (Tate, 1952b). condylocanine lengths of 13.2 to 14.1. So far, Tate only mentions the type locality (Mt. this shows considerable overlap and would Amos), but one of the original series is from support McKean and Price. Western Austra- Helenvale, which is probably also in the lian specimens I have seen, including those Cooktown region. The American Museum has at the Field Museum ofNatural History, show no other Australian material of this species. greater difference from Queensland speci- GENUS Hipposideros: Five species of this mens. Their forearm lengths are 35-37 and genus are known from Australia and all are condylocanine lengths are 12.2-12.6. I can- 10 AMERICAN MUSEUM NOVITATES NO. 2778 notjudge the color difference since most spec- and 1967 and from Bramston Beach collected imens in the American Museum of Natural by J. L. Harrison, all in Queensland. There History are or have been in alcohol for a are specimens of H. a. gilberti from near significant length of time (though the few I Katherine in the Northern Territory (col- have seen do indicate a color difference). The lected by W. Hosmer in 1960) and from Koo- size difference (at least as far as Western Aus- lan Island in Western Australia (collected by tralian specimens are concerned) is great W. H. Butler in 1965). In Western Australia, enough so that I considered the possibility the species is only known from the far north that gilberti is actually a representative ofthe (Kimberley). Indo-Malayan H. cineraceus. Hill (1963, pp. Hipposideros cervinus: This is the species 24, 35-36) distinguishes ater from cineraceus which until recently was called H. galeritus. by two characters, the presence of a dorsal However, Jenkins and Hill (1981) have dem- process on the zygomatic arch and the extru- onstrated sympatry between two members of sion (more or less) of the anterior upper pre- this complex in Borneo and have therefore molar from the toothrow. The former char- separated the more eastern species as H. cer- acter seems to hold, though in many skulls vinus, which still has a wide range extending the delicate zygomatic arch is broken at that from Sumatra to the New Hebrides. Virtually point, but not the latter character. Comparing the entire range in the Australian region is cineraceus skulls from Malaya with ater from included in H. c. cervinus which, however, in Australia and New Guinea, I see a great deal Australia is confined to the Cape York Pen- of variation in the placement of the anterior insula and adjacent islands. Tate (1952a) re- upper premolar, but no clear-cut difference corded this species from five localities in the between the two species. All gilberti skulls Cape York region. The American Museum available, which have the zygomatic arch in- of Natural History has no other specimens tact, do have a small but definite dorsal pro- from Australia. cess and should therefore be referred to H. Hipposideros semoni: Tate (1952a) listed ater. This then is the problem. If I knew only this as a subspecies of the New Guinea mus- the Queensland and Northern Territory pop- cinus, but Hill (1963, pp. 82, 86) separates ulations, I would be inclined to follow them on the specific level and I am inclined McKean and Price (1967) in synonymizing to agree with him. The main distribution is gilberti with H. a. aruensis. On the other hand, in New Guinea where it has a wide range. In if gilberti had never been described, I would Australia, however, it is confined to northern have no hesitation in recognizing the Western Queensland east ofthe Great Dividing Range. Australian populations as a different subspe- Tate (1952a) recorded it from four localities cies from those of Queensland and consid- and the American Museum has no material ering the Northern Territory populations as from any others. intergrades. The Northern Territory and Hipposideros stenotis: This species is en- Western Australian populations are probably demic to Australia but is closely related to continuous, whereas the Queensland popu- and allopatric with H. semoni so the two could lations (not known from west of the Great be included in the same superspecies. (I agree Dividing Range) are geographically separate. with Hill, 1963, that they should be retained Rather than synonymizing gilberti with as separate species.) The species occurs across aruensis and then describing the Western the dryer portions of tropical Australia from Australian populations as a separate subspe- northwestern Queensland to northeastern cies, I tentatively associate these with gilberti, Western Australia. Tate (1952a) did not re- recognizing that the typotypical population cord it but the American Museum of Natural is in the intergrade area between H. ater gil- History has four specimens collected by R. berti and H. ater aruensis. Tate (1952a) re- F. Peterson from 18 mi. w. Wollogorang in corded H. ater (under the name of H. bicolor the northeastern Northern Territory in 1959 albanensis) only from Lockerbie at the tip of which are the basis for one of Parker's (1973, Cape York. The American Museum also has p. 35) records. specimens of H. a. aruensis from Shipton's Hipposideros diadema: This is another Flat collected by J. Roberts in 1950, 1952, widespread species ranging from southeast- 1984 KOOPMAN: AUSTRALIAN BATS I1I ern Asia to the Solomon Islands. Two Aus- GENUS Myotis: Two species ofMyotis have tralian subspecies have been described, H. d. been recorded from Australia, the well-known reginae from the Cape York Peninsula and adversus (treated below) and australis (known H. d. inornatus (McKean, 1970a; McKean only by the type specimen that was supposed and Hertog, 1979) from the northwestern to have been collected near Sydney, New Northern Territory. The American Museum South Wales). The latter (unlike adversus) is has no specimens of the latter subspecies but a member of the subgenus Selysius. Except Dr. John Calaby has very kindly lent me two for insularum ofSamoa (likewise known only paratypes. Comparison of these with d. re- from the type and ofdubious status), the only ginae shows that whereas reginae is one of species of M. (Selysius) known from Celebes the larger subspecies, inornatus is one of the and the Lesser Sundas eastward is M. mu- smallest (condylobasal length 26.4-28.7 vs. ricola (formerly erroneously included in M. 23.9). I have previously (Koopman, 1982, p. mystacinus, but see Koopman, 1982, p. 17; 16) discussed the problem of the patchwork Hill, 1983, also recognizes ater as a distinct of large and small subspecies in this species. species in Celebes and the Mentawei Islands). Although its subspecific taxonomy is chaotic, This species probably does not occur east of the two Australian subspecies seem well dif- the Moluccas (see Koopman, 1982, p. 17) ferentiated from each other. Neither has any which makes the status of australis (and in- close affinity with the still larger H. d. dia- sularum) highly uncertain. I previously be- dema of the Lesser Sundas. Tate (1952a) re- lieved that both were based on mislabeled corded H. diadema reginae from three lo- material. However, the Field Museum has a calities on the Cape York Peninsula. The single specimen (FMNH 120121) collected American Museum of Natural History also at Geikie Gorge (Kimberley region), 3 De- has specimens from Shipton's Flat (collected cember 1976, by L. E. Schiller. The specimen by J. Roberts in 1951 and S. Breeden in 1962) is a female, in alcohol with the skull removed, and from Chillagoe (collected by R. F. Pe- with the following measurements: forearm terson in 1959). (35), hindfoot (9), ear from notch (13), con- Rhinonycteris aurantius: This (for emend- dylobasal length (11.9), maxillary tooth row ed spelling see Hill, 1982) is one of the only (4.5), width across last molars (4.8). These two endemic Australian genera (the other skull measurements compare with an unusu- being Macroderma). Confined to tropical ally small specimen of M. m. muricola from Australia but, like Hipposideros stenotis, R. Bali (AMNH 107523), the most similar spec- aurantius is absent from the east coast yet imen I could find, with the following: con- avoids the desert. Like H. stenotis, R. auran- dylobasal length (12.6), maxillary tooth row tius ranges from northwestern Queensland to (5.0), width across last molars (5.3). The two northeastern Western Australia, but also oc- skulls also differ in the less-reduced middle curs farther to the southwest in Western Aus- upper premolar and relatively more inflated tralia (Pilbara). Tate (1952a) did not record braincase of the Geikie Gorge skull as com- this species, but the American Museum of pared with that ofthe Bali skull. This is shown Natural History has specimens from near by the mastoid width, which in spite of the Katherine in the Northern Territory collected considerable difference in skull length is only by P. E. Aitken in 1966 and from Koolan slightly smaller (6.8 vs. 6.9). A comparison Island in Western Australia collected by W. of the Geikie Gorge specimen with infor- H. Butler in 1965. mation concerning the type specimen of aus- tralis (Dobson, 1978, pp. 317, 318; Tate, FAMILY VESPERTILIONIDAE 1941a, p. 555; Husson in McKean, 1970b) Nine genera of this cosmopolitan family is obviously in order. (Hill, 1983, regards have been recorded from Australia and all australis as a possible synonym of M. ater.) are known from north of the Tropic of Cap- In qualitative characters (post-calcaneal lobe ricorn. All but Murina and Kerivoula have on the calcar, tragus shape, size and position more than a single Australian species and of the middle upper premolar) the two are must be discussed as genera before individual quite similar. However, in the few measure- species are taken up. ments that can be taken on the type of aus- 12 AMERICAN MUSEUM NOVITATES NO. 2778 tralis (the skull is in fragments) the specimen Pipistrellus tenuis: I have previously is clearly considerably larger than that from (Koopman, 1973) discussed my concept of Geikie Gorge: forearm (39 vs. 35), maxillary this species and indicated its distribution in tooth row (5.8 vs. 4.5). I am therefore reluc- a general way. It has an extensive range from tant to refer the Geikie Gorge specimen to southeastern Asia to the New Hebrides, but M. australis, though, if the type of the latter there are only a few records from Australia. did come from New South Wales, the Geikie Tate (1 952a, p. 598) recorded the species (un- Gorge specimen could represent a small der the name papuanus) from three localities, northwestern geographical variant. Without all in the northern part of the Cape York more material I hesitate to describe it as new. Peninsula. Hall and Richards (1979, p. 41) The specimen was identified in the field as show the species extending south to Towns- Eptesicus douglasi, so Eptesicus from north- ville. The American Museum ofNatural His- ern Western Australia (and elsewhere) should tory also has a few specimens collected at be scrutinized with care. In this paper the Cape Bossut in northern Western Australia Geikie Gorge specimen is henceforth referred by Nelson, Butler, and Rosen in 1969. to as Myotis sp. (near australis). McKean and Price (1979) mention another Myotis adversus: This species (a member northern Western Australian locality and give of the subgenus Leuconoe) has an extensive measurements for another Cape York spec- range from Sumatra to the New Hebrides and imen. Most importantly, they record P. ten- its subspecies, M. a. macropus (confined to uis from two localities in the Northern Ter- Australia) likewise has a broad distribution ritory. One (Port Essington) is on the northern from extreme northeastern Western Austra- coast, approximately equidistant from the lia around the northern and eastern coasts to Cape York and Western Australian records. southeastern South Australia but apparently The other (Thring Creek) according to the nowhere extending far inland. Tate (1952a) coordinates they give should be in the arid recorded it from Cairns on the Cape York southern part of the Northern Territory. Dr. Peninsula. The only other American Mu- John Calaby, however, assures me that this seum material of this species from Australia is in error and that Thring Creek is actually is a series collected by W. H. Butler in 1973 just south of Van Diemens Gulf (12°14'S, at Mitchell River in the Kimberley region of 131°54'E) and therefore also in coastal Western Australia. Northern Territory. The skulls of the Amer- GENUS Pipistrellus: Three species have been ican Museum specimens fall into two rather recorded in Australia. By far the most dis- different morphological groups, according to tinctive is the large P. tasmaniensis, one of whether they come from Queensland (given the few species of bats which appears to be as the first range of measurements) or West- confined to the southern halfofAustralia. Its ern Australia (as the second range as given northernmost known occurrence is in ex- below). I see no difference in the condylobasal treme southeastern Queensland and all lengths (10.9-11.1 vs. 10.7-11.2), but for American Museum specimens are from maxillary tooth row lengths (3.9-4.1 vs. 3.6- southwestern Western Australia. Pipistrellus 3.7) and widths across the last molars (5.1- tenuis is now fairly well known from tropical 5.2 vs. 4.6-4.9), there is clearly a difference. Australia and is discussed below. The third Thus, it would appear that the Western Aus- species, P. javanicus, has an extensive dis- tralian specimens differ from those in tribution from southeastern Siberia to Java, Queensland by their relatively smaller rostra Celebes, and perhaps the Aru Islands (see Hill, (fig. 1). McKean and Price (1978) give skull 1983), but the only Australian record is pro- measurements for one specimen from Cape vided by two specimens in the British Mu- York and two from Thring Creek (the Port seum without definite locality. I have ex- Essington specimen unfortunately is repre- amined these specimens and can vouch for sented only by external measurements). Mea- their identity as P. javanicus. However, the surements for their Cape York specimen agree status of this species in Australia (assuming well with those in the American Museum. that the two specimens are not mislabeled as Dr. John Calaby has very kindly lent me the to locality) is not clear. two Thring Creek skulls from the CSIRO col- 1984 KOOPMAN: AUSTRALIAN BATS 13

0 .,.

...I g..r

'k., "-J %-qP t. 4,;If 'S = f '06 am'' Vk--

N:

