^ I L H Moon & Son Bookbinders ) i j RECORDS OF THE WESTERN AUSTRALIAN MUSEUM Volume 10, Part 1, 1982 Records of the Western Australian Museum Editorial Committee Chairman Natural Science Human Studies Publications Officer P.F. Berry G.R. Allen KJ. McNamara LM, Crawford GJ. Henderson M.E. Lofgren A.N. Browne The Records of the Western Australian Museum (Rec. West. Aust. Mus.) is published irregularly with up to four parts appearing each year. A series of supplements, usually devoted to recording the basic data and results of specific Museum studies, is also produced. The journal and supple- ments are available for sale or exchange, the current price per part being $2.50 plus postage. All but a few back issues are available and orders and enquiries should be addressed to the Book- shop, Western Australian Museum, Francis Street, Perth, W.A. 6000, Australia. Cover Holotype of Tympanocryptis lineata houstoni, drawn by Christine Bruderlin. © Western Australian Museum, 1982 ISSN 0312 3162 Published by the Western Australian Museum, Francis Street, Perth, Western Australia 6000. Printed in Western Australia by Frank Daniels Pty Ltd. Rec. West Aust. Mus. 1982 , 10 ( 1 ): 1-9 Four New Lerista (Lacertilia: Scincidae) from Western and South Australia G.M. Storr* Abstract 'rhe new taxa are L. picturata edivardsae, a member of the L. macropisthopus group from South Australia, and L. greeri, L. griffim and L. vertnicularis, members of the L. bipes group from Western Australia. Introduction Lerista, second-largest genus of Australian skinks, continues to yield undescribed species and subspecies. For descriptions of the L. macropisthopus and L. bipes groups and of the close relatives of the new taxa see Storr ( 1972, 1976). This paper is based on material in the South Australian Museum (specimens cited without prefix), Australian Museum (AM) and Western Australian Museum (WAM). New Taxa Lerista picturata edivardsae subsp. nov. Figure 1 Holotype R 17787 in the South Australian Museum, collected by G. Harold and T.M.S. Hanlon on 18 November 1979 in low open Eucalyptus-Casuaritia -woodland on brown limestone soil 6 km SEof Streaky Bay, South Australia, in 82°50'S, 134° 15 'E. Paratypes Ninety-seven specimens in the South Australian, Australian and Western Australian Museums. For details see Material. Diagnosis A subspecies of L. picturata with minute foreleg, didactyl hindleg and strong colour pattern, including 2 (rarely 4) dark dorsal lines and dark upper lateral stripe. Differing from L. p. picturata (Fry, 1914) in its much shorter foreleg, fewer dorsal lines (usually 2, V. 4) and fewer midbody scale rows ( 18-20, v. 20-22), and from L. p. bayriesi Storr, 1972 by its much stronger colour pattern and more numerous subdigital lamellae (9-14 under second toe, v. 8-10). * Department of Ornithology and Herpetology, Western Australian Museum, Francis Street, Perth, Western Australia 6000. 1 Four New Lerista from Western and South Australia Description Foreleg reduced to a papillose tubercle (N 85); hindleg bearing two well-developed toes (N 80). Snout-vent length (mm): 57-95 (N 89, mean 72.9). Length of appendages {9c SVL): hindleg 10.8-19.2 (N 80, mean 15.2); tail 75-100 (N 29. mkn 87.5); snout to foreleg 20.5-29.0 (N 81, mean 25.5). Nasals in contact (N 81). Prefrontals widely separated (N 81). Frontoparietals separated, each much smaller than interparietal (N 81 ). f^sually no nuchals, occasionally one or two on each side. Supraoculars 5, first and second in contact with frontal (N 42) or onlv the second in contact (48). SupraciliariesO -f 1 (N 7), 0 -h 2 (68)or0 + 5 (14). Temporals 2 (N 1) or 5 (85 ); upper