A, 11" 14

0 t1 4... 4 1 9f. r, S. .-- I .. -S-vs Tjw 1%. .s

FIG. 1. Skulls of Pipistrellus tenuis, dorsal (above) and ventral (below) views from left to right: AMNH 237820 (adult male from Taibesse, Timor); AMNH 216135 (adult male from Cape Bossut, Western Australia, holotype of westralis); AMNH 216136 (adult female from Cape Bossut, Western Australia); AMNH 154654 (adult male from Brown's Creek, Queensland). Approximately x 3. lections and my measurements for both are: highlands (Bonleo) to which the name P. t. condylobasal length (10.8), maxillary tooth sewelanus is applicable. I believe that the length (3.7), width across last molars (5.0). Western Australian population is most closely As with those from Queensland and Western related to P. t. sewelanus, but it is much Australia, there is no difference in condylo- smaller and therefore should not be included basal length, but in maxillary tooth row length, in this subspecies. I believe that the best way they clearly agree with those from Western to express the phenetic and phylogenetic re- Australia, whereas in width across the last lationships of this form is to describe it as molars, they are intermediate. Therefore the new. Queensland specimens (fig. 1) are clearly re- ferable to P. t. papuanus and have New Pipistrellus tenuis westralis, Guinea affinities. Those from Northern Ter- new subspecies ritory are probably intergrades but agree bet- ter with those from Western Australia. These HOLOTYPE: AMNH 216135, an adult male in turn, however, most closely resemble in collected by Gareth J. Nelson, W. H. Butler, their skull proportions specimens from the and Donn E. Rosen on 15-16 April 1969 at lowlands of Timor (Dili and Djamplong) as Cape Bossut (ca. 18°40'S, 12 1°30'E), Western represented in the American Museum ofNat- Australia. The holotype consists of an entire ural History, the Rijksmuseum van Naturr- specimen preserved in alcohol with the skull lijke Historie in Leiden, and the Zoologisches extracted and cleaned. Museum der Humboldt Universitat in Ber- DIAGNOSIS: A small subspecies of Pipis- lin. Finally, these intergrade with the sub- trellus tenuis (forearm length 27-30; condy- species in Celebes, Lombok, and the Timor lobasal length 10.7-11.2), thus being smaller 14 AMERICAN MUSEUM NOVITATES NO. 2778 than P. t. collinus or any but intergrade (low- 1966.) Furthermore, none ofthe southeastern land Timor) populations of P. t. sewelanus. Asian Eptesicus seem to be particularly closely It differs from P. t. murrayi, P. t. nitidus, P. related to any of the Australian Eptesicus, t. sewelanus, P. t. angulatus, and P. t. pon- which can be viewed as constituting an en- celeti in its shorter (3.6-3.7) maxillary tooth demic radiation. While there have been sev- row. It differs from the vast majority of P. t. eral subsequent papers which have dealt with papuanus (including those from Cape York) these species (Kitchener and Halse, 1978; in its combination of a shorter maxillary Carpenter, McKean, and Richards, 1978; Hall toothrow with a narrower width across the and Richards, 1979), the distinctions are none last molars (4.6-5.0). I know of no way to too clear and indeed seem to be rather subtle. distinguish P. t. westralis from the geograph- Fortunately, the American Museum has at ically distant P. t. tenuis. Specimens from the least some material of each of these five lowlands of Timor (forearm length, 27-29; species and this has been of great help in condylobasal length, 10.6-11.4; maxillary evaluating their differences. Unfortunately, tooth row length, 3.8-4.2; width across last the American Museum has no Eptesicus from molars, 4.9-5.2), while probably best referred southeastern New South Wales, the one area to P. t. sewelanus, show intergradation with where apparently four of the species occur in P. t. westralis. Specimens from the Northern close proximity. I have been forced to use Territory (see measurements above) are widely allopatric specimens which I am as- probably best referred to P. t. westralis but suming represent these four species in south- show intergradation with P. t. papuanus. eastern New South Wales. As a result I have ETYMOLOGY: The name refers to Western tried to evaluate them using the following Australia. skulls: one vulturnus from Victoria; eight reg- LOCALITY RECORDS: I have only seen the ulus from southwestern Western Australia; type and three topotypes from near the type one sagittula from Tasmania; 13 pumilus locality: AMNH 216134-216137, alcoholics, pumilus from northeastern Queensland skulls extracted and cleaned for 216135 and (Cooktown area). Besides size, the principal 216136. Almost certainly the specimen from skull character used by McKean, Richards, Roebuck Bay (referred to by McKean and and Price (1978) and Carpenter, McKean, and Price, 1978), however, ca. 60 mi. ne. Cape Richards (1978) is the slope of the forehead, Bossut is referable to this subspecies, as are distinguishing a "cave dweller" (with an those from the Pender area of the Dampier abrupt forehead) from a "forest dweller" (with Peninsula (see McKenzie, 1983, p. 50). I a gradually sloping forehead). Eptesicus would also refer specimens from Smith Point pumilus (along with the Western Australian and Thring Creek in the northern part of the E. douglasi) has the "cave dweller" skull, Northern Territory to P. t. westralis, but they whereas sagittula, regulus, and vulturnus have show intergradation with P. t. papuanus. "forest dweller" skulls. Comparing actual GENUS Eptesicus: Prior to 1976, only one skulls of the four species, I find the slope species of Eptesicus had generally been rec- character quite subtle, but nevertheless see ognized in Australia (e.g., Ride, 1970). How- fairly marked differences in skull shape be- ever, in the next two years (Kitchener, 1976; tween these two types. To me, however, this McKean, Richards, and Price, 1978), two new is better shown by rostral proportions, best species were described and two additional seen in palatal view. Eptesicus vulturnus, E. named forms that had been put in the species regulus, and (to a lesser extent) E. sagittula synonymy of E. pumilus were resurrected as have long narrow rostra, whereas E. pumilus full species. The five species now recognized (and also E. douglasi) have shorter, broader are all closely related and confined to Aus- rostra. This is probably associated with dif- tralia (including Tasmania and Lord Howe ferences in feeding habits, the "forest" type Island) and are widely separated geographi- being better equipped for feeding on soft in- cally from the nearest other Eptesicus in sects (such as moths) and the "cave" type for southern Thailand, E. demissus. (The Ma- hard insects (such as beetles). (See Freeman, layan species that Kitchener (1976) refers to 1979, for a discussion of some of the asso- has since been transferred to Philetor, see Hill, ciated adaptations.) Within the "forest" group 1 984 KOOPMAN: AUSTRALIAN BATS 15 the single vulturnus skull is smaller than any seum has a number of specimens collected of my regulus skulls, whereas the sagittula from the vicinity of Katherine by Spalding skull is near the maximum size for the regulus and Hosmer (1960), J. Owen (1962), and A. series. Outside of southeastern mainland J. Coventry (1963). W. H. Butler collected Australia, there are rarely more than two numerous specimens from tropical Western sympatric species and in each case they are Australia at the following localities: Kalum- clearly different in size. Thus in Tasmania, buru (1965); Parry Creek (1965); Koolan Is- the large sagittula occurs with the small vul- land (1973); King Edward River (1973); In- turnus, in southern Western Australia, the glis Gap (1965); Napier Range (1973); Tunnel larger regulus occurs fairly close to, and per- Creek (1973); 20 mi. north ofCallowa (1973); haps in contact with, the small pumilus cauri- Peawah, Mundabullangana (1963); Whim nus, though Kitchener and Vicker (1981) also Creek (1963); Woodstock Station (1963); list the small vulturnus. The relationship be- Tambrey (1964, 1973); Wittenoom Gorge tween the large douglasi and the small pum- (1973); Nodswell Creek (1963); Yardie ilus caurinus in northern Western Australia Homestead (1963); Barrow Island (1964, and Northern Territory is discussed below. 1966, 1967); Hermite Island, Montebello Ofthe five Australian species, vulturnus, reg- Group (1966). Two subspecies of E. pumilus ulus, and sagittula are all confined to the are currently recognized (McKean, Richards, southern part, south of the Tropic of Capri- and Price, 1978), p. pumilus in eastern corn, though vulturnus has recently been re- Queensland (and New South Wales), p. corded (Thompson, 1982) from just south of caurinus in western Queensland, Northern it in the Northern Territory. Only pumilus Territory, and Western Australia. These are and douglasi require separate treatment. distinguished on size and bacular form. Eptesicus pumilus: This is by far the most Judged by the distribution map in McKean, widespread of all the Australian Eptesicus, Richards, and Price (1978), only the Quamby extending over the entire continent except for and Mount Isa specimens among the Queens- the southern edges. Even though it reaches land material at the American Museum are the tip of Cape York near Somerset (AMNH referable to E. p. caurinus. All others are E. 155012), Eptesicus has never been recorded p. pumilus and I have listed them according- in New Guinea. Neither McKean (1972) nor ly. I have not examined any bacula, but I do Waithman (1979) record Eptesicus from there note that among these eastern Queensland and I have checked all specimens identified specimens referred to p. pumilus there is con- as Pipistrellus (the genus most likely to be siderable variation in size, some of which is confused with Eptesicus) from Western prov- geographical. The largest specimens (which I ince (Papua New Guinea) and southeastern have used in the above-mentioned compar- Irian Jaya (the parts of New Guinea closest isons with other species) are from the Cook- to Cape York) and all are indeed Pipistrellus town area (condylobasal length 11.9-12.7). tenuis. Over most of tropical Australia, E. Specimens from localities farther south (and pumilus is the only Eptesicus. Tate (1952a, mostly farther inland (Koombooloomba p. 599) recorded this species from eight lo- Creek, Mungana, Pentland) are clearly small- calities (the Pentland and Quamby specimens er (condylobasal length 11.2-11.9) and closely having been collected by G. Neuhauser in resemble specimens from Mount Isa and 1948). However, Tate omitted another lo- Quamby referred to E. p. caurinus (condy- cality (Lockerbie) which was worked by Tate lobasal length 11.2-11.7). This suggests in- and Van Deusen in 1948, though this is the tergradation between the two, but it must also northernmost locality for Eptesicus in Aus- be pointed out that the few specimens from tralia. In 1950 and 1951, J. Roberts collected north of Cooktown (Iron Range, Lockerbie) specimens from Helenvale and China Camp, are also small (condylobasal length 11.0-11.7) two localities south of Cooktown. Finally, in and it is difficult on geographical grounds to 1959 R. F. Peterson collected specimens at interpret these as intergrades. In any case, Lappa Junction, Mungana, and Koomboo- most of the Cooktown specimens were evi- loomba Creek (30 mi. s. Ravenshoe). From dently not available to Tate (1952a) when he the Northern Territory, the American Mu- wrote his paper, therefore it is not surprising 16 AMERICAN MUSEUM NOVITATES NO. 2778 that he referred all northern Queensland ma- from Yardie Homestead. In view of the fact terial to E. p. caurinus. The specimens from the Kitchener (1976) referred Yardie Home- Northern Territory (Katherine, together with stead specimens to E. pumilus and because specimens at the National Museum of Nat- Pilbara specimens (and those from farther ural History from Groote Eylandt, Oenpelli, south) seem to show intergradation, I am al- and 170 mi. e. ofDarwin) are typical caurinus locating specimens from Northwest Cape, (condylobasal length 10.8-11.5) as are spec- Barrow, and the Montebello Islands to E. imens from northeastern Western Australia pumilus, but leave them unidentified as to (Parry Creek, Inglis Gap, Kalumburu, Koo- subspecies. I cannot fit them into p. caurinus, lan Island, Tunnel Creek, Napier Range). allocation to p. pumilus seems ruled out on These specimens are, in fact, unusually small the basis of geography and bacular mor- (condylobasal length 10.3-11.0) and are im- phology, and description of a new subspecies portant in two respects; they come from the should only be done after a thorough revision region of the type locality of caurinus (Drys- ofE. pumilus (and perhaps all Australian Ep- dale), and they are more or less sympatric tesicus) throughout the continent. I am not with E. douglasi (see below for details). Spec- in a position to do this because ofinadequate imens from the Pilbara area to the southwest material from many areas. (Peawah, Tambrey, Woodstock) are larger Eptesicus douglasi: This recently described (condylobasal length 11.0-11.6) as are the few species (Kitchener, 1976) has only been re- that I have seen from south of the Tropic of corded from the Kimberley area (northeast- Capricorn (Kurara) in Western Australia ern Western Australia), but the National Mu- (condylobasal length 11.6-11.8). By far the seum of Natural History has two specimens largest Western Australian specimens I have from northeastern Northern Territory that I allocated to E. pumilus, however, come from would refer to this species. These were iden- extreme northwestern Western Australia tified by Johnson (1964) as E. pumilus cauri- (Yardie Homestead on the Northwest Cape, nus but the condylobasal lengths (11.9, 12.1) Barrow Island, and Hermite Island in the lie well outside the range of Northern Ter- Montebello group). With condylobasal ritory p. caurinus, but within that of Western lengths of 1 1.5-12.5, I thought it necessary Australian douglasi. All American Museum to compare these with the three other rela- specimens were collected by W. H. Butler at tively large Australian Eptesicus outside of only four localities: Inglis Gap (1965), Mt. southeastern Australia; they are regulus from Anderson (1965, 1973), Derby (1973), Lan- southern Western Australia, douglasi from gey Crossing (1965). At Inglis Gap it was northeastern Western Australia (Kimberley), found sympatric with E. pumilus caurinus, and p. pumilus from eastern Queensland but at Mt. Anderson, Derby, and Langey (particularly the Cooktown area). The large Crossing it occurred alone. Kitchener (1976) northwestern Western Australian population recorded E. douglasi from Tunnel Creek (the is clearly not regulus, using the character type locality), but the two American Museum mentioned above. However, I cannot distin- Eptesicus from this locality are clearly E. guish it from either douglasi or p. pumilus. pumilus caurinus. Specimens of E. douglasi These two taxa have never been satisfactorily (including the two National Museum speci- distinguished. Kitchener (1976), compared mens from Cape Arnhem) are clearly larger three bacula ofdouglasi with five bacula iden- than Kimberley pumilus caurinus (condylo- tified as p. pumilus, three from the eastern basal length 11.5-12.1 vs. 10.3-11.0), but Pilbara and two from Yardie Homestead (one there is some overlap with pumilus from far- ofwhich was figured). McKean, Richards, and ther to the southwest and complete overlap Price (1978), on the other hand, do not agree with the Northwest Cape and island popu- that these five specimens are p. pumilus since lations. My only reason for associating the their bacula are somewhat different from latter populations with E. pumilus specifi- eastern Australian specimens and imply that cally, rather than with E. douglasi, is that Kitchener's pumilus are all p. caurinus. I these northwest populations intergrade with would agree that this is probably true for the pumilus caurinus (in the Pilbara), but E. eastern Pilbara specimens but not for those douglasi occurs sympatrically (without inter- 1 984 KOOPMAN: AUSTRALIAN BATS 17 gradation) in the Kimberley. The two species base ofCape York (as is done here). However, may also be sympatric in the Northern Ter- they pointed out that all specimens from ritory, but the known localities are quite sep- northwestern Queensland could be regarded arate. Thus the picture as I see it is of two as intergrades. Van Deusen and Koopman species ofEptesicus in tropical Australia. Ep- (1971) believed that the gap in the recorded tesicus pumilus extends across the entire con- range of C. n. nigrogriseus between northern tinent with marked geographical variation, and extreme southeastern Queensland was an but with the large far eastern and far western artifact ofcollecting, but the map in Hall and infraspecific forms connected by intergrades Richards (1979) shows the gap still essen- with a small subspecies that spans the inter- tially unfilled and they imply that the species vening area. Eptesicus douglasi has a limited is now extinct in this isolated southeastern range clearly distinct from the populations of portion of its range. E. pumilus sympatric with it, but probably Chalinolobus picatus: Again, there is little not distinguishable from the species as a to add to the information in Van Deusen and whole. How this situation arose is unclear to Koopman (1971) except that the species is me at present. now also known from South Australia (Hall GENUS Chalinolobus: In the restricted sense, and Richards, 1979). Confined to eastern this genus is confined to Australia, southern Australia, west of the Great Dividing Range, New Guinea, Norfolk Island, New Caledo- it is known from four localities just north of nia, and New Zealand. However, I have pre- the Tropic ofCapricorn. The only specimens viously (Koopman, 1971) treated the African in the American Museum ofNatural History, Glauconycteris as a subgenus of Chalinolo- however, are from farther south in south- bus. Of the six currently recognized species western Queensland (Birdsville) and north- of the subgenus C. (Chalinolobus), only the western New South Wales. New Zealand tuberculatus does not occur in Chalinolobus morio: The overall distribu- Australia. Of the remaining five, only dwyeri tion ofthis species (which is confined to Aus- (of which I have seen no specimens) fails to tralia, including Tasmania) is unclear to me. reach tropical Australia (though picatus bare- Hall and Richards (1979) show the species ly gets across the Tropic ofCapricorn). These with a fairly extensive range in eastern Aus- four species are therefore the ones treated tralia (mostly east of the Great Dividing below. Range) as far north as the Townsville area. Chalinolobus nigrogriseus: There is little to Parker (1973) shows the distribution in the add to what has already been said by Van Northern Territory as confined to the far Deusen and Koopman (1971). Tate (1952a) south, not extending north of the Tropic of recorded this species (under the name of rog- Capricorn. Hamilton Smith's (1966) map ersi) from two localities. The American Mu- shows it confined to localities south of 30 seum has specimens from two other Queens- degrees in Western Australia, but he maps land localities, 24 mi. s. Burketown collected only cave occurrences and in some parts of by R. F. Peterson in 1959 and Karumba by its range it is a tree-dwelling bat. Ride (1970) G. Pawlowski in 1961. Two Northern Ter- gives the range only as "Southern Australia, ritory localities are also represented, 18 mi. including Tas.," but none ofthe localities list- w. Wollogorang by R. F. Peterson in 1959 ed by Kitchener and Vicker (1981) are north and 12 mi. e. Coolibah by Nelson, Butler, of28 degrees. Therefore, I know ofno records and Rosen in 1969. In tropical Western Aus- north of the Tropic of Capricorn except for tralia, W. H. Butler collected the species at the central Queensland ones given by Hall three localities, Ningbing and Parry Creek and Richards (1979). All American Museum (1965), and with Nelson and Rosen at North specimens are from farther south in south- Creek (1969), all in the Kimberley region. eastern Queensland (Bunya Mountains), Tas- Van Deusen and Koopman (1971) combined mania, and southwestern Western Australia western rogersi with eastern nigrogriseus (Busselton, Contine). (which had previously been treated as a sub- Chalinolobus gouldii: This is the most species of C. picatus) and drew the line be- widespread species in the subgenus Chali- tween the two subspecies at the southwestern nolobus, occurring over most of continen- 18 AMERICAN MUSEUM NOVITATES NO. 2778 tal Australia, Tasmania, Norfolk Island GENUS Nycticeius: I have previously (Troughton, 1967), and New Caledonia (Koopman, 1978) discussed the status ofthis (Koopman, 1971). Tate (1952a) does not re- genus and its species in Australia. Except for cord this species, whose distribution stops one species that also occurs in New Guinea, just short of the Cape York Peninsula. The all are endemic to Australia (not including American Museum has specimens collected Tasmania) and are widely separated geo- by G. Neuhauser in 1938 from two localities graphically from other species of the genus in northcentral Queensland (Pentland; Mal- in Africa (including extreme southwestern bon). From Western Australia there are spec- Asia) and North America (including Cuba). imens collected by W. H. Butler from La The Australian species fall into two groups, Grange Dam (1963) and Inglis Gap (1965) the large rueppelli (subgenus Scoteanax) and and in 1969 Nelson, Butler, and Rosen col- the smaller taxa (subgenus Scotorepens). lected specimens from Cape Bossut. All spec- There has never been any problem as to the imens I have seen from tropical Australia identity of rueppelli on a species level, but belong to the small northern C. g. venatoris, there has been much disagreement as to how and the contrast in size between northern many species of N. (Scotorepens) to recog- Queensland specimens on one hand and those nize. Estimates have gone from five (Mc- from New South Wales and South Australia Kean, 1966) to two (Ride, 1970). In 1978, I on the other is striking. I have seen only three examined this problem (Koopman, 1978) and specimens from the Northern Territory all in came to the conclusion that three should be the National Museum ofNatural History, but recognized. In view of continued disagree- Parker (1973, p. 37) refers specimens from ment with this point of view, I have reex- the northern part of the territory to C. g. amined the pattern ofspecies and population venatoris, and quotes McKean to the effect relationships to see how, ifpossible, the prob- that those from the southern part are inter- lems can be resolved. grades with C. g. gouldii. The single specimen Perhaps the best way to approach this I have seen from Darwin (USNM 237954) problem is geographically, starting in in the far north is referable to g. venatoris Queensland, where the greatest complexity (condylobasal length 14.1). On the other hand, exists, particularly if one adopts the five specimens from near Alice Springs (USNM species hypothesis (Hall and Richards, 1979). 284187) and the Horsehoe Bend ofthe Finke The largest taxon is influatus, which I pre- River (USNM 284188), both in the far south, viously (Koopman, 1978) recognized as one are indistinguishable from southern Austra- of three valid species. I have seen only two lian specimens of g. gouldi (condylobasal specimens, both males from central Queens- length 14.7, 15.4). In Western Australia, land, the type from Prairie in the British Mu- specimens from the three northern localities seum (BM) and one from 20 mi. s. Mt. Isa are definitely venatoris. The largest of these borrowed from the Commonwealth Scientif- (a skull extracted from the largest of the al- ic and Industrial Research Organization coholics from Cape Bossut) has a condylo- (CSIRO) in Canberra. The skulls have con- basal length of 14.2. This is the same as the dylobasal lengths of 14.8 and 15.2 and since smallest skull I have seen from southern females tend to be larger than males in N. Western Australia (150 mi. n. Geraldton), (Scotorepens), females would be expected to which was likewise extracted from the small- be even larger. Hall and Richards (1979) map est ofa large series ofalcoholics. Other West- additional localities from both north (Cook- ern Australian localities for C. g. gouldi (as town area) and south (southcentral Queens- represented by American Museum material) land) of Prairie and Mt. Ida. The next largest are Lake Nabberu and Gandak. I have seen named species is balstoni. I have seen only no specimens of C. gouldi from between Cape two Queensland specimens (CSIRO), both Bossut and Lake Nabberu. If C. gouldi occurs from Mt. Pluto (near Tambo). The male has in this intervening area (as according to Ban- a condylobasal length of 14.4, the female, nister, 1969, pp. 67, 69, and Kitchener and 14.5, both significantly smaller than the two Vicker, 1981, it does) this is where inter- influatus I have measured. Hall and Richards grades would be expected. (1979) state (p. 51) that balstoni reaches the 1 984 KOOPMAN: AUSTRALIAN BATS 19 coast near Rockhampton, but their map ing possible contact is between orion and bal- shows this taxon confined to the area west of stoni. These are apparently separated by the the Great Dividing Range and south of the Great Dividing Range, although Hall and Tropic ofCapricorn. McKean (1966) suggests Richards (1979, p. 51) imply sympatry with that influatus is conspecific with balstoni. This aquilo near Rockhampton. The most impor- hypothesis should be quite testable since the tant difference between orion and balstoni ap- maps in Hall and Richards (1979) show the pears to be the broader skull of the former, two species in close proximity in southern though the difference is not great. Compari- Queensland, where they should intergrade if son of the two females from Dorrigo (orion, conspecific. Currently recognized in coastal MCZ) with the female from Mt. Pluto (bal- Queensland are two taxa, which are smaller stoni, CSIRO) shows less difference in con- than typical balstoni but larger than greyi. dylobasal length (14.3, 14.4 vs. 14.5) and These are sanborni (originally described from maxillary tooth row length (5.1, 5.2 vs. 5.2) New Guinea) and aquilo, originally described than in width across the last molars (6.8 vs. as a subspecies ofthe coastal New South Wales 6.6). Obviously, more skulls from in and (and perhaps Victoria, see p. 53 of Hall and around the Great Dividing Range in northern Richards, but not their map) orion. These New South Wales should be compared. The two coastal Queensland taxa have never been British Museum has two relevant specimens. clearly distinguished. Hall and Richards An unsexed skin and skull from the Liverpool (1979) indicate slight size and color differ- Range has a maxillary tooth row length of ences, but it would appear that these are weak 5.0 and a width across the last molars of 6.6, subspecies at most. The critical area would thus agreeing better with balstoni than with seem to be the Tully-Innisfail region with orion. An alcoholic female from 40 mi. s. sanborni to the north and aquilo to the south. Moree had been identified as orion, but the Hall and Richard's text implies allopatry, but skull was not available to me. The smallest their maps suggest contact. To me, the dis- species of Nycticeius in Queensland is N. tinction between aquilo and true orion seems greyi, which also has the broadest range with- much clearer. Hall and Richards (1979) state in the state, being absent only from Cape York that aquilo occurs from Rockhampton north and most of the Pacific coast. It extends as and imply that specimens from extreme far north as the Gregory River near the Gulf southeastern Queensland are true orion. The of Carpentaria, reaches the Pacific coast (at specimens I have seen in the American Mu- least in the Townsville-Tully area), and ex- seum (AMNH), Museum of Comparative tends well to the south of Queensland west Zoology at Cambridge in Massachusetts of the Great Dividing Range. Over much of (MCZ), CSIRO, BM, and Los Angeles Coun- this area it is sympatric with influatus, bal- ty Museum ofNatural History (LACM) show stoni, or aquilo. In Queensland, the condy- otherwise. Those from Cooktown north have lobasal length measures 11.6-12.2 in males condylobasal lengths as follows: single male and 11.8-12.8 in females, which means that (12.8), five females (12.9-13.6); from the greyi is everywhere (at least in Queensland) Tully-Townsville area, two males (12.9), three distinct in skull size from any other popu- females (12.9-13.4); from Southport in ex- lation of Nycticeius with which it is sympat- treme southeastern Queensland, a single fe- ric. The above is my rationale for recogniz- male (13.2); from Dorrigo in northeastern ing, in Queensland and New South Wales, New South Wales, two females (14.3, 14.4). three species; N. influatus, N. balstoni (in- Southport and Dorrigo are only about 170 cluding orion, aquilo, and sanborni), and N. mi. apart. This means that either the tran- greyi. sition from aquilo to orion is sharp or else I have seen no specimens ofNycticeius from that the Southport and Dorrigo specimens Victoria and very little from South Australia happen to fall at the two extremes of a highly so I will proceed to the Northern Territory. variable population. Obviously, northeastern Since Parker (1973) does not distinguish be- New South Wales and southeastern Queens- tween greyi and balstoni, it is difficult to de- land is a critical area for understanding the termine the distribution of the three named orion-aquilo relationship. Another interest- forms which have been recorded from the 20 AMERICAN MUSEUM NOVITATES NO. 2778 territory. However, I have seen a fair amount have the type series ofbalstoni from Laverton ofmaterial (USNM, MCZ, BM, CSIRO) from in the southern part of the state, constituting which the following pattern emerges. All the only material ofnominate balstoni I have specimens from south of the Tropic of Cap- seen from Western Australia. These Laverton ricorn (Finke River, Horseshoe Bend; Her- specimens are smaller than most specimens mannsburg; Charlotte Waters) are large (con- ofnominate balstoni from the Northern Ter- dylobasal length 13.9-14.6 in males, 14.2- ritory and Queensland (condylobasal length 15.0 in females) and agree well with typical of the three female skulls, including the type balstoni. All specimens from north of 190S in the British Museum, 13.9-14.0). The other and east of 1340E (18 mi. w. Wollogorang; two named forms in Western Australia are Black Waterhole, Nicholson River; Upper caprenus and greyi, and this has been a major Gorge, Fish River; Alexandria) are small point ofdisagreement between me and West- (condylobasal length 11.8-12.6 in males, ern Australian mammalogists. The most im- 12.2-12.6 in females) thus agreeing well with portant material by far is a series in the Amer- Queensland greyi. Parker's map (1973, p. 39) ican Museum of Natural History from Cape shows five localities south of220S (two slight- Bossut, to which I have added specimens that ly north of the Tropic of Capricorn), which have since become available from Frazier presumably are typical balstoni and four lo- Downs and La Grange, two closely adjacent calities north of20°S and east of 1 340E, which localities. The suggestion has been made that are presumably greyi. Thus, there appears to this all represents one species, in which case be a gap between the two ranges where no it would be extremely variable. The condy- Nycticeius occurs. Material from the north- lobasal length of the entire series would be western Northern Territory presents more of 11.3-13.4 (males) and 11.6-14.2 (females). a problem. I have seen nine intact skulls from In my opinion, the series is much better seg- this area and only five of these come from regated into two species: greyi (males 11.3- the most critical portion of it (Cobourg Pen- 12.1, females 11.6-12.3) and balstoni ca- insula; DeafAdder Creek; Darwin) which in- prenus (males 12.8-13.4, females 13.0-14.2). cludes the type locality of greyi at Port Es- The female with a condylobasal length of 14.2 sington. I previously (Koopman, 1978) is interesting because it falls well within the allocated one ofthe Cobourg Peninsula spec- variation ofN. b. balstoni and is in fact larger imens and those from DeafAdder Creek and than any of the type series of balstoni (also Darwin to balstoni caprenus and left unal- females). Most of the remaining specimens I located a topotype of greyi (also from the have seen from Western Australia fall into Cobourg Peninsula). I did this because four the variability of either greyi or caprenus as of the specimens agreed better with Western known from the Cape Bossut area. The only Australian caprenus than with Western Aus- Western Australian specimen of Nycticeius tralian greyi. However, I am now dubious that has given me difficulty is a skull from that caprenus really occurs in the Northern Parry Creek, which is no longer available to Territory. The condylobasal length measure- me. It is ofa male with a condylobasal length ments of these extremely northern Northern of 12.5, well below that of the smallest male Territory specimens are 12.6 (for the single of caprenus from Cape Bossut. The skull is male) and 12.7-13.0 (for the four females), closer in size to that ofgreyi ofWestern Aus- which straddles the size hiatus between greyi tralia and since that is so, it is strange that it and caprenus as I have determined it in West- is larger than any of the female greyi from ern Australia. While more material from this Western Australia. Since Parry Creek is quite area is necessary for a definite decision, I am far north and fairly close to the Northern inclined to treat these as unusually large Territory border, I have considered the pos- greyi. Specimens from a little farther south the Creek (12 mi. e. Coolibah; Montejinni area; Wave sibility of allying Parry specimen Hill), with condylobasal lengths of 12.0 (two with the large greyi occurring around Van males) and 12.2-12.3 (two females) agree bet- Diemen's Gulf in northwestern Northern ter with greyi from farther east (and farther Territory. However, I have seen two speci- west). mens, one female and one unsexed from Passing on to Western Australia, first we Ningbing, which is also far to the north and 1 984 KOOPMAN: AUSTRALIAN BATS 21 even closer to the Northern Territory bound- Australia (where it is in contact with N. bal- ary. With condylobasal lengths of 13.0 and stoni caprenus), larger in western Queensland 13.2, they fall within the Cape Bossut ca- and most ofthe Northern Territory, and larg- prenus range. More material from extreme est in the extreme northwestern Northern northeastern Western Australia is necessary Territory. It was from this usually large pop- to determine the greyi-caprenus limits in that ulation, on the Cobourg Peninsula and just area. Besides the Parry Creek skull, I have south of Van Diemen Gulf that the type of seen skulls offour males and one female greyi greyi was drawn, and I believe that the un- from this area (Kununurra; Lissodell home- certainty as to where this population should stead; Carranya homestead) in two museums be allocated taxonomically has caused much (CSIRO and FMNH). Condylobasal lengths of the confusion. are 12.0-12.3 for males and 12.4 for the fe- Nycticeius balstoni: This is another wide- male. These lie at the upper end or slightly spread Australian species (according to my beyond the upper end of the Cape Bossut concept) which also occurs in southeastern greyi series. They agree better with specimens New Guinea. In northern Queensland it oc- in the Northern Territory immediately to the curs only east of the Great Dividing Range, east (Coolibah; Montejinni, Wave Hill) than but farther south it extends from the Pacific to those to the northeast (Darwin; Deaf Ad- Ocean across Victoria, New South Wales, der Creek; Cobourg Peninsula). southern Queensland, southern Northern To sum up this survey, I still believe that Territory and throughout much of Western there are three species of N. (Scotorepens): Australia. The two southern subspecies, b. greyi, balstoni (including sanborni, aquilo, orion and b. balstoni, are large, but the three orion, and caprenus), and influatus but have northern subspecies, b. sanborni, b. aquilo, indicated several critical areas where this the- and b. caprenus, are small. Nycticeius b. orion ory might be tested. The only important de- does not occur north of the Tropic of Capri- parture from the conclusions in my previous corn and N. b. balstoni extends at most slight- (Koopman, 1978) paper is that I no longer ly north of it in the Northern Territory. The believe that balstoni caprenus reaches the three small subspecies, however, all occur ex- Northern Territory, but would confine it to tensively north of the Tropic of Capricorn. the northern and western portions ofWestern Nycticeius balstoni sanborni and N. b. aquilo Australia. are found only to the east of the Great Di- Nycticeius greyi: As I now conceive this viding Range, meeting at about the south- species, it extends from the Pacific coast near eastern base of the Cape York Peninsula (if Townville south to the Victorian border, and indeed they are distinct). Nycticeius balstoni west across the northern part ofthe Northern caprenus (as I now conceive it) is confined to Territory to northeastern Western Australia Western Australia, extending from its north- at least as far as Cape Bossut. Tate (1952a, east corner southwest at least to Yanarrie p. 601) recorded this species from Pentland River (22°50'N, 11 5°E). I erroneously thought and Malbon. From Queensland, the Ameri- that one of these specimens came from Bar- can Museum ofNatural History now has two row Island. The only specimen of b. caprenus additional specimens from between Ca- I have seen from farther south is a single mooweal and Mt. Isa collected by P. Kraus female skull (no longer available to me) which in 1967 and 24 mi. s. Burketown by R. F. came (Bannister, 1969) from Contine Peterson in 1959. From the Northern Ter- (32050'S, 116050'E). It is a perfectly typical ritory, there are specimens from 18 mi. w. caprenus (condylobasal length 13.1) and Wollogorang by R. F. Peterson in 1959 and shows no indication ofintergradation with b. 12 mi. e. Coolibah by Nelson, Butler, and balstoni, although, as indicated above, some Rosen in 1969. From Western Australia, there specimens from farther north do show such is a large series from Cape Bossut by Nelson, intergradation. Tate (1 952a, p. 601) recorded Rosen, and Butler in 1969; W. H. Butler in N. b. sanborni from two localities in the 1973). As indicated above, N. greyi is small- Cooktown area of northern Queensland. All est in eastern Queensland (where it is in con- the remaining material ofthis species that we tact with N. balstoni aquilo) and in Western have is from Western Australia and belongs 22 AMERICAN MUSEUM NOVITATES NO. 2778 to N. b. caprenus. This includes specimens New Guinea, New Britain, Admiralties, Sol- from Frazier Downs (W. H. Butler in 1963), omons, and New Hebrides. Maeda (1982) La Grange (W. H. Butler in 1963), Ningbing recognizes three species, M. tristis in the Phil- (W. H. Butler in 1965), Yanarrie River (W. ippines, New Guinea, and New Britain, M. H. Butler in 1965), Cape Bossut (Nelson, But- bismarckensis from the Admiralties, and M. ler, and Rosen in 1969, W. H. Butler in 1973), melanesiensis from the Solomons and New Yeeda Creek (Nelson, Butler, and Rosen in Hebrides (though in a footnote, p. 47, he syn- 1969). onymized melanesiensis with Peterson's pro- Nycticeius influatus: This species is con- pritristis insularis). While both the Peterson fined to Queensland, where (according to Hall and Maeda arrangements can be derived from and Richards, 1979, p. 51) it has an extensive Hill's by simple splitting, Peterson's and distribution from the southern parts of the Maeda's are irreconcilable. Cape York Peninsula to just south of the Maeda's procedure (1982, p. 17 and per- Tropic ofCapricorn. The American Museum sonal commun.) seems to be as follows. With- of Natural History has no specimens. in the main Japanese islands, where Maeda's Nycticeius rueppelli: Hall and Richards earlier work was done and where his material (1979) give as the range ofthis species "coast- is most adequate, there is little geographical al eastern Australia from southern New South variation. From this observation, he has Wales to Ingham, Queensland." However, drawn the conclusion that this pattern exists specimens in the American Museum and the throughout the genus. He therefore interprets Museum of Comparative Zoology carry the any morphological change in going from one distribution somewhat farther north to Lake area to another as a species change. Numer- Barrine, definitely onto the Cape York Pen- ous species are therefore recognized (19 in insula. Tate (1952a) did not record this the area he covers), but these tend to have species, but the American Museum has a spotty, discontinuous ranges since similar ap- specimen from 30 mi. s. Ravenshoe collected pearing forms, even though widely separated by R. F. Peterson in 1959. geographically (with other species in be- GENUS Miniopterus: As previously dis- tween) are combined into the same species. cussed (Koopman, 1982), this genus is in Needless to say, my method is quite different taxonomic confusion with no two investi- and I will try to summarize it here (though I gators agreeing on what species to recognize use it for determining species in general, not or what to call them. Many of the respective just for Miniopterus). I first look for as many points of view have not been published but places as I can where sympatry occurs be- as ofthe end of 1981, it was evident that Hill tween closely related species. From study of (1971) and Peterson (1981a) differed rather these areas, I try to determine the kind and sharply. Since then, Maeda (1982) has pub- degree of difference which distinguishes the lished a classification of all the Miniopterus most similar pairs ofsympatric species in the ofthe Palearctic, Indo-Malayan, and Austra- group under consideration (in this case, Mi- lian regions which differs rather radically from niopterus). Then, in trying to interpret the those of either Hill or Peterson. This is per- specific status or its absence between related haps best shown in their treatments of the allopatric forms (usually separated by a Miniopterus tristis group which all have pub- marked barrier), I compare their kind and lished on. Ignoring subspecies and also the degree of difference with what I have found status of robustior of the Loyalty Islands, between the related sympatric species. A phe- which all three authors recognize as a sepa- notypic break between populations on the two rate species, we have the following alternative sides of a barrier is therefore not automati- treatments. Hill (1971, pp. 578-580) recog- cally treated as indicating a species difference, nized a single species, M. tristis extending but only if it is of such a nature that it com- from the Philippines to the New Hebrides. pares well with differences between related Peterson (1981 a) recognized two species, M. sympatric species. I do not claim that this tristis from the Philippines (and also Celebes, method is infallible, but I do believe that it from which neither of the other authors has is the most objective and efficient way to make seen specimens) and M. propritristis from such a decision. But whatever the method, 1984 KOOPMAN: AUSTRALIAN BATS 23 decisions are subject to change when new data er individuals of oceanensis. This is partic- become available. ularly true of a single skull from Panmure in Taking up Maeda's treatment ofAustralian southwestern Victoria. I therefore believe that Miniopterus we get the following pattern. The the Flinders Range specimens that Maeda re- larger Miniopterus of eastern Australia (usu- ferred to macrodens are simply unusually large ally regarded as a form of M. schreibersi) is oceanensis. However, a study ofgeographical described as a new species M. oceanensis (type variation of Miniopterus in its limited South locality Cape York), but to it are also referred Australian range would certainly be instruc- specimens from San Cristobal in the Solo- tive. The third large species of Miniopterus mons, St. Matthias in the Bismarcks, south- that Maeda recognized in Australia was esch- eastern New Guinea, Amboina in the Mo- scholtzi (type locality in the Philippines), with luccas, and a small area in Yunnan which he synonymizes orianae (type locality (southwestern China), and upper Burma. in the Northern Territory). To eschscholtzi Maeda's Australian material of oceanensis he also allocated specimens from South Aus- came from Queensland and New South Wales. tralia and from San Cristobal in the Solo- He saw no specimens from Victoria, but from mons. I have not seen any Solomon Islands South Australia, he had four specimens from material and suspect that the South Austra- the Flinders Range. These he allocated to lian record is in error (see Aitkin, 1975, for another new species (M. macrodens, type lo- a discussion of this problem in a different cality in Borneo) to which he also allocated species), but again careful study of South specimens from Timor (presumably the Australian material would be rewarding. I species Goodwin, 1979, called M. magnater), have compared true eschscholtzi from the Moluccas (where apparently sympatric with Philippines with orianae from Northern Ter- oceanensis), Java, and several areas in south- ritory and Western Australia and can see no eastern Asia. I have looked at all Miniopterus special resemblance. To me they are as dif- skulls in the American Museum of Natural ferent from one another as either is from Eu- History which seem relevant to the problem ropean schreibersi (which Maeda puts in a at hand. Specimens from Hainan (China), different species group) and would put them Burma, Malaya, Bali, and Timor are all from all in the same species for which the oldest within or near areas where Maeda allocated name would be M. schreibersi. As explained specimens to his macrodens (and seem to below, I regard oceanensis as conspecific with agree with his concept of that species), but I orianae and would regard both as subspecies have seen no specimens from the Flinders ofM. schreibersi (and thus in agreement with Range. I have also looked at typical magnater Hill, 1983). Summing up, I would recognize from New Guinea. Also from New Guinea the following taxa oflarger Miniopterus from but from localities farther to the south and Australia and surrounding areas: M. mag- east, I have seen specimens which I previ- nater magnater from northern and western ously (Koopman, 1982) referred to M. schrei- New Guinea; M. magnater macrodens from bersi, but without allocation to subspecies. Timor (in agreement with Hill, 1983, as to Maeda saw one of these southeastern New status, but not as to allocation of these spec- Guinea specimens (as well as typical mag- imens); M. schreibersi oceanensis from east- nater) and referred it to oceanensis. However, ern Australia; M. schreibersi ssp. from south- as I previously (Koopman, 1982) stated, these ern New Guinea; M. schreibersi orianae from southeastern New Guinea (including East Northern Territory and Western Australia. Papuan Island) specimens are smaller than Maeda (1982) recognized only one addi- those from eastern Australia and I therefore tional Australian species of Miniopterus, M. would not refer them to oceanensis. (Hill, australis. He accepted the now conventional 1983, has also recorded two specimens from view that the type locality is in the Loyalty southwestern Irian Jaya.) From study of the Islands, but also allocated specimens from extensive series of eastern Australian ocean- New Caledonia, New Hebrides, eastern Aus- ensis in the American Museum, I note that tralia, "New Guinea," Amboina (including some specimens approach macrodens from the type of tibialis), and Borneo (including Timor closely, though obviously simply larg- the type of witkampi). Maeda recognized 24 AMERICAN MUSEUM NOVITATES NO. 2778