secondary larger than primary (N 72) or subequal to it (7) or smaller (5); lower secondary much the smallest except in one specimen where it is fused to primary. Upper labials 5 (N 1) or 6 (85); first count due to fusion of second and third. Midbody scale rows 18 (N 48), 19 (4) or 20 (25). Lamellae under longer toe 9-14 (N 77, mean 11.2). Upper surface pale brownish-grey (silvery-grey in life, at least in holotype) to greyish- brown. Two blackish dorsal lines from neck to tip of tail, each passing through a series of para\ ertebral scales; occasionally outside of these a series of blackish dorsal spots which rarely {59c of specimens) coalesce to form an additional pair of lines. Head variably marked with blackish, most markings taking form of a thick discontinuous margin to scales. Wide blackish upper lateral stripe from nasal to end of tail. Posterior edge of upper labials thickly margined with blackish-brown. Lower labials sometimes brown- Figure 1 Holotype oi Lensta picturata edwardsae photographed in life by Gregorv Harold. 2 G.M. Storr Distribution Arid and semi-arid lowlands of southern South Australia west of the Flinders and Mt Lofty Ranges (Figure 2). Geographic Variation The most distinctive population is that inhabiting the Nuyts Archipelago. These skinks are noticeably longer and stouter than those from elsewhere, e.g. SVL 75-95 mm (N 14, mean 87. 1; v. 37-90, N 75, mean 70.3). They share two traits with the population 3 Four New Lerista from Western and South Australia on the Investigator Group (also off the west coast of Eyre Peninsula), namely short hindleg (11.2-14.8% of SVL,N 18, mean 13.2; v. 10.8-19.1 N 62, mean 15.8 elsewhere) and short snout to foreleg (20.5-23.6% of SVL, N 18, mean 22.1; v. 21.6-29.0, N 63, mean 23.9 elsewhere). The populations east and west of Spencer Gulf dif fer in two respects. On average the eastern skinks are darker, partly due to the greater extent of the head markings. They are also notable for the high frequency of individuals with only one supraocular in contact with frontal (88%, v. 33% west of Spencer Gulf). Remarks Geographically, the subspecies L. p. picturata, L. p. baynesi and L. p. edwardsae form a west-to-east sequence. Morphologically, however, they do not form a sequence. The various character gradients from picturata to baynesi are steep, but none of them extends to edwardsae. Presumably baynesi and edwardsae have evolved independently from picturata. Indeed the latter and edwardsae may yet prove to be in contact in Western Australia north of the Nullarbor Plain. Derivation of Name After Adrienne Edwards of the Department of Herpetology, South Australian Museum. Material South Australia Ooldea (1794); 23 km N Koonibba Mission (3r42'S, 133°26T) (15018); S. Childara Paddock (31°37'S, 134°32'E) (13745); 12 km SSE Lake Everard HS (WAM R69930-I): near Ceduna (3722, 4294); West 1., Nuyts Archipelago ( 15983); St Francis 1., Nuyts Archipelago (715 a-g, 12880 a-/); Pine Lodge HS (1 1705-6, 15579 a-6); Thurlga Woolshed (32°40'S, 135°40'E) (1 1703-4); Minnipa Hill (549); Buckleboo (1588 a-c, 8799); Lake Gilles Conservation Park (14381, 17989); False Bay (6083); W hyalla (14350) and 25 km NW (19061) and 15 km N (18707-8); near Minaro Downs HS (14193. 14203 a-e)\ Carappee Rocks (5752); Mt Wedge (3358); Flinders L, Investigator Ciroup (10211. 10225, 10227); Pearson L, Investigator Group (10235); Reevesbv L. Sir Joseph Banks Group (12988; A.