another species, M. paululus from the Phil- Australia. Miniopterus s. orianae, on the oth- ippines, java (including the type of short- er hand (if the alleged South Australian rec- ridgei), and Rennell Island in the Solomons. ord is ignored), is confined to northwestern This is essentially the arrangement of names Northern Territory and northeastern West- that Peterson (1982b) adhered to, but he has ern Australia. Though M. s. oceanensis ex- concluded that the type of australis actually tends slightly west of the Great Dividing came from eastern Australia and that M. aus- Range at Lyndhurst Station, there is still a tralis does not occur in either New Caledonia wide gap between the ranges of the two sub- or the Loyalty Islands. Hill (1983) disagrees species. While the subspecies remain distinct, with these conclusions, but I still see prob- the difference is not great and there is even lems. If we follow Peterson, then M. a. aus- a little overlap. Thus five skulls from Pent- tralis occurs in eastern Australia, southeast- land and Lyndhurst have condylobasal ern New Guinea, and on Tagula Island in the lengths of 15.1-15.8 and the skull from north Louisiade archipelago; M. australis ofuncer- of Katherine has a condylobasal length mea- tain subspecies occupies northern and west- suring 14.8. Furthermore, the largest of a se- ern New Guinea; M. paululus is in Timor. ries of 25 skulls from Darwin in the National The third species in Timor (besides mag- Museum ofNatural History (USNM 248203) nater, macrodens, and paululus) is M. pus- has a condylobasal length of 15.2. As indi- illus, which does not occur in either Australia cated above, both subspecies are probably or New Guinea, but does range from India endemic to Australia. to the Aru Islands (southwest ofNew Guinea) Miniopterus australis: Tate (1952a) record- and perhaps occurs on some islands east of ed this species from four localities (based on New Guinea and Australia. American Museum material) in eastern Miniopterus schreibersi: Tate (1952a) re- Queensland north ofthe Tropic ofCapricorn. corded this species from seven localities There are also specimens from Port Stuart (based on American Museum specimens) in (G. Neuhauser in 1938); Cowie Bay (J. Rob- northeastern Queensland. There is also erts in 1950); Innisfail (J. L. Harrison in 1959); Queensland material in the American Mu- Chillagoe, Mt. Garnet, and Mungara (R. F. seum from a number of other localities. In Peterson in 1959). Like M. s. oceanensis, M. 1937-1938 G. Neuhauser obtained speci- a. australis extends slightly west of the Great mens from Cunjeboi, Pentland, Port Stuart Dividing Range but does not extend as far (probably = Port Stewart), and Somerset. south (only to northeastern New South Tate and Van Deusen also collected speci- Wales). Also unlike M. s. oceanensis, it oc- mens from Lockerbie, but they were omitted curs in New Guinea, but has no representa- from Tate's (1952a) account. In 1950-1951 tive in the Northern Territory or Western J. Roberts collected specimens from Mt. Australia. Amos (Phoenician), Ayton, Cooktown, and Murina florium: This species has recently Green Hills (south of Cooktown); J. L. Har- (Richards et al., 1982) been recorded from rison in 1958-1959 from Bramston Beach near Atherton on the Cape York Peninsula, and Ingham; R. F. Peterson in 1959 from Queensland. The species is distributed from Koombooloomba Creek (south of Raven- the Lesser Sundas to New Guinea. Though shoe) and Lyndhurst Station; R. M. Ryan and three subspecies are recognized by Laurie and L. Wassell from 5 mi. se. Coen. From the Hill (1954), their type localities are all in the Northern Territory, there is a single specimen Lesser Sundas and Moluccas and it is uncer- from north of Katherine collected by A. J. tain what subspecific name should apply to Coventry in 1963. All Western Australian material from New Guinea and Australia. It specimens were obtained by W. H. Butler, should be pointed out, however, that the either in 1965 (Ningbing; Parry Creek) or measurements (supported by the scale oftheir 1973 (Mitchell River; Port Warrender). All photographs) that Richards et al. (1982) give specimens from eastern Australia are M. s. are unusually large for Murinaflorium. Thus oceanensis, which has a wide distribution for their specimen vs. the largest of the Ce- from islands off the northern end of Cape lebes and New Guinea florium skulls in the York to Victoria and southeastern South American Museum, we obtain: condylobasal 1 984 KOOPMAN: AUSTRALIAN BATS 25 length (15.8, 14.2); maxillary tooth row length of the largest Nyctophilus skull I have mea- (5.8, 5.2) and the latter measurement is also sured (18.0 for a specimen of N. timoriensis larger than that for the types of lanosa and major from southwestern Western Austra- toxopei (5.6, as given by Tate, 1941), which lia). There has been considerable confusion are currently considered to be conspecific with in the past as to what species to recognize in M. florium. Hill (1983), however, gives a Australia and, as will be seen, my arrange- condylobasal length of 15.7 and a maxillary ment is somewhat different from any pro- toothrow of 5.6 for a skull from Goram in posed in the past. Starting with the smallest the Moluccas. The American Museum has species that lacks a highly modified postnasal no Australian material of Murina. elevation (not to be confused with the true Kerivoula (Phoniscus) papuensis: Although noseleaf that is anterior to it), thus corre- Hill (1965) and others (e.g., Hall and Rich- sponding to degrees 1 and 2 ofThomas (1915), ards, 1979) have treated Phoniscus as a sep- we have N. walkeri. In the past considerable arate genus, I have given reasons for includ- uncertainty has existed concerning this species ing Phoniscus in Kerivoula as a subgenus (e.g., Ride, 1970, p. 165), but it is perfectly (Koopman, 1982, p. 22). The only species of distinct with at least one good series referred this subgenus in the Australian region (see to it. Next in size is N. arnhemensis, which Hill, 1955) is K. (P.) papuensis, known only is also quite distinct. The larger populations from tropical Queensland and eastern New with the less derived nasal elevation are treat- Guinea. Tate (1952a) did not record it, but ed by Ride (1970) as belonging to two species, the American Museum of Natural History bifax and timoriensis. Hall and Richards has a single specimen obtained by J. Roberts (1979), however, separate N. gouldi (previ- in 1950 at Shipton's Flat. Hall and Richards ously regarded as the southeastern mainland (1979) record it from two additional localities subspecies of N. timoriensis) as a separate in eastern Queensland and also (surprisingly) species and after studying all the Nyctophilus in northwestern Queensland. This unusual material at the British Museum, I agree, and distribution suggests that papuensis may also believe that at least near sympatry can be occur farther west. Hall and Richards (1979) demonstrated. However, there is complete suggested that it may be extinct, but I see no allopatry between gouldi and bifax and com- basis for this statement. Like other members parison of their skulls shows close resem- of its subgenus (see Hill, 1965), it is indeed blance. There are other characters which rare in collections from throughout its range. Thomas (1915) pointed out, but, as I will Besides the Shipton's Flat specimen men- explain below, these do not seem to be of tioned above, the American Museum has only sufficient magnitude to characterize species. two specimens, both from Papua New The last Australian Nyctophilus species to be Guinea, one from Milne Bay province, the considered, N. geoffroyi, differs from the rest other from Morobe province. Hill (1965) saw in having the postnasal elevation much more only two specimens, the type from Central highly developed (degree 3 ofThomas, 1915). province in Papua New Guinea and another Whereas there are several names which have specimen from Rockhampton in Queens- been applied to variants, there seems to be land. Whether the species is genuinely rare general agreement, at least since Thomas in nature or whether its habits have pre- (1915), that only a single species is repre- vented it from being collected efficiently is as sented. I therefore recognize five species of yet uncertain. Nyctophilus in Australia, walkeri, arnhemen- GENUS Nyctophilus: This genus is virtually sis, gouldi, timoriensis, and geoffroyi, all of confined to Australia and New Guinea, which are known north of the Tropic of Ca- though the type of N. timoriensis is supposed pricorn. I have previously (Koopman, 1982) to have come from Timor, and McKean discussed the four species ofNyctophilus that (1975) has described a (?subfossil) species occur in New Guinea. Two of them, gouldi from Lord Howe Island. This species (N. (previously called bifax) and timoriensis also howensis) is larger than any species known occur in Australia. Ofthe other two, microtis living. The condylobasal length measure- is obviously very closely related to arnhe- ment McKean gives (21.4) contrasts with that mensis and their relationship will be dis- 26 AMERICAN MUSEUM NOVITATES NO. 2778

cussed in the N. arnhemensis account. The the National Museum ofNatural History and other species, microdon, is much more dis- the British Museum from several localities in tinct. I believe that it is most closely related northern Australia and Papua New Guinea. to N. geoffroyi, albeit considerably more From this comparison, it becomes evident primitive. Like gouldi, timoriensis, and geof- that the most important difference is in ear froyi, N. microdon has the ears connected by length, arnhemensis clearly having longer ears a high band. However, the postnasal eleva- (19-22 vs. 14-17) and this is also reflected tion is considerably better developed than in in the slightly better development ofthe band gouldi or timoriensis, though not as highly connecting the ears (Hill and Koopman, 1981) developed as in geoffroyi. Nyctophilus micro- and the larger bullae (which I have not tried don resembles N. geoffroyi and differs from to measure). I would therefore retain arnhe- N. arnhemensis and N. gouldi in its smaller mensis as a species distinct from N. microtis. molars and less massive rostrum, suggesting Perhaps they could be included in the same that microdon and geoffroyi are adapted for superspecies. feeding on softer (and perhaps smaller) insect Nyctophilus gouldi: Tate (1 952a) recorded prey than are other Nyctophilus (with the pos- this species (under the name bifax) from three sible exception of walkern). localities on the Cape York Peninsula. The Nyctophilus walkeri: Until fairly recently, specimen from Ravenshoe, which he iden- this species was known only from the holo- tified as N. geoifroyi pallescens, however, is type, collected in northwestern Northern also referable to N. gouldi. Furthermore, the Territory. More recently, the species has been American Museum has a specimen obtained recorded from two areas in northeastern on the Atherton Tableland (9 mi. sse. Ra- Western Australia (Kimberley) and the Na- venshoe) by H. C. Raven in 1922, and three tional Museum of Natural History (USNM) from Jackeroo collected by J. Roberts in 1950. has a series of 11 specimens from Jim Jim The only other tropical Australian specimens Creek (150-160 mi. se. Darwin). The USNM in the American Museum are two from Willie series is sympatric with a series of arnhe- Wollie Spring (ca. 20 mi. wnw. Poonda) in mensis from the same localities, showing that the Pilbara region of Western Australia col- they are perfectly distinct species (condylo- lected by Nelson, Butler, and Rosen in 1969. basal length 1 1.4-12.5 vs. 13.4-14.1). Though Hall and Richards (1979) showed a marked the American Museum has no specimens of hiatus in eastern Queensland between the this species, clearly it is well established at ranges of gouldi (not extending north of the least in northwestern Northern Territory and Tropic ofCapricorn) and bifax (south to about northeastern Western Australia. Ingham), amounting to some 600 mi. The Nyctophilus arnhemensis: This species was only specimen I have seen from western described (Johnson, 1959) after Tate's (1 952a) Queensland is one in the British Museum paper appeared. The American Museum has from Cloncurry, which Thomas (1915) re- a single specimen from northwestern ferred to bifax, though in its great condylo- Queensland (24 mi. s. Burketown) collected basal length (16.0), it agrees better with by R. F. Peterson in 1959. There is also a Northern Territory daedalus (15.1-16.2) than single specimen from northwestern Northern with eastern Queensland bifax (14.7-15.4). Territory (Adelaide River, Tortilla flats) col- All Northern Territory daedalus I have seen lected by Bolton and Parker in 1967 and a and all records I know of (Parker, 1973) are series from Cape Bossut in northern Western from the extreme northwestern portion. I have Australia collected by Nelson, Butler, and seen no specimens I would refer to N. gouldi Rosen in 1969. The Queensland and Western from South Australia and only two additional Australian localities are the easternmost and ones from Western Australia (besides the two westernmost known for the species. As John- from Willie Wollie Spring mentioned above). son (1959, p. 185; 1964, p. 481) made clear, One is from a place called Eureka (which I no comparison was made between arnhe- have not been able to find); the specimen mensis and the New Guinea microtis. This I (British Museum) tag gives as additional lo- have done using specimens in the American cality information "N. Territory W. A." which Museum and measurements of specimens in I would interpret as meaning the northeastern 1984 KOOPMAN: AUSTRALIAN BATS 27

(Kimberley) portion of Western Australia. N. gouldi are of a single male from the Drys- The other, also in the British Museum, is dale River (McKenzie et al., 1977) and one from Swan River in the southwest. All these male and three females from the Dampier specimens are females and the two from Wil- Peninsula (McKenzie, 1983, p. 47). At least lie Wollie Spring have the same forearm and the Drysdale River specimen was identified ear measurements. I have extracted the skull as bifax on the basis of its bifurcate baculum from one of the Willie Wollie alcoholics and although this Western Australian locality is compared this specimen with those from Eu- on the other side of daedalus geographically. reka (first) and Swan River (last), thus in north Kitchener and Vicker (1981) list three spec- to south order: forearm length (44, 42, 40); imens as bifax (all from north of 180) and 24 ear from notch (-, 24, 27); condylobasal Western Australian specimens as cf. gouldi length (15.6, 15.7, 15.4); width across last (all south of3 10). Obviously, there is still much molars (6.9, 7.0, 6.5); maxillary tooth row to be learned about the precise relationships length (6.3, 6.1, 5.8). When these Western of bifax and daedalus. For the present I rec- Australian specimens are compared with g. ognize a single species, N. gouldi, with three gouldi from southeastern Australia and with subspecies; N. g. gouldi across southern Aus- g. daedalus from the Northern Territory, it tralia, N. g. daedalus in northwestern and becomes evident that in forearm length, the northcentral Australia, and N. g. bifax in Eureka and Willie Wollie Spring specimens northeastern Australia. The first and last are agree better with daedalus (38-44) than with most distinct from one another, with N. g. gouldi (38-41), whereas in ear length and daedalus being somewhat intermediate in width across last molars, the Swan River morphology between N. g. gouldi and N. g. specimen agrees better with gouldi (24-28, bifax. 6.5-7.0) than with daedalus (20-25, 6.7-7.2). Nyctophilus timoriensis: With the removal I am therefore inclined to identify the Eureka of gouldi, the characterization of this species and Willie Wollie Spring specimens as N. g. as the largest living species of the genus be- daedalus and the one from Swan River as N. comes clearer. Two subspecies are recogniz- g. gouldi. However, this decision is based on able in southern Australia, the very large ma- very few specimens and it is obvious that jor in the west and the somewhat smaller much more material from southern and sherrini in the east. Although the implication Western Australia will have to be analyzed ofHallandRichards's(1979, pp. 59, 61) maps before these relationships can be anything are that sherrini (described from Tasmania) more than tentative. It should be possible to belongs to N. gouldi, I have seen the type in do so now that it is clear that gouldi is not the British Museum and it belongs in N. ti- conspecific with N. timoriensis. Thomas moriensis. (On inspection, it is clear that the (1915) characterized gouldi as having the type of sherrini is somewhat immature and postnasal elevation medium (degree 2), its resultant smaller measurements have ev- whereas daedalus and bifax were character- idently been responsible for its association ized as having it poorly developed (degree 1). with gouldi.) Hall and Richards (1979) However, if the Willie Wollie specimens are showed N. timoriensis as being confined to daedalus, then in this respect they are inter- the area west of the Great Dividing Range. mediate between northeastern Queensland However, the only eastern Australian main- bifax and southeastern Australian gouldi. land specimen in the British Museum is from Nyctophilus bifax was originally distin- the Richmond River in extreme northeastern guished from both gouldi and daedalus pri- New South Wales. I know of no record of N. marily on the basis of its bifurcate baculum. timoriensis in eastern Australia north of the However, I know of no study of the bacula Tropic of Capricorn. The only specimens in ofthe various species of Nyctophilus that has the American Museum are from southwest- taken variation into account so that it is cer- ern Western Australia and all but one of the tainly not clear to me how constant the bac- Western Australian specimens in the British ular difference is between bifax and daedalus. Museum are from the same area. The one The only published records I know from exception is unfortunately immature (a male Western Australia ofthe species I have called from Margaret River in southern Kimberley). 28 AMERICAN MUSEUM NOVITATES NO. 2778