\I R79708); Port Lincoln (AM R6378-9, AM R6380 a-b)\ 24-28 km SSE Port Augusta (14270, 14272 a-b, 14279, 14297); Mambray Creek (13965); between Port Piiie and Port Augusta (5457 a-e); 6 km N Port Germein (12427 a-d, 12593); Port Pirie (12373): 3 km W Warnertown (10850-1); Port Broughton (9416, 14134); Price 11982-4); Ardrossan (11978-81); 7 km NW Stansbury (12499); Tiddy Widdy Beach, Yorke Peninsula (16982 a-h); Port Wakefield (4294); Parham (9298, 14604 a~b, 15613); 8 km W Dublin (14035): Port Prime (17875, 17878). Lerista greeri sp. nov. Holotype R23005 in the Western Australian Museum, collected by G.M. Storr and A.M. Douglas on 1 September 1964 at 8 km SSE of Derby, W.A., in 17°22'S, 123°4()'E. Paratypes Kimberley Division (W.A.) Lake Argyle (WAM R400()l-2); 23 km NNE Dunham River HS (WAM R23080); Point Torment (WAM R58571); Derby (WAM R20295-6, 20333-7) and 24 km SSE (WAM R182 10, 32169, 32343); 4 G.M. Storr Christmas Creek (VVAM R57149); Wolf Creek Meteorite Caater (WAM R64044-5); (irannv Soak, Cardiner Range (19°()7'S, 128°5!VE) (WAM R51239, 51242). Diagnosis A member of the L. bipes ^roup with movable eyelid, two toes and no trace of forelimb (including groove). Most like L. hipes (Fischer, 1882) and L. lahialis Storr, 1972, but distinguishable from hipes by its 6 (rather than 5) upper labials and 20 (rather than 18) midbody scale rows, and from labialis by having 2 (rather than 1) supraocular in contact with frontal, and 1 or 2 supraciliaries (rather than none). Description Snout-vent length (mm): 26-62 (X 20, mean 46.8). Length of appendages {% SVL): hindleg 13.9-21.2"(N 16, mean 17.0); tail 72-91 (N 6, mean 84). Nasals narrowly separated (N 18) or in short contact (2). No prefrontals (apparently fused to second loreal, as in L. bipes and L. labialis). Nuchals 1-3 (N 20, mean 2.0^ Supraoculars 3, first two in contact with frontal. Supraciliaries 0 + 1 (N 15) or 0 + 2 (4). Loreals 2, second high with acute apex. Preocular 1. Temporals 3, upper secondary Figure 3 Map of northern Western Australia showing location of specimens Lerista greeri, L. grifpni and L. vermicularis. 5 Four New Lerista from Western and South Australia usually largest, lower secondary much the smallest. Upper labials 6 (N 20). Midbody scale rows 19 (N 1) or 20 (19). Lamellae under longer toe 8-10 (N 20, mean 8.9). Upper surface pale brown or pale reddish brown marked with dark brown as follows: scattered spots or smudges on head; a line of dots through each paravertebral series, dots sometimes coalescing to form a narrow stripe; occasionally a laterodorsal series of faint dots; and a wide upper lateral stripe from nostril to end of tail. Ventral and lower lateral surfaces whitish. Distribution Far northern Western Australia (semi-arid zone of south and east Kimberley), mainly | on sandy and loamy soils (Figure 3). Remarks This species is morphologically intermediate between L. bipes and L. labialis; it has the supraoculars and supraciliaries of the former, and the upper labials and midbody scale rows of the latter. Lerista greeri could thus be close to the common ancestor of these species. However it is now probably sympatric with both of them. In the Kimberley bipes generally ranges north to Coulomb Point, the Edgar Ranges and Gregory Salt Lake, i.e. to the south of greeri, but there is an isolated population considerably further north (if WAM R37698 from Manning Creek is correctly identified as bipes). In the Kimberley, labialis has been collected in the southern semi-arid zone near Mt North, Mt Percy and Fitzroy Crossing. A single specimen (WAM R188) collected in November 1913 by W.B. Alexander on one of the Wallabi Islands, Houtman Abrolhos, is indistinguishable in scutellation from L. greeri. However it is now entirely devoid of coloration. Derivation of Name After Allen E. Greer of the Australian Museum. Lerista griffini