Kitchener and Vicker (1981) list a number in molar teeth between the single American of Western Australian specimens (under the Museum skull from northern Queensland and name of major), all but two ofwhich are from the four from southern Queensland, New south of the Tropic of Capricorn. The two South Wales, and Victoria, the type of pa- localities (Mount Bruce and "Weeli Wooli cificus (a male) is smaller in condylobasal Spring") are in the Pilbara and if correctly length than any of the southeastern Austra- identified would constitute additional tropi- lian males I have measured (14.2 vs. 14.4- cal Western Australian records. In view of 15.4). Unfortunately, I have only seen one the fact that the two Willie Wollie Springs male skull from northeastern Australia, the specimens I have seen are N. gouldi, I am type ofpallescens from Alexandria, Northern somewhat skeptical of these records. How- Territory, which has a condylobasal length of ever, in the British Museum there is also an 14.1. It should be noted that in N. geoffroyi, adult female from Port Essington in north- males tend to be smaller than females, though western Northern Territory and Hill and Pratt as far as I know this has not been pointed (1981) recorded two specimens from New out before. In view of Tate's earlier (1941) Guinea, which they tentatively referred to N. remarks, it is odd that in his later (1952a) t. timoriensis. I am inclined to refer the two paper, he made no mention of the name pa- northern Australian specimens (Margaret cificus in reference to northern Queensland River and Port Essington) also to N. t. ti- material. I am therefore inclined to follow moriensis, leaving open the question as to Thomas in using the name pallescens rather whether or not the type of timoriensis ac- than pacificus for the subspecies in northern tually came from Timor. Queensland and the Northern Territory. I Nyctophilus geoffroyi: Tate (1 952a) record- would be inclined to place the geographical ed this species from two localities, but as ex- boundary between pallescens and pacificus in plained above, the Ravenshoe specimen was eastern Australia somewhere between Pent- misidentified and only the Pentland speci- land and the southeastern corner of Queens- men is actually N. geoJfroyi. The only other land. Hall and Richards (1979, p. 61) allo- tropical Australian specimen in the Ameri- cated not only inland Queensland but even can Museum ofNatural History was collected inland New South Wales material to palles- by W. H. Butler at Yarraloola in Western cens, but it is not clear where they would put Australia. Unlike the other Australian species specimens from coastal localities, since they of the genus, whose ranges are more or less mention unicolor only from Tasmania. All confined to the periphery of the continent, Western Australian specimens I have seen I N. geoffroyi occurs across its center (see Hall have allocated to N. g. geoffroyi, though my and Richards, 1979, p. 61; Parker, 1973, p. basis for doing this is very shaky. Thomas 41) and this makes the delimitation of sub- (19 15) distinguished g. geoffroyi from g. pal- species more difficult. Thomas (1915) rec- lescens only on color ("rather dark" vs. "much ognized three subspecies, g. geoffroyi in paler"). All American Museum material of southwestern Western Australia, g. palles- g. geoffroyi (mostly from southwestern West- cens in Northern Territory and northwestern ern Australia) is either in fluid or as a skeleton Queensland, and g. pacificus in southeastern only). The only dry skins of N. geoffroyi in Australia (including Tasmania) and he has the American Museum are the one from been generally followed. Tate (1941 b, p. 504), Pentland (identified as g. pallescens) and one however, on the basis of tooth size of the from Mandurama (near Blayney) in New types, believed that the type ofpacificus (de- South Wales (identified as g. pacificus). These scribed from "Islands of the Pacific") origi- should be very different in color since Thom- nated from northeastern rather than south- as ( 19 1 5) characterized g. pacificus as "dark." eastern Australia. What would appear to be Yet I see very little difference in color be- the consequence of this would be to make tween these two specimens. I have borrowed paci.ficus a senior synonym ofpallescens, with two skins from the Field Museum which come unicolor then becoming the oldest name for from the Kimberley region (northwestern the southeastern subspecies. Although I am Western Australia). They too are very similar unable to detect any relative size difference in color to the Pentland and Mandurama 1 984 KOOPMAN: AUSTRALIAN BATS 29 skins. Thomas (1915) also gave the condy- land," Hill stated that he examined only one lobasal length of g. geoffroyi as 14 and that specimen besides the type. Three subspecies ofg. pallescens as 15.3. My measurements of of planiceps were recognized, the nominate southern and central Western Australian form from "South and west Australia": p. skulls give condylobasal measurements of loriae from Papua, and p. cobourgiana from 13.3-14.1 for males and 13.8-15.2 for fe- the Northern Territory and northern Queens- males. Specimens from the Kimberley region land (Inkerman). Specimens from northeast- (borrowed from the Field Museum) have ern Queensland, which had been identified condylobasal lengths of 13.8-14.5 for eight by Tate (1 952a) as loriae, are mentioned, but males and 14.3 for the single female. None it is not clear whether they should be allo- of my condylobasal measurements of g. pal- cated to p. loriae or to p. cobourgiana. No lescens approaches 15.3, the largest being a mention is made of other northeastern female from Killalpanima (east of Lake Eyre Queensland specimens that Tate referred to and one of the specimens seen by Thomas), norfolcensis (=norfolkensis) but presumably which measures 14.7 (the male type from Al- Hill was inclined to follow Tate in his allo- exandria which Thomas also saw measuring cation, since he included Queensland in the only 14.1). It seems evident that much work range of norfolkensis. Felten (1964) adopted remains to be done to elucidate the subspe- a different arrangement. He agreed with Hill cies of N. geoffroyi, assuming that subspecies in recognizing norfolkensis as a distinct can usefully be recognized. species, but allocated Hill's second specimen (from New South Wales) to planiceps, leaving norfolkensis known only from the type. How- FAMILY MOLOSSIDAE ever loriae (with cobourgiana) was separated During the last two decades, all Australian as a species from planiceps and a new sub- molossids have generally (e.g., Ride, 1970) species (1. ridei) was proposed. To this latter been included in one genus, Tadarida. Free- subspecies were referred all northern Queens- man (1981), however, has made a good case land specimens including those that Tate sep- for dividing it into several of which three arated as norfolkensis and loriae as well as (Mormopterus, Tadarida, Chaerephon) are the Inkerman specimen. Felten also allocated recognized in Australia. There is but a single specimens from Buchanan's Island (which he species each in Tadarida (australis) and places in Shoal Bay, north of Darwin in the Chaerephon (jobensis) and few taxonomic Northern Territory) to 1. cobourgiana, where- problems exist for either ofthese within Aus- as Hill (196 1) had identified them as p. plani- tralia. Mormopterus, however, presents sev- ceps. It is therefore evident that the situation eral serious taxonomic problems involving is more complicated than simply deciding Australian populations, which are still not whether or not planiceps and loriae are con- resolved. specific. The two principal characters by GENUS Mormopterus: I have previously which Felten (1964) distinguished loriae from (Koopman, 1982) discussed some of the planiceps are the less flattened skull and the problems involving this genus in the Austra- presence of a gular pouch in males. I have lian region, but I have since studied the sit- examined all specimens in the American Mu- uation in Australia in considerably greater seum of the planiceps-loriae complex, in- depth. Hill (1961) revised all the Indo-Aus- cluding a number of males in alcohol from tralian members ofthe genus Tadarida (sensu northeastern Queensland. I am unable to find lato) and recognized three Australasian anything that I would call a gular pouch. Only species of Mormopterus. One of these (bec- three skulls are available from within the caril) has been generally recognized as dis- range ofplaniceps as given by Felten (1964). tinct and will be discussed below. The other The two from Victoria are indeed greatly flat- two (norfolkensis and planiceps) have pre- tened, the single skull from Western Australia sented much more of a problem, in part be- less so. (One of the two Western Australian cause of uncertainty as to the type locality of skulls I previously referred to (Koopman, either species. Although the range of norfolk- 1982, p. 24) turns out to be actually M. bec- ensis was given as "Victoria north to Queens- carii.) Felten (1964, p. 3) expressed this as a 30 AMERICAN MUSEUM NOVITATES NO. 2778 regression diagram ofcondylobasal length vs. norfolkensis by "wing membrane attaches height of the skull at the posterior margin of near ankle joint; forearm 28-34 mm. "vs." the palate. Felten's diagram shows loriae and wing membrane attaches approximately 1/3 of planiceps to be well separated, but if I have tibia's length from the ankle; forearm 34-40 measured the Western Australian specimen mm." This is supplemented by additional in- correctly in the American Museum (AMNH formation in the text in that the neck pouch 197172 from Contine), it has a condylobasal is stated to be present in males but rudimen- length of 15.3 and a skull height of3.5, which tary in females of both loriae and norfolk- places it between loriae and planiceps. Fur- ensis. The wing membrane is stated to arise thermore, the skulls offour specimens ofthis from "mid leg" in planiceps, whereas in nor- group in the National Museum of Natural folkensis the wing membrane is confusingly History from Gloucester (northeastern New stated to arise from the ankle. The forearm South Wales) and Farina (northeastern South length is stated to be 32-38 mm. in planiceps. Australia), though well within the geograph- The important characters to check would ical range of planiceps as given by Felten, therefore seem to be presence or absence of show the following range of condylobasal banding on ventral hairs, presence or absence lengths and skull heights (as given by Felten): ofa gular pouch in males, place ofattachment 14.9, 4.1; 15.3, 3.8; 14.9, 3.6; 14.9, 3.3. Only of the wing membrane, and forearm length. the last (a female from Farina) falls into the All American Museum material is clearly planiceps range as given by Felten, the other within the geographical range of either plan- three fall into the loriae range. The two iceps alone (three from Victoria) or loriae American Museum specimens from Victoria alone (50 from northeastern Queensland). I likewise fall outside the planiceps range (14.8, have checked all these specimens for the 4.3; 15.0, 4.1). A specimen in the Field Mu- above characters, skins for ventral hair color, seum of Natural History from Hermanns- alcoholics for gular pouch and wing mem- burg (southern Northern Territory falls bare- brane attachment, and both for forearm ly inside the planiceps range (15.9, 3.5). In length. As stated above, I have been unable short, I cannot duplicate the small skull height to find a gular pouch in any ofthem. The two that Felten gives for most ofhis planiceps on Victorian skins do agree fairly well with Hall any Australian Mormopterus skull available and Richards's (1979) color characterization to me. The ranges of loriae and planiceps as of planiceps, but those from northeastern given by Felten are allopatric (loriae in north- Queensland show considerable variation, the ern Northern Territory and northern Queens- skins Tate (1952a) referred to norfolkensis land, planiceps from New South Wales to (Mossman, Cairns) showing a pronounced southern Western Australia). Hall and Rich- ventral hair banding, whereas those he re- ards (1979, p. 37), however, showed a small ferred to loriae (Helenvale) having it poorly area ofsympatry in southeastern Queensland developed. In the single Victorian alcoholic, (where incidentally they also show Mormop- while the wing membrane attachment could terus norfolkensis occurring). Ifthere are real- hardly be called "mid leg," it is definitely ly three species of Mormopterus in south- above the ankle. The northeastern Queens- eastern Queensland, this is extremely land alcoholics on the other hand show con- interesting and the three taxa should be readi- siderable variability. Most show an attach- ly characterized in this limited area. Unfor- ment at the ankle, several show an attachment tunately, I have difficulties with the charac- higher up, and there is a certain amount of ters of the three species as given by Hall and arbitrariness in scoring. All the specimens I Richards (1979, p. 33). They separated these have studied have a forearm length of less three alleged species on the basis of two cou- than 34 mm. so none seems to conform to plets. First, planiceps is separated from the norfolkensis. All three Victorian specimens other two by "shafts of ventral fur one con- have forearms of slightly more than 32 mm. tinuous color, no neck pouch" vs. "shafts of and seem to be good planiceps. The northeast ventral fur light at base, darker in the center Queensland specimens, however, though on and lighter at the tips; neck pouch may be geographical grounds clearly within the range present." Secondly, loriae is separated from of loriae ridei, show considerable variability 1 984 KOOPMAN: AUSTRALIAN BATS 31 in and in part discrepancy from the characters p. planiceps is definitely known is in southern which Hall and Richards have assigned to Northern Territory, where Parker (1973, p. loriae. Taking account of my problems with 36) shows planiceps (which he distinguished the delimitation of loriae from planiceps as from loriae) extending almost to 22°S. Ban- given by Felten (1964) and by Hall and Rich- nister (1969, p. 70) records planiceps from ards (1979), I remain skeptical of their dis- the Yannarie River (in the Pilbara region of tinction, particularly as sympatric reproduc- Western Australia), just north of the Tropic tively isolated species. Until an analysis of of Capricorn. However, the Mormopterus, the situation in southeastern Queensland is specimen we have from that locality (part of made showing two (or three) clearly distinct the samhe collection that Bannister reports) is, species within this critical area I will continue as explained below, actually M. beccarii. On to include loriae in M. planiceps and thus the other hand, Kitchener and Vicker (198 1) find myself in agreement with Winter and list specimens identified as planiceps, loriae, Allison (1980, p. 34). I am still not clear about and cf. beccarii, all from the Pilbara region. the status of norfolkensis, but at least sensu So this area, like southeastern Queensland, Hall and Richards (1979), its range does not would repay further analysis. Except for these extend north ofthe Tropic ofCapricorn. This two areas, however, there appears to be a leaves two tropical Australian Mormopterus broad gap between the ranges ofp. ridei and species to be considered, M. planiceps and p. cobourgiana to the north and p. planiceps M. beccarii. to the south. Mormopterus planiceps: As discussed Mormopterus beccarii: Until recently (Hill, above, Tate (1 952a) allocated material ofthis 1961; Koopman, 1982), this species was species from northern Queensland to two known only from the Moluccas and New species, "Nyctinomys norfolcensis" and Guinea (including the Louisiades). Win- "Nyctinomus loriae." Combining his records ter and Allison (1980, p. 34) record this for these two, we have four localities. The species from two localities on the Cape York American Museum also has specimens from peninsula (as I would define it), Hill (1983) China Camp and Lord's Prayer (both south mentions two Queensland localities, one in of Cooktown and collected by J. Roberts in the southeast, just south ofthe Tropic ofCap- 1951), Grasstree (J. Roberts in 1952), and ricorn, and Kitchener and Vicker (1981) list Gordonvale (R. F. Peterson in 1959). All these 22 specimens from the Pilbara district. are from northeastern Queensland and are, I Nevertheless, the first specimen that I have believe, referable to M. p. ridei. I might add definitely identified is a single female from that, as indicated above, the skins Tate al- Yannarie River (Western Australia, Pilbara located to the two species do differ in color, district). This specimen (obtained by W. H. and furthermore all "norfolcensis" have fore- Butler in 1965) was previously misidentified arms longer than 30 mm. and all "loriae" as planiceps by both Bannister (1969, p. 70) have forearms shorter than 30 mm. I can and myself(Koopman, 1982, p. 24). It differs make no such separation in the large Ship- from planiceps and agrees better with bec- ton's Flat series of alcoholics and am unable carii in its larger skull, higher brain- to understand how Tate divided the part of case, and more inflated rostrum. It does this series that was then available to him. show differences from available New Guinea Though the American Museum has no spec- material of beccarii, however, in its longer imens to document this, M. planiceps ex- forearm, smaller skull, and less-reduced an- tends across the Northern Territory to West- terior upper premolar (indistinguishable from ern Australia (McKenzie et al., 1977; that of M. planiceps). Measurements of the Kitchener, 1978). In the north, two subspe- Yannarie River female (AMNH 197749) to cies are recognized, p. ridei (in northeastern be compared with Hill's (1961, p. 47) are: Queensland) and p. cobourgiana (in north- forearm (39), condylobasal length (15.7), "in- western Northern Territory). Allocation of terorbital" width (4.1), braincase width (7.4), other northern populations are, however, at m3-m3 width outside alveoli (7.4), c''-c width present uncertain. The only area north of the outside alveoli (4.3), c-m3 length outside al- Tropic ofCapricorn from which the southern veoli (5.7), pm4-m3 length outside alveoli 32 AMERICAN MUSEUM NOVITATES NO. 2778

(4.7). It is evident that the Yannarie River the broad hiatus between them. Tadarida a. specimen has a considerably longer forearm australis barely extends north of the Tropic than any of the Amboina or New Guinea of Capricorn in Queensland (Hall and Rich- material but is shorter in all skull dimensions. ards, 1979), Northern Territory (Parker, While there are, therefore, rather pronounced 1973), and Western Australia (several Pilbara differences from either of the two currently localities listed by Kitchener and Vicker recognized subspecies ofM. beccarii, I would (1981). All American Museum specimens are allocate the Yannarie River specimen to this from farther south in Victoria, southwestern species. The Field Museum has a series of Queensland (Birdsville), and southern West- Mormopterus from the Kimberley district, ern Australia (Elduna; Malura Sork; "wheat previously identified as loriae. With condy- belt"; Shark Bay Turnoff). lobasal lengths of 15.6 to 17.0, m3-m3 widths Chaerephon jobensis: Tate (1952a) record- of 7.2-7.9, and c-m3 lengths of 5.8-6.3, it is ed this species under the name of "Nyctino- evident that they are much too large for loriae mus colonicus" from two localities in north- but agree well with the Yannarie River bec- ern Queensland. The American Museum also carii. Furthermore, I have measured, but not has specimens from Lucinda collected by L. critically compared, a skull in the Museum D. Crossan in 1959. There are also numerous of Comparative Zoology from Coen on the specimens from tropical Western Australia: Cape York Peninsula (MCZ 29109). This is 25 mi. n. Hall's Creek (Butler in 1965); Mt. a male with a condylobasal length of 17.8, Anderson (Butler in 1965); Woodstock (But- an m3-m3 width of 8.4 and a c-m3 length of ler in 1965); Derby (Butler in 1965 and 1973); 6.7. Hill's (1983) measurements ofa Queens- Willie Wollie Spring, ca. 20 mi. wnw. Poonda land male and female are: condylobasal (17.9, (Nelson, Butler, and Rosen in 1969). This is 16.9), m3-m3 width (8.2, 7.3), c-m3 length the only species of Chaerephon in the Aus- (6.5, 6.2). The MCZ skull is considerably tralian region. Felten (1964) recognized four larger than the Yannarie River female skull subspecies ranging from New Guinea and and thus agrees best with a Fergusson Island Western Australia to the Fijis. In Australia, male skull (see Koopman, 1982, p. 24) with C. jobensis appears to be confined to areas, the same condylobasal length. I have been north of the Tropic of Capricorn (Hall and unable to compare the two skulls, however. Richards, 1979; Parker, 1973; Bannister, Probably, the specimen from Coen does rep- 1969), though Kitchener and Vicker (1981) resent a population of beccarii closely related record a specimen from barely south of it. to those ofNew Guinea. (Hill, 1983, allocates The only possible exceptions are old, doubt- his Queensland specimens tentatively to b. ful records from South Australia (Aitken, astrolabiensis.) Waithman (1979) records 1975). The American Museum of Natural beccarii from two localities in the Western History also has an old specimen from the province of Papua, New Guinea, just across Wood Jones collection (AMNH 160320) la- Torres Straits from Cape York, but he gives beled only "South Australia." As Aitken no measurements. (1975) pointed out, these Wood Jones spec- Tadarida australis: I previously (Koop- imens were probably collected at a time when man, 1982) presented evidence for treating the Northern Territory (where the species is the New Guinea kuboriensis as a subspecies well known) was part of South Australia. In ofthe chiefly southern Australian T. australis Australia, the distribution of C. jobensis is and follow Hill (1961) in synonymizing atra- therefore largely allopatric with Tadarida tus with T. a. australis. Incidentally, this re- australis (though both species occur in New moves the presence of a white flank stripe as Guinea). Currently all Australian popula- a diagnostic character of T. a. australis (vs. tions of C. jobensis are allocated to C. j. co- a. kuboriensis) since Thomas (1924), in his lonicus and comparing specimens from as far description of atratus, states that the type apart as Lucinda on the Pacific coast of lacks "any trace of the whitish line along the Queensland, Woodstock, and Willie Wollie edge of the body below," thus leaving size as Spring in northwestern Western Australia re- the only constant character separating the two veals no significant differences. Chaerephon subspecies. Most of tropical Australia lies in jobensis colonicus seems therefore to be an 1 984 KOOPMAN: AUSTRALIAN BATS 33 endemic subspecies, widespread across trop- an extensive distribution in Australia, going ical Queensland, Northern Territory, and across the entire continent, extending from Western Australia. at least the base of the Cape York Peninsula to at least some distance south of the Tropic of Capricorn. Five species (Pteropus alecto, DISCUSSION P. scapulatus, Nyctophilus gouldi, Mormop- Having taken up the various families, gen- terus planiceps, Chaerephon jobensis) also era and species of tropical Australian bats, I occur in New Guinea. The only other way I would now like to discuss their distributions know of in which distributions of this group and relationships under several headings. can be classified is between those species that These are the distributional patterns within tend to avoid the really arid areas (Pteropus tropical Australia and with the three major alecto, Nyctophilus gouldi, Chaerephon job- adjoining areas: temperate Australia, New ensis) and those that do not (Pteropus scapu- Guinea (particularly its south central low- latus, Taphozous georgianus, Saccolaimus lands), and the Lesser Sunda Islands (partic- flaviventris, Macroderma gigas, Eptesicus ularly their largest, nearest, and best worked pumilus, Chalinolobus gouldii, Nyctophilus island, Timor; see Goodwin, 1979). In the geoffroyi, Mormopterus planiceps). As would following analyses, I am omitting three species be expected for such widespread species, the supposed to have come from tropical Aus- latter category predominates. tralia and discussed in the accounts above CAPE YORK SPECIES: Ofthe 25 species which (Pteropus brunneus, Myotis sp. (near austral- are not widespread in tropical Australia but is), Pipistrellus javanicus). All are too poorly occur on at least part of the Cape York Pen- known, at least in Australia, for anything very insula, only six (Dobsonia moluccensis, Sac- useful to be said. colaimus mixtus, Rhinolophus philippinen- sis, Hipposideros cervinus, H. semoni, Murina GEOGRAPHICAL RELATIONSHIPS WITHIN florium) are confined to Cape York in the TROPICAL AUSTRALIA sense that they do not occur south of Towns- ville nor appreciably west of the Great Di- Since tropical Australia extends some 800 viding Range. Two others (Pteropus conspi- mi. north-south and over 2000 mi. east-west cillatus, Taphozous australis) are similarly and exhibits a variety of environments from confined to the eastern tropics (assuming that rain forest to desert, it is not surprising that all more western records of T. australis are its 49 species analyzed here should show a misidentified), but do extend south of the great variety ofdistributional patterns. Aside Cape York Peninsula. Five additional species from 11 widespread species, most of the re- are likewise strictly eastern but reach farther mainder may be said to center on Cape York, south to southeastern Queensland (Nycti- which was probably an important part of the mene robinsoni), New South Wales (Syco- corridor through which a majority ofthe Aus- nycteris australis, Nycticeius ruppellii, Min- tralian bats reached their present distribu- iopterus australis) or Victoria (Rhinolophus tional areas. These 26 species which occur on megaphyllus). Two species (Saccolaimus sac- Cape York are taken up as a group, though colaimus, Hipposideros diadema) though not as will be pointed out, many extend far from extending south of the Cape York Peninsula, Cape York either to the south or the west. do have apparently isolated populations in Among the 12 species of tropical Australian the northern part of the Northern Territory. bats that do not reach Cape York, I have Four additional species (Macroglossus mini- recognized a "western mesic" group of three mus, Hipposideros ater, Pipistrellus tenuis, species, a "southern marginal" group of four Miniopterus schreibersi) have a similar dis- species (which form a group transitional to tribution except that they also reach the Kim- that of the few exclusively temperate species berley region (northeastern Western Austra- that are discussed in the following section), lia). None extend appreciably south of the and finally two species that I have not been Cape York Peninsula except for Miniopterus able to fit into any distributional pattern. schreibersi that extends all the way to Vic- WIDESPREAD SPECIES: Eleven species have toria. Of the remaining species, all except 34 AMERICAN MUSEUM NOVITATES NO. 2778