sp. nov. Figure 4 Holotype R75543 in the Western Australian Museum, collected by P. Griffin and G. Harold on 12 March 1981 at Kununurra, Western Australia, in 15°47'S, 128°44'E. Paratypes Kimberley Division (W.A.j Ninbing (WAM R27913): Point Springs, Weaber Range (WAM R26774); 18 km NE Kimberley Research Station (WAM R17105); 21 km SE Kununurra (WAM R23108) and 37 km SE (WAM R231 13); Martins Well (8 km S Lombadina) (WAM R60849-50): mainland opposite Packer I, (WAM R60904); 4 km S to 7 km N Coulomb Point (WAM R58472, 60844, 60855-6, 60870, 60906-7, 61362); Broome (WAM R1257, 14112 a-b, 29159). Northern Territory Bullo River HS (WAM R60337). 6 G.M. Storr Figure 4 Holotype Lerista griffini photographed in life by Philip Griffin. Diagnosis A member of the L. bipes group with movable eyelid, two toes and no trace of forelimb (including groove). Most like L. bipes, L. greeri and L. labialis, but differing in its greater size, stouter body, darker coloration, low Hat-topped second loreal and lack of preocular. Description Snout-vent length (mm); 47-67 (N 22, mean 57.2). Length of appendages (% SVL): hindleg 12.9-18.3 (N 21, mean 15,5); tail 68-106 (N 10, mean 91). Nasals very narrowly separated (N 16) or in point contact (3). No prefrontals (apparently fused to frontal). Nuchals 0-3 (N 21, mean 1.7). Supraoculars 3, first two in contact with frontal. Supraciliaries 0 + 1 (N 4) or 0 + 2 (18). Loreals 2, second much wider than high. No preocular (fused to second loreal). Temporals 3, upper secondary slightly larger than or subequal to primary, lower secondary much the smallest. Upper labials 5 (N 22). Midbody scale rows 20 (N 21) or 21 (1). Lamellae under longer toe 8-11 (N 22, mean 9.2). 7 Four New Lerista from Western and South Australia Upper surface brown or reddish-brown marked with dark brown as follows: scattered spots or smudges on head; a paravertebral stripe or line of dots from occiput to end of tail; occasionally a laterodorsal series of small indistinct spots; and a wide upper lateral stripe from nostril to end of tail. Ventral and lower lateral surfaces whitish. Distribution Disjunct in far north of Western Australia: semi-arid zone of west Kimberley (Dampier Land) and of east Kimberley (lower Ord valley), and extending from latter into extreme north-west of Northern Territory (Figure 3). Remarks This species is sympatric with L. bipes on the west coast of Dampier Land and parapatric (perhaps marginally sympatric) with L. greeri in the Cockatoo Springs-Lake Argyle region. Derivation of Name After Philip Griffin, formerly of the Western Australian Museum, in appreciation of his excellent assistance in 1979-81. Lerista vermicularis sp. nov. Holotype R73814 in the Western Australian Museum collected on 17 August 1981 by MJ. Bamford at Dragon-tree Soak. Western Australia, in 19°39'S, 123''23'E. Paratypes Eastern Division (W.A.) McLartv Hills (19°30'S. 123°30'E) (WAM R73815-6); Joanna Spring (2()°()5'S, 124°12T) (WAM R738I8). N orth-Western Division (W.A.) 63 km SE Wallal (WAM R6()157). Diagnosis A small slender member of the L. bipes group with immovable eyelid, two toes, and no trace of forelimb (including groove). Most like L. labialis, L. greeri and L. bipes, but lacking a preocular and lower secondary temporal, and having an immovable eyelid, smaller supraoculars, body angular at change from lateral to ventral surface, and dark upper lateral stripe not so well dehned. Further distinguishable from L. bipes and L. green by absence of supraciliaries, and from L. lahialis by fewer labialis (5, v. 6) and midbody scale rows (16-18, v. 20). Distinguishable fromL. griffimhy fewer supraoculars (2, v. 