Mormopterus beccarii are vespertilionids and Australia (south of the Tropic of Capricorn), the known range of M. beccarii (Cape York, there are four (Pteropus poliocephalus, Cha- southeastern Queensland, Kimberley and linolobus picatus, C. morio, Tadarida aus- Pilbara regions of Western Australia) does tralis) which extend a relatively short dis- not fit any known pattern. Of the remaining tance into tropical Australia. Except for T. six species, two (Nycticeius influatus, Keri- australis, these extensions are only in the east voula papuanus) are Queensland endemics and indeed two (P. poliocephalus, C. picatus) (except that K. papuanus, but not N. influa- are confined to eastern Australia. Presumably tus, also occurs in New Guinea). Both occur the absence of desert barriers in the east has on the Cape York Peninsula (N. influatus ap- facilitated northward spread of southern ele- parently only marginally), but barely extend ments. south of the tropics, and both occur in rela- MISCELLANEOUS SPECIES: Two species tively arid regions in northwestern Queens- (Taphozous hilli, Nyctophilus timoriensis) land. Both are also known only by a few spec- cannot be fitted into any ofthe above patterns imens which may explain some of their at present. In the case of N. timoriensis, this apparent resemblances and differences. Ofthe may be resolved once the confusion in the remaining four species, all are fairly wide- records between it and N. gouldi is disentan- spread. Myotis adversus, Chalinolobus nigro- gled. Until this is done, little can be said about griseus, and Nycticeius greyii (the latter only the distributional pattern. Taphozous hilli, marginally in the Cape York Peninsula) are however, does seem to present a unique pat- similar in that all occur across tropical Aus- tern. It is the only Australian bat that appears tralia and extend south ofthe Tropic of Cap- to be confined to arid areas, avoiding more ricorn in the east but not (apparently) in the mesic regions to the north, east, and south. west. Finally, Nycticeius balstoni occurs in both tropical Queensland and tropical West- ern Australia and across temperate Australia, GEOGRAPHICAL RELATIONSHIPS WITH but apparently not across tropical Australia. TEMPERATE AUSTRALIA WESTERN MESIC SPECIES: Three species Excluding the poorly understood Mor- (Taphozous kapalgensis, Eptesicus douglasi, mopterus norfolkensis, five species (Pipistrel- Nyctophilus walkeri, all Australian endemics) lus tasmaniensis, Eptesicus vulturnus, E. reg- are absent from Queensland but are confined ulus, E. sagittula, Chalinolobus dwyeri) are to mesic areas in extreme northern Northern confined to the large part of Australia south Territory and, except for T. kapalgensis, also of the Tropic of Capricorn. As indicated in the Kimberley region (extreme northeast- above, four other species (Pteropus polio- ern Western Australia). Except for the fact cephalus, Chalinolobuspicatus, C. morio, Ta- that they are absent from Cape York, their darida australis) are largely confined in Aus- distributions in Australia are quite compa- tralia to the temperate zone. Since both P. rable with those of Macroglossus minimus, poliocephalus and T. australis have close (and Saccolaimus saccolaimus, Hipposideros ater, in the case of Tadarida conspecific) relatives H. diadema, and Pipistrellus tenuis men- in New Guinea, it is evident that with the tioned above. Three additional species (Hip- exception of Eptesicus and Chalinolobus, posideros stenotis, Rhinonicteris aurantius, temperate Australia has not been an impor- Nyctophilus arnhemensis) have somewhat tant evolutionary center for bats. Of the 24 similar distributions except that they extend remaining species of bats that occur south of into somewhat more arid regions (but not the Tropic of Capricorn, three also belong to into the desert), thus reaching northwestern the genera Eptesicus and Chalinolobus (E. (but not northeastern) Queensland and, in the pumilus, C. nigrogriseus, C. gouldii). Tem- case of Rhinonycteris, also the Pilbara region perate Australia may have been involved in of northwestern Western Australia. H. ste- the endemic Australian radiations of Tapho- notis is closely related to the Cape York H. zous (T. georgianus, T. hilli) Nycticeius (N. semoni. greyii, N. balstoni, N. influatus, N. rueppelli), SOUTHERN MARGINAL SPECIES: Besides a and Nyctophilus (N. gouldi, N. timoriensis, few species which are confined to temperate N. geoffroyi), though the last almost certainly 1 984 KOOPMAN: AUSTRALIAN BATS 35 also involved tropical Australia and even New the Cape York Peninsula with the exception Guinea. The remaining 13 species all have ofNyctophilus timoriensis and Tadarida aus- their closest (usually only congeneric) rela- tralis. As mentioned above, the true distri- tives to the north and are represented in tem- bution of N. timoriensis is poorly known but perate Australia only by peripheral popula- it almost certainly does not occur on the Cape tions. Pipistrellus tasmaniensis is a special York Peninsula. It may have used the western case. It is a very isolated species, not at all part ofthe land bridge since it is known from closely related to the tropical P. tenuis and both the Northern Territory and New Guinea. of highly uncertain relationships within its Tadarida australis is, as discussed above, almost cosmopolitan genus. represented by widely separated populations in Australia (a. australis) and New Guinea (a. kuboriensis) and the precise nature of the GEOGRAPHICAL RELATIONSHIPS WITH former connection is unknown. As men- NEW GUINEA tioned above, five New Guinea species (Dob- New Guinea has a rich tropical bat fauna sonia moluccensis, Saccolaimus mixtus, of some 70 species (see Koopman, 1982), Hipposideros cervinus, H. semoni, Murina much of which occurs in the southern low- florium) have only a limited distribution on lands, separated from Cape York and north- Australia (Cape York Peninsula). Five other eastern Northern Territory (Arnhemland) species (Macroglossus minimus, Saccolai- only by the shallow Torres Straits and Ara- mus saccolaimus, Hipposideros ater, H. dia- fura Sea. These were dry land at various low dema, Pipistrellus tenuis) occur also in trop- sea level periods during the Pleistocene ep- ical areas to the west (Northern Territory, och, the most recent ending only about 10,000 Western Australia). Mormopterus beccarii years ago. Thus a broad land connection, pre- might also be included in this category since, sumably with a varied environment (wetter though poorly known in Australia, it does in the east and dryer in the west) connected occur on the Cape York Peninsula and in the two now separate land masses. It should northern Western Australia, but also in be pointed out that it was, of course, a low- southeastern Queensland. Yet another species land connection so that species confined to (Pteropus conspicillatus) is confined to trop- mountains would not be able to use it. After ical eastern Queensland but does extend south each reunion following a period of separa- ofthe Cape York Peninsula. For all the species tion, however, lowland species would be ex- mentioned above except for Pteropus alecto, pected to disperse in both directions. Since P. scapulatus, Nyctophilus timoriensis, and the original vegetation of New Guinea was Tadarida australis, the main distributional chiefly forest and the original vegetation of areas as well as the ranges of related species Australia chiefly savanna and grassland, dis- are to the north ofAustralia so dispersal from persal from the north mainly has involved New Guinea to Australia is strongly indicat- forest elements, whereas dispersal from the ed. south mainly has involved savanna elements. Besides Pteropus alecto and scapulatus, Of the 49 tropical Australian species here there are four species (Chalinolobus nigro- discussed, 27 also occur in New Guinea, griseus, Nycticeius balstoni, Nyctophilus though Pteropus alecto and P. scapulatus gouldi, N. timoriensis), which have a much barely get across Torres Straits. I am not in- more restricted distribution in New Guinea cluding Taphozous australis in this figure than in Australia. It is therefore probable that since the single New Guinea record (Port all six species reached New Guinea from Aus- Moresby) is almost certainly accidental or er- tralia. As mentioned above, C. nigrogriseus roneous. Four additional species are repre- and N. balstoni are each part ofan Australian sented in the New Guinea region by close radiation. The radiation of Nyctophilus in- relatives: Pteropus poliocephalus (P. macro- volved both areas since each has endemic tis), Nyctimene robinsoni (P. major), Nycto- species lacking in the other (N. walkeri, N. philus arnhemensis (N. microtis), and N. geof- arnhemensis, N. geoffroyi, in Australia; N. froyi (N. microdon). All species that are shared microtis, N. microdon, in New Guinea). between New Guinea and Australia occur on Some eight cases require more extensive 36 AMERICAN MUSEUM NOVITATES NO. 2778 discussion. Syconycteris australis has an ex- it has only a single Australian subspecies. The tensive Australian distribution from Cape subspecies situation to the north ofAustralia York to northeastern New South Wales but is confused, but there are related species (M. only a single subspecies occurs. North ofNew paululus, M. pusillus) which are present in Guinea, the species has a range from the Mo- the north but absent in Australia. Again, a luccas to the Bismarcks and East Papuan Is- northern origin and subsequent dispersal to lands with considerable geographical varia- the south seems the most likely explanation. tion. Moreover a second species (Ziegler, Kerivoula (Phoniscus) papuanus is known by 1982) has been described from New Guinea. only a few specimens from eastern New A northern origin with southward dispersal Guinea and northeastern Australia. How- is therefore indicated. Rhinolophus mega- ever, its closest relatives (K. atrox and K. phyllus has an even more extensive Austra- jagoril) are confined to the Sunda and Wal- lian distribution, from Cape York to Victo- lacean portions of the Indo-Malayan region. ria. However, as explained above, I believe Again, northern origin followed by south- there is only a single subspecies throughout ward dispersal is most likely. Mormopterus this area. While the known New Guinea range planiceps presents a more complicated prob- is small, the species occurs in the Bismarcks lem. Since it has an extensive Australian and and East Papuan Islands and three well- (apparently) a restricted New Guinea distri- marked subspecies can be recognized. Fur- bution, southern origin seems most probable. thermore, several closely related species However, its closest relatives are norfolkensis (simplex, keyensis, borneensis) are known (if indeed this is a distinct species) and bec- from islands to the west of New Guinea. carii. Both are poorly known from a distri- Again, northern origin and southward dis- butional standpoint, so there is some doubt persal are indicated. Myotis adversus has a as to how to treat Mormopterus planiceps. still more extensive Australian distribution Finally, there is Chaerephon jobensis, which from Cape York south to Victoria and also has an extensive tropical Australian distri- west to northeastern Western Australia, but bution, but representing a single subspecies. with only one subspecies. North ofAustralia, The species also has an extensive northern M. adversus has an extensive range from distribution, extending from New Guinea to southeastern Asia to the New Hebrides in the Fijis and in this area is represented by which there are several subspecies. Further- three subspecies. Furthermore, its closest rel- more, there are several close relatives (e.g., ative is C. plicata of the Indo-Malayan re- M. hasselti, M. horsfieldi) on the islands of gion. A northern origin and southward dis- the Sunda shelf. Again, northern origin and persal is therefore most probable. Thus we southward dispersal are indicated. Miniop- see that of the 27 species shared between terus schreibersi (as I delimit this species) has tropical Australia and New Guinea, only a very extensive range from Europe and Af- Pteropus alecto, P. scapulatus, Chalinolobus rica at least as far east as the Solomon Islands. nigrogriseus, Nycticeius balstoni, Nyctophilus Within this area, numerous subspecies are gouldi, N. timoriensis, and perhaps Mormop- recognized and other related species (e.g., M. terus planiceps and Tadarida australis are magnater) also occur. In Australia, there is likely to have moved from Australia. For the (according to my view) only a single species, remaining 19 species, a New Guinea origin with two well-marked subspecies, and a fairly is most probable. This is, of course, to be extensive continental distribution. Miniop- expected since bats show greatest diversity in terus schreibersi tends to avoid the dryer areas, wet tropical areas and New Guinea is much however, and is therefore absent from large richer than Australia in wet tropical habitats. areas of the south and west. It also has no very close relatives in Australia, M. australis belonging to a somewhat different group GEOGRAPHICAL RELATIONSHIPS WITH THE within the genus. Again, northern origin and LESSER SUNDAS southward dispersal seems indicated. Mini- opterus australis has a distribution very sim- The Lesser Sunda Islands, the area which ilar to that of Syconycteris australis and like approaches Australia to the northwest, is here 1 984 KOOPMAN: AUSTRALIAN BATS 37 defined as all the islands from Lombok to exception, there are no species in New Guinea, Tanimbar (=Timorlaut). This area is some- but two in the Lesser Sunda Islands, the latter what arbitrarily defined since the more west- seems the most probable source for the Aus- ern islands show closest mammalogical affin- tralian species. Taphozous melanopogon is ity with Java, whereas the eastemmost islands recorded (Goodwin, 1979) from as far east (particularly Tanimbar) have their closest af- as Timor and there is also a specimen in finity with New Guinea. However, there is the American Museum from the Kei Is- no evidence that any of these islands have lands. Taphozous longimanus leucopleurus been connected with either of the two large was described from Flores and, as mentioned land masses of the Sunda and Sahul shelves. above, was compared in the original descrip- Goodwin (1979) recognized 22 species ofbats tion of T. kapalgensis. Males of T. melano- as almost certainly occurring on Timor and pogon have long, usually black, hair in the there appear to be some 15 additional species gular region and no trace of a gular sac in not known from Timor, but reliably recorded either sex. This condition is unknown in any from other Lesser Sunda Islands. Many of of the Australian species. Males of T. Ion- these 37 species have no close relatives in gimanus, on the other hand, have a naked Australia. In the following discussion only gular area with a well-developed sac, a con- those species which either occur or are rep- dition found in three of the four Australian resented in Australia are considered. species (hilli, kapalgensis, australis). It also Pteropus alecto: Besides Australia and ex- shares small skull size with hilli, broad sphen- treme southern New Guinea, this species oc- oid pits with australis, and a weak anterior curs on several islands from Bawean (on the mandibular emargination with kapalgensis. Sunda shelf) and Celebes to Sawu (=Savu, Therefore, there do seem to be special rela- just west of Timor), over this area being rep- tionships between T. longimanus and at least resented by three subspecies, whereas all three of the Australian species. Whereas ra- Australian (and New Guinea) populations are diation ofa single stock into four species sub- referable to P. a. gouldi. This seems to be a sequent to reaching Australia is most prob- clear instance ofderivation of the Australian able, this stock would seem to have been either (and through them the New Guinea) popu- T. longimanus or a close relative and to have lations from the Lesser Sundas. come from the Lesser Sunda Island area. Macroglossus minimus (=lagochilus): As GENUS Saccolaimus: As discussed above, mentioned above, this species has an exten- there are three species of this genus in Aus- sive distribution from the Malay Peninsula tralia, one endemic (flaviventris) and two to the Solomons, including Timor. Goodwin shared with New Guinea (mixtus, saccolai- (1979) referred Timor specimens to M. m. mus), the latter also having an extensive Indo- lagochilus, whereas Australian specimens are Malayan range, including the Lesser Sunda here referred to M. m. nanus (=pygmaeus). Islands, where it is the only species of Sac- These subspecies are distinguished by An- colaimus. As mentioned above in the New dersen (1912) primarily by the narrower mo- Guinea section, clearly there has been recent lars and premolars of nanus. This distinction interchange between Australian and New seems to hold (although with considerable Guinea involving T. mixtus and T. sacco- variation), the few adult Australian skulls I laimus, but the ultimate derivation of the have seen agreeing with those from New Australasian species is less certain. Certainly Guinea in having narrower cheek teeth than the endemic Australian T. flaviventris is the the few skulls I have seen from Timor. While most distinctive ofthe three species and could a revision of the genus is certainly in order, have had a separate origin from S. saccolai- derivation ofthe Australian populations from mus or a close relative coming into Australia Timor seems considerably less likely than from the Lesser Sundas. Saccolaimus sac- from New Guinea. colaimus itself could have entered Australia GENUS Taphozous: With the exception of from either the Lesser Sundas or New Guinea a single record (probably accidental) of T. (or even perhaps both) since it is now known australis, all four Australian species are en- from the Northern Territory as well as Cape demic to that continent. Since, with the above York. Saccolaimus mixtus, however, judged 38 AMERICAN MUSEUM NOVITATES NO. 2778 by its very restricted Cape York distribution species on Cape York (P. t. papuanus) almost in Australia and its much more extensive New certainly came from New Guinea, P. t. wes- Guinea distribution, almost certainly did not tralis of Western Australia probably came enter Australia from the Lesser Sundas. from the Lesser Sundas. The two lineages have Rhinolophus megaphyllus Subgroup: Be- evidently come together in the Northern Ter- sides megaphyllus, this subgroup (called the ritory and hybridized, since I would interpret simplex subgroup by Tate and Archbold, the two specimens I have seen from the 1939) includes, in the Lesser Sundas, R. sim- Northern Territory as intergrades. plex of Lombok, Sumbawa, and Komodo Murinaflorium: In the Lesser Sundas, this (specimen in the American Museum) and species is known from Sumbawa and Flores. keyensis annectens of Wetar. Though this Too little is known about geographical vari- group is in need of taxonomic revision, it is ation in this species to say whether the Cape evident that neither of these two taxa is as York population is more closely related to closely related to m. megaphyllus (=ignifer) those ofNew Guinea or to the Lesser Sundas. as the three New Guinea subspecies. This, The known Australian distribution, however, plus the absence ofm. megaphyllus from west makes Lesser Sunda derivation highly im- of the Great Dividing Range makes it un- probable. likely that R. megaphyllus reached Australia Nyctophilus timoriensis: There is still con- from the Lesser Sundas. siderable doubt as to whether or not this Rhinolophus philippinensis: Known rec- species really occurs on Timor (see Hill and ords for this species are very spotty. Besides Pratt, 1981). Even ifthis old unsubstantiated R. p. robertsi of Cape York, there is R. p. record is correct it would obviously have achilles of the Kei Islands, and the newly de- nothing to do with the origin ofthe otherwise scribed R. p. montanus of Timor (Goodwin, endemic Australia-New Guinea genus Nyc- 1979, p. 112), together with several subspe- tophilus. cies from farther north. The species has not, This concludes the species or species groups as yet, been recorded from New Guinea. The that are shared between Australia and the distances are so great among these separate Lesser Sundas. Of these only two (Pteropus areas that it seems impossible, at this time alecto and Pipistrellus tenuis sewelanus-wes- to decide by what route R. p. robertsi reached tralis) show clear evidence of recent deri- its present range. vation of Australian populations from con- Hipposideros diadema: This widespread specific ones on the Lesser Sundas. For two species is represented by H. d. diadema in others (Taphozous, Myotis (Selysius)), an the islands from Java to Timor. As men- earlier derivation with subsequent speciation tioned above, this subspecies is quite differ- on Australia seems the most likely explana- ent from either of the two Australian sub- tion. For all others, either derivation from species. Furthermore, the subspecies closest the Lesser Sundas is improbable, or the evi- to Timor (H. d. inornatus) is particularly dif- dence is equivocal. ferent from it, whereas H. d. reginae of Cape York is very similar to H. d. pullatus of the THE ORIGIN OF THE AUSTRALIAN New Guinea mainland. Derivation of the BAT FAUNA Australian populations from the Lesser Sun- das is highly improbable. It is now generally agreed (see e.g., Hallam, Myotis (Selysius): As discussed above, this 198 1) that early in the Cenozoic, Australia lay subgenus probably does not occur in New far to the south of its present position and Guinea but is represented by M. muricola in close to ifnot actuallyjoined with Antarctica. the Lesser Sundas (east to Flores and Sumba). Both the monotremes and the marsupials It is therefore likely that the very poorly clearly have had a long history in Australia. known australis and the Kimberley species The former may even have been in Australia were derived from the Lesser Sundas. since the breakup of Pangaea in the Jurassic, Pipistrellus tenuis: I have discussed this judged by the extreme distance of their re- species at length above. Although the sub- lationship with other mammals and their re- 1 984 KOOPMAN: AUSTRALIAN BATS 39