3), supraciliaries (0, v. 1-2), temporals (1-2, v. 3) and midbody scale rows (16-18, v. 20-21). Description Snout-vent length (mm): 32-41.5 (N 5, mean 37.8). Length of appendages (% SVL): hindleg 15.2-21.3 (N 5, mean 18.3), tail 73-89 (N 4. mean 80). Nasals separated. No prefrontals. Frontal slightly wider than long. Nuchals 1-2 (N 5, 8 G.M. Storr mean ] .8). Supraocuiars 2. small, first in contact with frontal No supraciliaries. Loreals 2. No preocular (fused to second loreal). Temporals usually 2, primary the smaller and occasionally fused to fourth labial Upper labials 5. Midbody scale rows 16 (N 2) or 18 (3). Lamellae under longer toe 7-11 (N 5, mean 8.8). Upper surface buff or very pale reddish brown, marked with dark brown as follows: 1-5 small irregular spots on frontonasal and frontal; 0-4 lines of small spots on back, passing through centre of dorsal scales, central (paravertebral) pair extending on to tail, where spots become larger, darker and more closely spaced; and upper lateral stripe from nasal to end of tail broken on head, variably defined on body, more conspicuous on tail Ventral and lower lateral surfaces whitish. Distribution Crests of dunes in Great Sandy Desert (northern interior of Western Australia) (Figure 3). Derivation of Name From Latin vermiculus (little worm). Acknowledgements I am grateful to Drs T.D. Schwaner and A.E. Greer for the loan of specimens in the South Australian and Australian Museums respectively. I am also grateful to Dr Greer for pointing out the diagnostic characters of Lerista greeri and L. griffini. References Storr, G.M. (1972). The genus Lerista (Lacertilia, Scincidae) in Western Australia./. Proc. R. Soc. West. Aust. 54 : 59-75. Storr, G.M. (1976). Revisionary notes on the Lerista (Lacertilia, Scincidae) of Western Australia. Rec. West. Aust. Mus. 4 : 241-256. Received 6 March 1981 Accepted 17 December 1981 Published 30 June 1982 9 Rec. West. Aust. Mus. 1982 , 10 ( 1 ): 11-34 The Distribution of Earthworms in the Perth Metropolitan Area Ian Abbott* Abstract Thirteen species of earthworm were collected on the coastal plain near metropolitan Perth between 1977 and 1980. Six of these species were introduced following European settlement in 1829. The present known distribution of each species in the metropolitan region is mapped at a scale of 1:400 000; that of the more common species in temperate Western Australia is mapped at a smaller scale (1:8 500 000). All introduced species were found only in disturbed habitats (gardens, man-made parks, etc.) but only two of the seven native species were confined to undisturbed habitat, principally woodland or swampland. Settlement of the region by European man has resulted in the fragmentation of the range of some of the native species, but the replacement of woodland by gardens has enabled the introduced species to establish. It is unlikely that the local decline in distribution of native earthworm species in the metropolitan area is a result of the introduction of peregrine earthworm species. Each species is keyed out using external features. Introduction In 1905 Professor Wilhelm Michaelsen, one of the world’s authorities on the taxonomy of earthworms, spent six months collecting in the settled districts of temperate Western Australia as part ol the Hamburg/south-western Australia expedition. Although metropolitan Perth then comprised a small area (Figure 1), Michaelsen examined 23 sites in and around suburban Perth as well as six localities in the Darling Range to the immediate east (Michaelsen 1907). These collections were important for two reasons: Perth had been settled by Europeans only 76 years earlier, and outside the suburban limits there were still extensive areas of pristine Banksia woodland. Michaelsen recorded hve species around Perth: one lumbricid Aporrectodea trapezoides (as Helodrilus caliginosus [Savigny, 1826]) and five megascolecids Diplotrema cornigravei (as Eodrilus cornigravei Michaelsen, 1907), Graliophilus levis (as Plutellm levis Michaelsen, 1907), Microscolex dubius ' (Fletcher, 1887) and Woodwardiella libferti (as Woodwardia libferti Michaelsen, 1907). Because lumbricid earthworms are not native to Australia (Jamieson * Department of Soil Science and Plant Nutrition, University of Western Australia, Nedlands, Western Australia 6009. Present address: Institute of Forest Research and Protection, Hayman Road, Como, Western Australia 6152. Michaelsen also recorded Microscolex phosphorem (Duges, 1837). According to Jamieson (1974a), this species is indistinguishable from M. dubius. 11 Earthworms in the Perth Metropolitan Area Figure 1 Approximate suburban limits of’Perth metropolitan area. 1905 (dotted area) and 1979 (inside dashed line). After Morison 1979 and Department of Lands and Surveys. Perth, 1979. ▲ Sites sampled in 1905 (Michaelsen 1907): •, sites sampled 1977-1980. 12 Ian Abbott 1981), it is certain that A. trapezoides was introduced after 1829. In addition, it is almost certain that M. duhius was also introduced (Ljungstrdm 1972). The remaining species are indigenous to Western Australia. In the late I92()s, Ada Jackson collected at a small but undisclosed number of sites ‘in the neighbourhood of Perth’, and added five species to Michaelsen’s list (Jackson 1931). These were the lumbricids Eisenia fetida^ (as E. foetida (Savigny, 1826) and Eiseniella tetraedra^ (Savigny, 1826), and the megascolecids Amynthas corticus^ (as Pheretima heterochaeta (Michaelsen, 1903), Megascolex imparicystLs Michaelsen, 1907 and M. longicystis Nicholls and Jackson, 1926. Both Michaelsen and Jackson described new species collected from the Darling Range behind Perth: Graliophilns strelitzi (as Plutelhis strelitzi Michaelsen, 1907), Wo/avfofex /2()ri^'775wMichaelsen, 1907, Af Michaelsen, 1907, Woodwardiella molaeleonis (as Woodwardui molaeleonis Michaelsen, 1907), Graliophdm candidits (as Plutelhis Jackson, 1931) and W. magna^ (as Woodwardia wiagna Jackson, 1931). Michaelsen also collected the lumbricid Bimastos partuLs''^ {-ds Helodnlxis parvus [Eisen, 1874]). These species are omitted from consideration here because the Darling Range in 1905 and the late 1920s was only sparsely settled, and was not part ol suburban Perth as it is now. The Darling Range native earthworm fauna is more properly considered in the context of the northern jarrah forest. The purpose of this paper is to document the distribution in the Perth metropolitan area of all species of earthworm found there in the late 1970s. Between 1977 and 1980 collections of earthworms were made at 70 sites in and around suburban Perth (Figure 1). The scale of mapping, 1 :400 000, is large enough that any real expansion or contraction of distribution of species in the future will be detectable. A key and glossary of specialist terminolog)^ is also provided so that all species can be readily identified from external features. It is hoped that this key will enable naturalists to be better able to realize the significance of new material as it is collected. All the specimens examined for this paper, with details of their places of collection, have been lodged in the Western Australian Museum. Species Diagnoses and Distributions Thirteen earthworm species were collected (Table 1). Five of these had not been recorded before: Aporrectodea caligiiiosa, two indeterminate Megascolex species, and two indeterminate species belonging to a non-perichaetine genus. These await formal description by a specialist. None of these four indeterminate species can be matched with the descriptions of Michaelsen (1907), so they will probably prove to be new to science. I failed to relocate Diplotrema cornigravei and Graliophilus levis, collected earlier by Michaelsen. According to Jamieson (1971: 502), D. cornigravei is ‘widespread in swamps on the Swan Coastal Plain (pers. obs.)’