striction (as far as we know) to Australia and hypothesis that some part of the Australian New Guinea (the southern part of which was bat fauna was there when Australia com- part ofAustralia before it moved northward), menced its northward drift in the Eocene (or either as living animals or fossils. From reached there across broad expanses of open known fossil evidence, it is likely that the ocean afterward) highly improbable. Rather marsupials did not reach Australia much be- a percolation of numerous species across a fore the beginning ofthe Cenozoic. Ifso, they series of relatively narrow water gaps from probably had to get across at least one (albeit the north and west, starting in the Miocene, narrow) water gap from South America via seems indicated. Some of these species that Antarctica, since otherwise it is difficult to reached Australia relatively early either dif- understand why edentates and South Amer- ferentiated into endemic genera or radiated ican ungulates (which also occurred in South out into a group of endemic species (Tapho- America at the beginning of the Cenozoic) zous, Eptesicus, Chalinolobus, Nycticeius). In did not also reach Australia. Between the be- the case of Taphozous, there is still a related ginning of the Eocene and sometime in the species on islands nearby, whereas in the oth- Miocene (some 30 million years) it is prob- er three (as also with Nyctophilus whose ra- able that monotremes and marsupials were diation surely also involved New Guinea), the only land mammals in Australia. there is now a wide gap between the ranges Starting sometime in the Miocene, how- of the Australian species and their closest ever, Australia (including what is now south- Asian (or African) relatives. ern New Guinea) came close to a northern Prior to 1982, the few known pre-Pleis- chain of islands extending from southeastern tocene fossil bats were represented by too lit- Asia through what is now northern New tle material to determine their affinities. Sige Guinea and beyond. Whereas a number of et al. (1982), however, have now described a groups ofplacental mammals presumably oc- Miocene species from northwestern Queens- curred at the western end of this chain, only land which they refer to the otherwise French the murid rodents and five families of bats Miocene subgenus Hipposideros (Brachip- occurred on the more eastward islands with posideros). What is more, they point out a which the Australian plate became associ- special resemblance between this subgenus ated. This is shown by the absence of any and the living endemic Australian genus Rhi- other placentals from the Moluccas, New nonycteris. I have compared their figures with Guinea, or the Bismarcks with the exception skulls ofthe three groups ofHipposideros now ofrecent human introductions. Both the bats in Australia, together with the three other and the murid rodents had reached Australia Australian region hipposiderines (Anthops, by the Pliocene and, as we shall see, at least Aselliscus, Rhinonycteris) and agree with one bat species is now known as fossil from them. Thus it would appear that we can now the Miocene. trace the origin of one of the two endemic One characteristic of both the rodent and Australian bat genera back to the Miocene bat faunas ofAustralia (and also New Guinea) when it was still referable to an apparently is the low level of endemism, which hardly widespread Old World subgenus. Further- goes above the tribal level even when the more, Hill (1982) has regarded Rhinonycteris Bismarcks, Solomons, and Moluccas are as being specially related to the African added. Thus, among the bats, the tribe Nyc- Cloeotis and Triaenops. Therefore, all three tophilini (including Nyctophilus and Pharo- genera could have been derived from the tis), which is almost endemic to Australia and widespread Miocene Brachipposideros. Mac- New Guinea, is probably the most distinct. roderma, the other endemic Australian bat As mentioned above, there are only two en- genus, is most closely related to the Indo- demic Australian genera of bats (Macroder- Malayan Megaderma, which extends east at ma, Rhinonycteris); only three in New Guinea least to Celebes and perhaps to the Moluccas, (Aproteles, Paranyctimene, Pharotis); and no but is not known from either the Lesser Sun- additional genera endemic to Australia and das or New Guinea. Macroderma is known New Guinea together. To me this makes any fossil from the Pleistocene, when it occurred 40 AMERICAN MUSEUM NOVITATES NO. 2778 farther south in Australia than today (Ham- long to two endemic sugenera (Scoteanax and ilton-Smith, 1966), but it is not known how Scotorepens). The subgenus NyctiCeius in- long it has been in Australia. cludes the remaining species but these are Of the other endemic Australian species, I widely separated from those ofAustralia and have already discussed Pteropus polio- from one another. In Africa and extreme cephalus (an Australian representative of the southwestern Asia there is N. (N.) schlieffeni New Guinea P. macrotis), Nyctimene robin- whereas in North America and Cuba there is soni (an Australian representative of N. ma- the N. (N.) humeralis group. As with Chali- jor of the New Guinea area), the species of nolobus, this is obviously a relict distribution the genus Taphozous, Saccolaimus flaviven- with occurrence at one time in southern and tris, Hipposideros stenotis (a dry country rep- eastern Asia including the Malay archipelago. resentative of the mesic Australia-New Nyctophilus (along with its close endemic Guinea H. semoni), and Myotis australis and New Guinea relative, Pharotis) has generally its Kimberley representative. I have no idea been associated taxonomically with the North what is the closest relative of the endemic American and Cuban Antrozous and the southern Australian Pipistrellus tasmanien- Middle American Bauerus (Engstrom and sis. This leaves the endemic Australian rep- Wilson, 1981). A closely related early Pleis- resentatives of the genera Eptesicus, Chali- tocene genus, Anzanycteris, from North nolobus, Nycticeius, and Nyctophilus. America has also been described (White, As discussed above, though the Australian 1969). However, considerable doubt has been Eptesicus seem to be closely related among expressed as to whether the North American themselves; their relationship to other Ep- and the Australasian genera are really each tesicus is far from clear. They show no special others closest relatives or whether instead they relationship to the geographically closest are independently derived from Nycticeini. species (in mainland southeast Asia) and may It has even been suggested (see Pine, Carter, be more closely related to the Eptesicus ca- and LaVal, 1971) that the Antrozous group pensis group of Africa (see Koopman, 1975, is specially related to the Palearctic nycticeine p. 405). In view ofour ignorance ofEptesicus genus Otonycteris, to which it bears a strong relationships, it is difficult to say. Another external resemblance. In any case, it is clear possibility is that in view of the probably that the Nyctophilus group has no close rel- polyphyletic origin of Eptesicus (see Koop- atives in the Indo-Malayan region at present. man, 1975, p. 406), Australian "Eptesicus" It has been in New Guinea and Australia long are really independently derived from a small enough to have undergone a considerable Pipistrellus such as P. tenuis, which would adaptive radiation, including the evolution explain its wide geographical separation from ofa New Guinea endemic (Pharotis imogene) other Eptesicus. that is currently regarded as a separate genus. I previously (Koopman, 1971) discussed my reasons for uniting Glauconycteris APPENDIX: LOCALITIES OF of Africa with Chalinolobus subgenerically. TROPICAL AUSTRALIAN BATS IN The subgenus Chalinolobus is, of course, not THE AMERICAN MUSEUM OF endemic to Australia but also occurs in NATURAL HISTORY southeastern New Guinea and on three Pa- cific islands (New Caledonia, Norfolk, New Pteropus alecto gouldi Zealand). There is, however, a great gap be- QUEENSLAND: no locality (AMNH 108867- tween Australia and Africa where there is at 108868); Brown's Creek, Pascoe River (AMNH present no member of the genus Chalinolo- 154550-154553, 155005); Coen (AMNH 108860- if I am there almost cer- 108863); Lockerbie, 10 mi. wsw. Somerset (AMNH bus, though, correct, 154547); Normanton (AMNH 183581); Portland tainly were representatives in southern Asia Roads (AMNH 155548-155549). at one time, which have since disappeared. WESTERN AUSTRALIA: Kalumburu (AMNH I previously (Koopman, 1978) discussed 197132); Mitchell River (AMNH 236505- the Australian species of Nycticeius in rela- 236508); Parry Creek (AMNH 197123-197131); tion to other members of the genus. Briefly, Tambrey (AMNH 197133-197135); Tunnel Creek the Australian (and New Guinea) species be- (AMNH 197136-197138, 236510-236511). 1984 KOOPMAN: AUSTRALIAN BATS 41

Pteropus conspicillatus conspicillatus land Roads (AMNH 154717-154719); Possession QUEENSLAND: locality uncertain (AMNH Island (AMNH 154714-154715); Wenlock, Ba- 156963,156966-156967); Babinda Creek (AMNH tavia River (AMNH 154723-154728). 66150-66155); Coen (AMNH 108864); Cook- town (AMNH 155318); Julattan (AMNH 154242- Taphozous georgianus troughtoni 154246); Lake Barrine (AMNH 107332-107334); QUEENSLAND: Chillagoe Caves (AMNH Peach River (AMNH 154565-154568, 154570); 66145, 154847-154849, 183556-183557); Lappa Shipton's Flat (AMNH 155303-155317, 155319- Junction (AMNH 183558-183560); Mungana 155324). Caves (AMNH 183551-183555); Rifle Creek, Mt. Isa (AMNH 162707-162708); Pentland (AMNH Pteropus scapulatus 107727, 107769-107770, 109268-109269); QUEENSLAND: Byfield (AMNH 162661- Quamby (AMNH 107720-107725, 107731, 162662); Coen (AMNH 108865-108866,153500, 107733-107737, 107740-107741, 107746- 154571-154572); Cooktown (AMNH 155318); 107754, 107761, 109263-109267); Mt. Etna, Flaggy (AMNH 156964-156965); Green Hills Rockhampton (AMNH 162706). (AMNH 156989-156708); Hann River (AMNH NORTHERN TERRITORY: 3 mi. n. Kather- 154573-154577); Helenvale (AMNH 154579- ine (AMNH 160387-160390, 220075); 18 mi. n. 154585); Peach River (AMNH 154564, 154569); Wollogorang (AMNH 183437-183440, 183537- Shipton's Flat (AMNH 154578, 156968-156988); 183550). Wenlock, Batavia River (AMNH 154554-154563). WESTERN AUSTRALIA: Frazier Downs Taphozous georgianus georgianus (AMNH 197139-197142); Ningbing (AMNH WESTERN AUSTRALIA: Barrow Island 197143-197149); Parry Creek (AMNH 197150- (AMNH 197558); Black Elvire River (AMNH 197152); Stockade Creek (AMNH 236509). 197176-197177); 20 mi. n. Callowa (AMNH 236515); Inglis Gap (AMNH 197559); Kalum- Dobsonia moluccensis magna buru (AMNH 197179-197180); King Edward QUEENSLAND: Coen (AMNH 154589); Iron River (AMNH 236525-236529); Manning Creek Range (AMNH 154586-154588); Peach River (AMNH 236530-236536); Mt. Anderson (AMNH (AMNH 154590-154591). 197175,236537); Napier Downs (AMNH 236538- 236540); Ningbing (AMNH 197181); Parry Creek Nyctimene robinsoni (AMNH 197182); Peawah, Mundabulangana (AMNH 197183-197185); Tambrey (AMNH QUEENSLAND: Captain Billy Creek, "Heath- 197186-197188, 197560, 236516-236518); Tun- lands" (AMNH 222758-222759); Mission Beach nel Creek (AMNH 197189-197190, 236541- (AMNH 196643-196644); Portland Roads 236542); Whim Creek (AMNH 197191-197193); (AMNH 154759); Shipton's Flat (AMNH 196630, Willie Wollie Spring (AMNH 216130-216132); 222726). Wittenoom Gorge (AMNH 236519-236524); Woodstock (AMNH 197194-197199); Yardie Macroglossus minimus nanus Homestead (AMNH 197200). QUEENSLAND: Seagren's Farm (AMNH 154741-154743). Taphozous hilli WESTERN AUSTRALIA: Mitchell River NORTHERN TERRITORY: Tennant Creek (AMNH 236512-236514). (AMNH 160450). Syconycteris australis australis Saccolaimus mixtus QUEENSLAND: no locality (AMNH 108843, 108845, 108848, 108856-108857, 108859); Ship- QUEENSLAND: Brown's Creek, Pascoe River ton's Flat (AMNH 222725). (AMNH 154720-154722). Taphozous australis Saccolaimus saccolaimus nudicluniatus QUEENSLAND: no locality (AMNH 153516); QUEENSLAND: Babinda Creek (AMNH Albany Island (108826-108828); Cairns-Moss- 66144). man Road, Rex's Lookout (AMNH 154850- 15485 1); Coen (AMNH 154729-154731, 154852- Saccolaimus flaviventris 154855); Cowie Bay Cave (AMNH 156946- QUEENSLAND: Malbon (AMNH 107729, 156958); Newcastle Bay (AMNH 154716); Port- 107742); Pentland (AMNH 107755-107760). 42 AMERICAN MUSEUM NOVITATES NO. 2778

WESTERN AUSTRALIA: Inglis Gap (AMNH 162663-162666); Cromarty (AMNH 162667- 197174); Ningbing (AMNH 197201-197202). 162668); Lyndhurst Station (AMNH 183510- 183514). Macroderma gigas (incl. saturata) Rhinolophus philippinensis robertsi QUEENSLAND: Mt. Etna, Rockhampton (AMNH 162669-162674). QUEENSLAND: Helenvale (AMNH 157034); NORTHERN TERRITORY: 18 mi. w. Wol- Mt. Amos, Phoenician (AMNH 157055-15707 1). logorang (AMNH 183435). WESTERN AUSTRALIA: Kalumburu (AMNH Hipposideros ater aruensis 197210); Koolan Island(AMNH 197203-197209, QUEENSLAND: Lockerbie (AMNH 155666- 236543); 20 mi. s. Marble Bar (AMNH 197211); 155668); Shipton's Flat (AMNH 155394,194173, Tunnel Creek (AMNH 236544-236546). 196631); Bramston Beach (AMNH 194224- 194229). Rhinolophus megaphyllus megaphyllus (incl. ignifer) Hipposideros ater gilberti QUEENSLAND: no locality (AMNH 108829- NORTHERN TERRITORY: 13 mi. s. Kath- 108836, 108839-108842, 108844, 108846- erine (AMNH 160376-160377). 108847, 108849-108850, 108852, 108854- WESTERN AUSTRALIA: Koolan Island 108855, 108858); Mt. Carbine (AMNH 107365); (AMNH 197212). Coen (AMNH 108818-108825, 154959-154996, 220076); Cairns (AMNH 154592-154593); Julat- Hipposideros cervinus cervinus tan-Mossman Road (AMNH 154594-154595, 154904-154908); Portland Roads (AMNH QUEENSLAND: Newcastle Bay (AMNH 154596-154600); Iron Range (AMNH 154601- 154669); Somerset (AMNH 154670-154685, 154615, 154909-154952); Pascoe River, Brown's 154857-154878); Thursday Island (AMNH Creek (AMNH 154616); upper Nesbit River 154686-154687); Iron Range (AMNH 154691- (AMNH 154617-154620); upper Peach River 154696, 154708, 154881-154883, 154885- (AMNH 154621-154624); Laura (AMNH 154888); Coen (AMNH 154889-154897). 154626); Mt. Finnigan (AMNH 154627-154630, 154997-154999); Shipton's Flat (AMNH 154631- Hipposideros semoni 155000, 155251-155255, 155257- 154632, QUEENSLAND: no locality (AMNH 108851); 155259, 155261-155269, 155271, 155273, Coen (AMNH 154709); Cooktown (AMNH 155275, 155277-155280, 155282, 155284- 154856); Iron Range (AMNH 154707); Upper 155285, 155287-155288, 155291, 155293, 154710-154712). 155295-155298, 155302); Walter Hill range Nesbit River (AMNH (AMNH 154664); Irvine Bank (AMNH 154898- 154903); Wenlock (AMNH 154953-154958); Hipposideros stenotis Cooktown-Laura railroad (AMNH 155001- NORTHERN TERRITORY: 18 mi. w. Wol- 155003); Collingwood (AMNH 155256, 155260, logorang (AMNH 183441-183442, 183563- 155276, 155299); Home Rule (AMNH 155270, 183564). 155283, 155286, 155290, 155301, 157009- 157011); Grass Tree (AMNH 155272, 155274, 155281, 155292, 155294, 155300, 155407- Hipposideros diadema reginae 155414); Helenvale (AMNH 155289); Boiling QUEENSLAND: Coen (AMNH 108853); Springs (AMNH 155404-155405); Mt. Poverty Cairns (AMNH 154665); Chillagoe (AMNH (AMNH 155406); Stucky's Gap (AMNH 157012); 183565-183578); Iron Range (AMNH 154688- Wyalla (AMNH 157013); Green Hills (AMNH 154690, 154697-154706, 154879-154880, 157014); Ayton (AMNH 156924, 157015- 154884); Shipton's Flat (AMNH 160284, 194174- 157018); Middle Normanby (AMNH 157019- 194175). 157021); Mt. Amos, Phoenician (AMNH 157022- 157053); Lappa Junction (AMNH 183445- Rhinonycteris aurantius 183446, 183515-183520); Chillagoe (AMNH 183509); Mt. Cook (AMNH 194155-194158); NORTHERN TERRITORY: 16 mi. s. Kath- Endeavor Bridge (AMNH 194159-194170); erine (AMNH 199979-199980). Mclvor (AMNH 194171-194172); Innisfail WESTERN AUSTRALIA: Koolan Island (AMNH 194230-194238); Mt. Etna (AMNH (AMNH 197213-197216). 1984 KOOPMAN: AUSTRALIAN BATS 43

Myotis adversus macropus 197727); Montebellos, Hermite Island (AMNH QUEENSLAND: Cairns (AMNH 154635- 197698, 197702); Nodswell Creek (AMNH 154636, 155010). 197621); 6 mi. ne. Yardie Homestead (AMNH WESTERN AUSTRALIA: Mitchell River 197710-197726). (AMNH 236547-236577). Eptesicus douglasi Pipistrellus tenuis papuanus WESTERN AUSTRALIA: Inglis Gap (AMNH QUEENSLAND: Archer River (AMNH 197596); Langey Crossing (AMNH 197598- 154655); Brown's Creek (AMNH 154654); Upper 197620; Mt. Anderson (AMNH 197566-197569, Peach River (AMNH 154656). 236604-610); Derby (AMNH 236578-236596). Pipistrellus tenuis westralis Chalinolobus nigrogriseus nigrogriseus WESTERN AUSTRALIA: Cape Bossut (AMNH 216134-216137). QUEENSLAND: Seagren's Farm (AMNH 154661-154663); Shipton's Flat (AMNH 155230, Eptesicus pumilus pumilus 155232-155234, 155401-155402); Karumba (AMNH 160243). QUEENSLAND: Alderbury, 25 mi. out on Cooktown-Laura Railway (AMNH 154659); Black Mountain, 15 mi. s. Cooktown (AMNH 154657- Chalinolobus nigrogriseus rogersi 154658); Chillagoe (AMNH 154760-154762); QUEENSLAND: 24 mi. s. Burketown (AMNH China Camp, south ofCooktown (AMNH 194176- 183431). 194185); Helenvale, south of Cooktown (AMNH NORTHERN TERRITORY: 18 mi. w. Wol- 156959-156962); Iron Range (AMNH 154651- logorang (AMNH 183432, 183579); 12 mi. e. 154653); Irvinebank (AMNH 154772-154774, Coolibah (AMNH 216194). 155015); Lappa Junction (AMNH 183536); Lock- WESTERN AUSTRALIA: Ningbing (AMNH erbie, 8 mi. wsw. Somerset (AMNH 155012); 196711, 197261-197262); Parry Creek (AMNH Mungana caves (AMNH 183534-183535); Pent- 196710); North Creek (AMNH 216195-216196). land (AMNH 107728, 107790-107795); Koom- booloomba Creek, 30 mi. s. Ravenshoe (AMNH 183383, 183495-183496). Chalinolobus gouldi venatoris QUEENSLAND: Pentland (AMNH 107764- Eptesicus pumilus caurinus 107765, 107800, 109275); Malbon (AMNH QUEENSLAND: Mount Isa (AMNH 162702- 109274). 162705); Quamby (AMNH 107726, 107732, WESTERN AUSTRALIA: Inglis Gap (AMNH 107739, 107772-107789, 109270-109273). 197248); La Grange Dam (AMNH 107250- NORTHERN TERRITORY: 3-5 mi. n. Kath- 107251); Cape Bossut (AMNH 216183-216193). erine (AMNH 160378-160386, 160391-160400, 220084-220085). Nycticeius greyi WESTERN AUSTRALIA: Inglis Gap (AMNH QUEENSLAND: Pentland (AMNH 107796- 197575); Kalumburu (AMNH 197577-197586); 107798, 109278); Malbon (AMNH 109279); 24 Parry Creek (AMNH 197622-197630); King Ed- mi. s. Burketown, Gregory River (AMNH 183429); ward River (AMNH 236597); Koolan Island between Camooweal and Mt. Isa (AMNH 196642). (AMNH 236598-236603); Napier Range (AMNH NORTHERN TERRITORY: 18 mi. w. Wol- 236611-236621); Tunnel Creek (AMNH 236622- logorang (AMNH 183430, 183432-183433); 12 236623); Peawah, Mundabullangana (AMNH mi. e. Coolibah, Victoria River (AMNH 216138). 197631-197636); Tambrey (AMNH 197637- WESTERN AUSTRALIA: Cape Bossut 197642, 236632-236639); Whim Creek (AMNH (AMNH 216141-216146, 216150-216157, 197643); Woodstock Station (AMNH 197644- 216176- 197650, 197705-197709); 20 mi. n. Callowa 216161-216166, 216168-216174, (AMNH 236624-236631); Wittenoom Gorge 216179, 236645). (AMNH 236640-236642). Nycticeius balstoni sanborni Eptesicus pumilus ssp. QUEENSLAND: Seagren's Farm, 10 mi. w. WESTERN AUSTRALIA: Barrow Island Cooktown (AMNH 154660); Cooktown (AMNH (AMNH 197570-197573, 197699-197704, 154775). 44 AMERICAN MUSEUM NOVITATES NO. 2778