. The only specimens of these species in the Western Australian Museum are those collected by Michaelsen and none of these is suitable for study now. Both species are therefore omitted from the diagnoses and key. Ail species diagnoses, except where otherwise noted, are based on the material collected, and are composite. The external coloration and maximum length are taken from specimens preserved as follows: held in 35-40% alcohol for one minute: transferred to 4% formaldehyde for 3 hr; and preserved in 75% alcohol ■ Introduced to Western Australia. According to Jamieson (1970: 105), this species is a synonym of W. affinis (Michaelsen, 1907). 13 Earthworms in the Perth Metropolitan Area Figure 2 Known distribution in 1977-79 of Aporrectodea caliginosa (k) and A. trapezoides (•) in Perth metropolitan area. 14 Ian Abbott Family Lumbricidae Aporrectodea caliginosa (Savigny, 1826) Enterion caliginosa?>^^/\gny, 1826: 180. Allolobophora turgida Eisen, 1873: 46. Aporrectodea caliginosa—Orley 1885: 22. no7j Helodrihis caliginosiis—Mich'dehQn 1907:229; — Michaelsen 1911: 142; —Jackson 1931: 126. Diagnosis (external) Length: to 70mm. Colour: pale yellow; pale fawn; pale yellow anterior to segment 15, otherwise light grey; fawn except dorsally on segments 15-29 which are dark grey. Male pores: on segment 15, spreading slightly to neighbouring segments, and lying between b and c setal lines. Clitellum: variants noted are 27-35, 28-34, 28-1/235, 1428-^235, 29-34 (3 specimens), 29-35 (2 specimens), 29-!435- Tubercula pubertatis: Segments 31-33, rarely 31-34, 30-33, or 32-34; always very narrow on the middle segment, and occurring between b and c setal lines. Genital tumescences; between a and b setal lines and usually paired. Variants noted are 9-11; 9-11, 27, 30-34; 27 (RHS), 30, 32-33; 9-11, 27; 27 (LHS), 30, 32-33. Setae: closely paired. Behaviour when handled: sluggish. Gates (1972a) provides a very detailed description of this species. Distribution Rare (Figure 2), recorded from only five sites— all gardens (Table 1). From the distribution map it is evident that this species is found in the higher rainfall zone. The range outside the metropolitan area is shown in Figure 9. Distribution outside Australia: cosmopolitan (Gates 1972a). Aporrectodea trapezoides (Duges, 1828) Lumbricus trapezoides Duges, 1828: 289. Aporrectodea trapezoides— Orley 1885; 22. Helodrilus caliginosm—M\ch^chen 1907: 229; —Michaelsen 1911: 142; —Jackson 1931: 126. All of Michaelsen s specimens held in the W. A. Museum and labelled Helodriliis caliginosus belong to A. trapezoides and not A. caliginosa (pers. obs.). Diagnosis (external) Length: to 115 mm. Colour: pink to segment 15, thereafter grey; dark grey to segment 13, thereafter browm; all grey but darker dorsally; greyish-fawn; dark grey but nearly purple dorsally; fawn; pale yellow. Male pores: paired on segment 15 with papillae extending on to segments 14 and 16, and lying between setal lines b and c. (In immature specimens without clitellum and tubercula pubertatis, the male pores do not extend bevond segment 15.) Clitellum: white, fawn or light grey; usually on segments 27-34, occasionally on 27-35, rarely *426-34, 26-34, *426- 14^5, 26-35, 27-33, 28-34 or 28-35. Tubercula pubertatis: usually on segments 31-33, 30-33 or ‘/230-33; rarely on 30-32, 30-!434, !