Nycticeius balstoni caprenus Kerivoula (Phoniscus) papuensis WESTERN AUSTRALIA: Frazier Downs QUEENSLAND: Shipton's Flat (AMNH (AMNH 197263); La Grange (AMNH 197264- 155403). 197267); Ningbing (AMNH 197268); Yannarie River (AMNH 197269-197271); Cape Bossut Nyctophilus arnhemensis (AMNH 216139-216140, 216147-216149, 216158-216160, 216167, 216175, 216180, QUEENSLAND: 24 mi. s. Burketown (AMNH 236643-236644); Yeeda Creek (AMNH 216181). 183436). NORTHERN TERRITORY: Adelaide River, Tortilla Flats (AMNH 222760). Nycticeius rueppelli WESTERN AUSTRALIA: Cape Bossut QUEENSLAND: Ravenshoe, Koombooloom- (AMNH 216197-216200, 216682-216685). bah Creek (AMNH 183376). Nyctophilus gouldi bifax Miniopterus schreibersi oceanensis QUEENSLAND: Ravenshoe (AMNH 66147); Atherton Tableland, 9 mi. sse. Ravenshoe (AMNH QUEENSLAND: Cunjeboi (AMNH 107340- 66146); Mt. Finnegan (AMNH 154749); Upper 107352); Mt. Spurgeon (AMNH 107353-107364); Nesbit River (AMNH 154746); Shipton's Flat Pentland (AMNH 107766-107768, 107799, (AMNH 154747-154748, 154750, 155227- 109276); Port Stuart (AMNH 108805-108816); 155229, 155395-155397, 160244, 194192- Somerset (AMNH 108817); Cairns (AMNH 194194, 222727); Jackeroo (AMNH 155398- 154633-154634); Thursday Island (AMNH 155400). 154637-154650, 154776-154826); Walter Hill Range (AMNH 154744); Possession Island Nyctophilus gouldi daedalus (AMNH 154745, 154751-154753, 154835- 154846); Chillagoe (AMNH 154827-154834); WESTERN AUSTRALIA: Willie Wollie Spring, Lockerbie (AMNH 155011, 155013); Shipton's ca. 20 mi. wnw. Poonda (AMNH 216686-216687). Flat(AMNH 155226,194190-194191); Mt. Amos, Phoenician (AMNH 156904-156923); Ayton Nyctophilus geoffroyi pallescens (AMNH 156925-156945); Koombooloomba QUEENSLAND: Pentland (AMNH 109277). Creek (AMNH 183374-183375, 183468-183471); Lyndhurst Station (AMNH 183478-183479); Nyctophilus geoffroyi geoffroyi Green Hills (AMNH 194188); Cooktown (AMNH 194189); Bramston Beach (AMNH 194240- WESTERN AUSTRALIA: Yarraloola (AMNH 19424 1); Ingham (AMNH 194242-194245); 5 mi. 197279). se. Coen (AMNH 220088). Mormopterus planiceps ridei Miniopterus schreibersi orianae QUEENSLAND: Cairns (AMNH 154736- NORTHERN TERRITORY: Kintore Cave, 154740, 155014); Helenvale (AMNH 154756- north of Katherine (AMNH 220087). 154758); Mossman (AMNH 154732-154735); WESTERN AUSTRALIA: Ningbing (AMNH Shipton's Flat (AMNH 155235-155250,155419- 197272-197274); Parry Creek (AMNH 197275- 155432); Grasstree (AMNH 155415-155418); 197276); Mitchell River (AMNH 236752- Lord's Prayer, south of Cooktown (AMNH 236779); Port Warrender (AMNH 236780). 194195); China Camp, south of Cooktown (AMNH 194196); Gordonvale (AMNH 183589). Miniopterus australis australis Mormopterus beccarii ssp. QUEENSLAND: Port Stuart (AMNH 108804); WESTERN AUSTRALIA: Yannarie River Upper Peach River (AMNH 157754); Shipton's (AMNH 197749). Flat (AMNH 154755, 155231, 194186-194187); 4 mi. s. Lappa (AMNH 154763-154771); Cowie Chaerephon jobensis colonicus Bay (AMNH 156903); 10 mi. n. Rockhampton (162679-162701); Chillagoe (AMNH 183472- QUEENSLAND: Lucinda (AMNH 160294- 183475); 3 mi. w. Mt. Garnet (AMNH 183476- 160295); Pentland (AMNH 109281); Malbon 183477); Mungana (AMNH 183480-183483); In- (AMNH 107762-107763, 107801-107804, nisfail (AMNH 194239). 109280). 1984 KOOPMAN: AUSTRALIAN BATS 45

WESTERN AUSTRALIA: Derby (AMNH 1981. Systematics of Antrozous dubiaquercus 197155-197170,236787-236798; 25 mi. n. Hall's (Chiroptera: Vespertilionidae), with Creek (AMNH 197173); Mt. Anderson (AMNH comments on the status of Bauerus Van 197154); Woodstock (AMNH 197171); Willie Gelder. Ann. Carnegie Mus., vol. 50, Wollie Spring (AMNH 216688-216691). pp. 371-383. Felten, H. LITERATURE CITED 1964. Zur Taxonomie indo-australischer Fle- dermause der Gatting Tadarida (Mam- Aitken, P. malia: Chiroptera). Senck. Biol., vol. 45, 1975. Two new bat records from South Aus- pp. 1-13. tralia with a field key and checklist to Freeman, P. D. the bats ofthe state. Sth. Australian Nat., 1979. Specialized insectivory: beetle-eating vol. 50, pp. 9-15. and moth-eating molossid bats. Jour. Allen, G. M. Mammal., vol. 60, pp. 467-479. 1933. Two new bats from Australia. Jour. 1981. A multivariate study of the family Mo- Mammal., vol. 14, pp. 149-151. lossidae (Mammalia: Chiroptera): mor- Andersen, K. phology, ecology, evolution. Fieldiana 1912. Catalogue of the Chiroptera in the col- Zool., n.s., no. 7, i-vii + 173 pp. lection of the British Museum. 2nd ed., Goodwin, R. E. vol. 1: Megachiroptera. London, i-ci, 1979. The bats of Timor: systematics and 854 pp. ecology. Bull. Amer. Mus. Nat. Hist., Bannister, J. L. vol. 163, pp. 73-122. 1969. A list of the species of mammals col- Hall, L. S. lected by W. H. Butler for the Archbold 1981. The biogeography ofAustralian bats. In Collections ofthe American Museum of Keast, A. (ed.), Ecological biogeography Natural History and for the Western of Australia. The Hague-Boston-Lon- Australian Museum 1963-66. W. Aus- don, vol. 3, pp. 1557-1583. tralian Mus. Ann. Rpt., 1966-1967. Hall, L. S., and G. C. Richards Barghoorn, S. F. 1979. Bats of eastern Australia. Queensland 1977. New material of Vespertiliavus Schlos- Mus. Booklet, no. 12, i-vi, 66 pp. ser (Mammalia, Chiroptera) and sug- Hallam, A. gested relationships of Emballonurid 1981. Relative importance of plate move- bats based on cranial morphology. ments, eustasy, and climate in control- Amer. Mus. Novitates, no. 2618, pp. 1- ling major biogeographical changes since 29. the Early Mesozoic. In Nelson, G., and Carpenter, S. M., J. L. McKean, and G. C. Rich- D. E. Rosen (eds.), Vicariance biogeog- ards raphy: a critique. New York, pp. 303- 1978. Multivariate morphometric analysis of 330. Eptesicus (Mammalia: Chiroptera) in Hamilton-Smith, E. Australia. Australian Jour. Zool., vol. 1966. The geographical distribution of Aus- 26, pp. 629-638. tralian cave dwelling Chiroptera. Int. Daniel, M. J. Jour. Speleology, vol. 11, pp. 91-104. 1975. First record of an Australian fruit bat Hill, J. E. (Megachiroptera: Pteropodidae) reach- 1961. Indo-Australian bats of the genus Ta- ing New Zealand. New Zealand Jour. darida. Mammalia, vol. 25, pp. 29-56. Zool., vol. 2, pp. 227-231. 1963. A revision of the genus Hipposideros. Dobson, G. E. Bull. Brit. Mus. (Nat. Hist.) Zool., vol. 1878. Catalogue of the Chiroptera in the col- ll,pp. 1-129. lection ofthe British Museum. London, 1965. Asiatic bats ofthe genera Kerivoula and i-xlii, 867 pp. + 30 pls. Phoniscus (Vespertilionidae), with a note Douglas, A. M. on Kerivoula aerosa Tomes. Mamma- 1962. Macroderma gigas saturata (Chirop- lia, vol. 29, pp. 524-556. tera, Megadermatidae). A new subspe- 1966. A review of the genus Philetor (Chirop- cies from the Kimberley Division of tera, Vespertilionidae). Bull. Brit. Mus. Western Australia. Western Australia (Nat. Hist.) Zool., vol. 14, pp. 371-387. Nat., vol. 8, pp. 59-61. 1971. Bats from the Solomon Islands. Jour. Engstrom, M. D., and D. E. Wilson. Nat. Hist., vol. 5, pp. 573-581. 46 AMERICAN MUSEUM NOVITATES NO. 2778

1982. A review of the leaf-nosed bats Rhi- 1971. Taxonomic notes on Chalinolobus and nonycteris Cloeotis and Triaenops (Chi- Glauconycteris (Chiroptera, Vespertili- roptera: Hipposideridae). Bonn. Zool. onidae). Amer. Mus. Novitates, no. Beitr., vol. 33, pp. 165-186. 245 1, pp. 1-10. 1983. Bats (Mammalia: Chiroptera) from 1973. Systematics of Indo-Australian Pipis- Indo-Australia. Bull. Brit. Mus. (Nat. trellus. Period. Biol. (Zagreb), vol. 75, Hist.) Zool., vol. 45, pp. 103-208. pp. 113-116. Hill, J. E., and K. F. Koopman 1975. Bats ofthe Sudan. Bull. Amer. Mus. Nat. 1981. The status of Lamingtona lophorhina Hist., vol. 154, pp. 353-444. McKean and Calaby, 1968 (Chiroptera: 1978. The Genus Nycticeius (Vespertilioni- Vespertilionidae). Bull. Brit. Mus. (Nat. dae), with special reference to tropical Hist.) Zool., vol. 41, pp. 275-278. Australia. In Olembo, R. J., F. B. Cas- Hill, J. E., and T. K. Pratt telino, and F. A. Mutere (eds.), Pro- 1981. A record of Nyctophilus timoriensis ceedings ofthe Fourth International Bat (Geoffroy, 1806) (Chiroptera: Vesper- Research Conference. Nairobi, pp. 165- tilionidae) from New Guinea. Mam- 171. malia, vol. 45, pp. 264-266. 1982. Results of the Archbold Expeditions. Iredale, T., and E. L. G. Troughton No. 109. Bats from Eastern Papua and 1934. A check-list of the mammals recorded the East Papuan Islands. Amer. Mus. from Australia. Australian Mus. Sydney Novitates, no. 2747, pp. 1-34. Mem., no. 6, xi + 122 pp. Laurie, E. M. O., and J. E. Hill Jenkins, P. D., and J. E. Hill 1954. List of Land Mammals of New Guinea, 1981. The status of Hipposideros galeritus Celebes and adjacent islands 1758-1952. Cantor, 1846 and Hipposideros cervinus London, 175 pp. + 3 pls. (Gould, 1854) (Chiroptera, Hipposide- Lekagul, B., and J. A. McNeely ridae). Bull. Brit. Mus. (Nat. Hist.) Zool., 1977. Mammals ofThailand. Bangkok, i-li + vol. 41, pp. 279-294. 758 pp. Johnson, D. H. Maeda, K. 1959. Four new mammals from the Northern 1982. Studies on the classification of Miniop- Territory of Australia. Proc. Biol. Soc. terus in Eurasia, Australia, and Mela- Washington, vol. 72, pp. 183-187. nesia. Honyurui Kagaku (Mammalian 1964. Mammals of the Arnhem Land Expe- Science), suppl. no. 1, pp. 1-176. dition. In Spect, R. L. (ed.), Records of McKean, J. L. the American-Australian scientific ex- 1966. Some new distributional records of pedition to Arnhem Land, vol. 4, Zo- Broad-Nosed Bats (Nycticeius spp.). ology. Melbourne, pp. 427-515. Victorian Nat., vol. 83, pp. 25-30. Kitchener, D. J. 1 970a. A new subspecies of the horseshoe bat, 1976. Eptesicus douglasi, a new Vespertilio- Hipposideros diadema from the North- nid bat from Kimberley, Western Aus- ern Territory, Australia. West. Austra- tralia. Rec. Western Australia Mus., vol. lianNat., vol. 1,pp. 138-140. 4, pp. 295-301. 1 970b. Status of Myotis australis (Dobson). 1978. Mammals of the Ord River area, Kim- Australian Bat. Res. News, no. 9, pp. berley, Western Australia. Ibid., vol. 6, 4-6 (cited by permission of McKean). pp.189-219. 1 972a. Notes on some collections ofbats (order 1980. Taphozous hilli sp. nov. (Chiroptera, Chiroptera) from Papua-New Guinea Emballonuridae), a new Sheath-tailed and Bougainville Island. CSIRO Div. bat from Western Australia and North- Wildlife Res. Tech. Paper, no. 26, pp. ern Territory. Ibid., vol. 8, pp. 161-169. 1-35. Kitchener, D. J., and S. A. Halse 1972b. A Nyctimene (Chiroptera: Pteropodi- 1978. Reproduction in female Eptesicus reg- dae) specimen supposedly from Wee ulus (Thomas) (Vespertilionidae) in Jasper, New South Wales. Australian South-western Australia. Aust. Jour. Mammal., vol. 1, p. 47. Zool., vol. 26, pp. 257-267. 1975. The bats of Lord Howe Island with the Kitchener, D. J., and E. Vicker description of a new nyctophiline bat. 1981. Catalogue of modern mammals in the Ibid., vol. 1, pp. 329-332. Western Australian Museum 1895 to McKean, J. L., and G. R. Friend 1981. Perth, 184 pp. 1979. Taphozous kapalgensis, a new species Koopman, K. F. of Sheath-tailed Bat from the Northern 1984 KOOPMAN: AUSTRALIAN BATS 47

Territory, Australia. Victorian Nat., vol. 1970. A guide to the native mammals ofAus- 96, pp. 239-241. tralia. Melbourne, i-xiv, 249 pp. McKean, J. L., G. R. Friend, and A. L. Hertog Sanborn, C. C. 1980. Occurrence of the Sheath-tailed Bat 1931. Bats from Polynesia, Melanesia, and Taphozous saccolaimus in the Northern Malaysia. Field Mus. Nat. Hist., Zool. Territory. N. Terr. Nat., vol. 4, p. 20. Ser., vol. 18, pp. 7-29. McKean, J. L., and A. L. Hertog Sanborn, C. C., and W. J. Beecher 1979. Extension ofrange in the Horseshoe Bat. 1947. Bats from the Solomon Islands. Jour. Ibid., vol. 1, p. 5. Mammal., vol. 28, pp. 387-391. McKean, J. L., and W. J. Price Sige, H., S. Hand, and M. Archer 1967. Notes on some Chiroptera from 1982. An Australian Miocene Brachipposi- Queensland, Australia. Mammalia, vol. deros (Mammalia, Chiroptera) related 31,pp. 101-119. to Miocene representatives from France. 1978. Pipistrellus (Chiroptera: Vespertilioni- Palaeovertebrata, vol. 12, pp. 149-172. dae) in Northern Australia with some Tate, G. H. H. remarks on its systematics. Ibid., vol. 1941a. Results of the Archbold Expeditions. 42, pp. 343-347. No. 39. Review of Myotis of Eurasia. Bull. Amer. Mus. Nat. Hist., vol. 78, McKean, J. L., G. C. Richards, and W. J. Price pp. 537-565. 1978. A taxonomic appraisal of Eptesicus 1941 b. Results of the Archbold Expeditions. (Chiroptera: Mammalia) in Australia. No. 40. Notes on Vespertilionid Bats. Australian Jour. Zool., vol. 26, pp. 529- Ibid., vol. 78, pp. 567-597. 537. 1942. Results of the Archbold Expeditions. McKenzie, N. L. No. 48. Pteropodidae (Chiroptera) ofthe 1983. Mammals. Wildlife Res. Bull. Western Archbold Collections. Ibid., vol. 80, pp. Australia, no. 11, pp. 40-53. 331-347. McKenzie, N. L., A. Chapman, W. K. Youngson, 1952a. Results of the Archbold Expeditions. and A. A. Burbidge No. 66. Mammals of Cape York Pen- 1977. The mammals of the Drysdale River insula with notes on the occurrence of National Park, North Kimberley, West- rain forest in Queensland. Ibid., vol. 98, ern Australia. Wildlife Res. Bull. West- pp. 563-616. ern Australia, no. 6, pp. 79-86. 1952b. Results of the Archbold Expeditions. Parker, S. A. No. 67. A new Rhinolophus from 1973. An annotated checklist ofthe native land Queensland (Mammalia, Chiroptera). mammals of the Northern Territory. Amer. Mus. Novitates, no. 1578, pp. Rec. S. Australian Mus., vol. 16, pp. 1- 1-3. 57. Tate, G. H. H., and R. Archbold 1939. Results of the Archbold Expeditions. Peterson, R. L. No. 24. Oriental with spe- 1981 a. Systematic variation in the tristis group Rhinolophus, bats of the cial reference to material from the Arch- of the bent-winged genus bold Collections. Ibid., no. 1036, pp. 1- Miniopterus (Chiroptera: Vespertilioni- 12. dae). Canadian Jour. Zool., vol. 59, pp. Thomas, 0. 828-843. of a new 1981b. The systematic status of Miniopterus 1894. Diagnosis Pteropus from the australis and related forms. Bat. Res. Admiralty Islands. Ann. Mag. Nat. Hist., vol. 48. ser. 6, vol. 13, p. 293. News, 22, p. 1904. New species of Pteropus, Mus, and Po- Pine, R. H., D. C. Carter, and R. K. LaVal gonomys from the Australian region. 1971. Status of Bauerus Van Gelder and its Novitates Zool., vol. 11, pp. 597-600. relationships to other nyctophiline bats. 1915. Notes on the Genus Nyctophilus. Ann. J. Mammal., vol. 52, pp. 663-669. Mag. Nat. Hist., ser. 8, vol. 15, pp.493- Richards, G. C., L. H. Hall, P. M. Helman, and 499. S. K. Churchill 1924. A new subspecies of Nyctinomus aus- 1982. First discovery of a species of the rare tralis. Ibid., ser. 9, vol. 1 5, pp. 455-456. tube-nosed insectivorous bat (Murina) Thompson, B. G. in Australia. Australian Mammal., vol. 1982. Records of Eptesicus vulturnus (Thom- 5, pp. 149-151. as) (Vespertilionidae: Chiroptera) from Ride, W. D. L. the Alice Springs area, Northern Ter- 48 AMERICAN MUSEUM NOVITATES NO. 2778

ritory. Australian Mamm., vol. 5, pp. 1979. A report on a collection of mammals 69-70. from southwest Papua, 1972-1973. Troughton, E. L. G. Australian Zoologist, vol. 20, pp. 313- 1925. A revision ofthe Genera Taphozous and 326. Saccolaimus (Chiroptera) in Australia White, J. A. and New Guinea, including a new 1969. Late Cenozoic bats (subfamily Nycto- species, and a note on two Malayan philinae) from the Anza-Borrego Desert forms. Rec. Australian Mus., vol. 14, of California. Univ. Kans. Mus. Nat. pp. 313-341. Hist., Misc. Publ., no. 51, pp. 275-282. 1930. A new species and subspecies of fruit- Winter, J. W., and F. R. Allison bats (Pteropus) from the Santa Cruz 1980. The native mammals of Cape York group. Ibid., vol. 18, pp. 1-4. Peninsula-changes in status since the 1967. Furred animals of Australia. 9th ed. 1948 Archbold Expedition. In Stevens, Sydney, i-xxxii, 384 pp. N. C., and A. Bailey (eds.), Contem- Van Deusen, H. M., and K. F. Koopman porary Cape York Peninsula. Brisbane, 1971. Results of the Archbold Expeditions. pp. 3 1-44. No. 95. The genus Chalinolobus (Chi- Ziegler, A. C. roptera, Vespertilionidae). Taxonomic 1982. The Australo-Papuan genus Syconyct- review of Chalinolobus picatus, C. nig- eris (Chiroptera: Pteropodidae). With the rigriseus, C. rogersi. Amer. Mus. No- description ofa New Papua New Guinea vitates, no. 2468, pp. 1-30. species. Occas. Papers Bernice P. Bish- Waithman, J. op Mus., vol. 25, no. 5, pp. 1-22.