/230-i434, 31-34 or 31-33; occurring between setal lines b and c. The tubercula may be either translucent or the same colour (whitish) as the clitellum. Genital tumescences: paired, lying between a and b setal lines in any of the following arrange- 15 Earthworms in the Perth Metropolitan Area Figure 3 Known distribution in 1977-79 pAsenia fetida in Perth metropolitan area. 16 Ian Abbott merits: 9-11, 32-34; 9-11; 9-11, 27-29, 31-34; 9-11, 27-34; 9-11, 27-30, 32-33; 9-11, 27-29 29/30, 31/32; 9-11, 30-35; 9-11,30-31, 32-34; 9-11, 30-34; 9-11, 26-35; 9-11, 26-28, 31-33; 9-11, 30, 32-34; 30, 32-33. Setae; closely paired with aa > he. Behaviour when handled: sluggish. Gates (1972a) provides a detailed description. Distribution Recorded at 30 sites (Figure 2), associated with gardens and other man-made areas (Table 1). This species is found more frequently in the Darling Range than Eiseniafetida. In 1905, A. trapezoides was recorded at four widely distributed sites in the metropolitan area. Range outside metropolitan area; widespread (Figure 10), extending farther into semi-arid parts of south-western Australian than A. caliginosa. Range outside Australia: cosmopolitan (Gates 1972a). Table 1 Frequency of occurrence of earthworm species in metropolitan Perth region 1977- 1980. Family and species Uncleared sites ' Frequency Man-affected sites ^ Total sites Lumbricidae "^Aporrectodea caliginosa 0 5 5 *A. trapezoides 0 30 30 "^Eiseniafetida 0 38 38 "^Eiseniella tetraedra 0 1 1 Megascolecidae ^Amynthas cortiens 0 8 8 Megascolex imparicystis 0 3 3 M. longicystis 0 1 1 M. sp. indet. A 0 16 16 M. sp. indet. B 2 0 2 "^Mkroscolex dubiiLS^ 0 46 46 Woodwardiella libferti 4 0 4 sp. indet. C 0 1 1 sp. indet. D 0 1 1 * These species are introduced to Western Australia. ' Areas that still carry native vegetation with substantial undergrowth of native species, though some weeds may be present. ^ Includes gardens, plant nurseries, man-made parks, compost areas, stables, lawns, pine plan- tations, orchards, weedy roadside verges, areas with native trees but with completely weedy undergrowth (native understorey destroyed). ^ Records from adjacent islands are omitted. 17 Earthworms in the Perth Metropolitan Area Figure 4 Known distribution in 1977-79 o{ Eiseniella tetraedra (k), Megascolex longicystis (y), Megascolex sp. indet B. (■), Woodwardiella lihferti (•). sp. indet C. (*), and sp. indet D. (®) in Perth metropolitan area. 18 Ian Abbott Eisenia fetida (Savigny, 1826) Enterioii fetidum Savigny 1826: 182. Eisenia foetida- M'Am 1877: 45; -Michaelsen 1907: 228; -Michaelsen 1911: 142; -Jackson 1981: 126. Diagnosis (external) Common name: Tigerworm, Brandling, Red Wriggler. Length: to 105 mm. Colour: dorsally purplish-red, ventrally fawn or pale yellow and posteriorly with yellow between the segments. Sometimes purplish-red all over except for the ventral parts of the first —20 segments, which are yellow. Male pores; paired, on segment 15 between b and c setal lines. Papillae do not extend onto segments 14 and 16. Clitellum; fawn or off-white, covering on most specimens segments 26-32, occasionally 1/224-33, 25-30, 25-32, 26-31, 26- 1432, 26-1433, 26-33, or 27-32. Tubercula pubertatis: between b and c setal lines on segments 28-30 or 28-1431 on majority of specimens examined, occasionally on 25-30, 27- 1430, 28-31, 1428-32 or 29-31. Genital tumescences: usually paired, but variably developed: 9-13, 19 (RHS), 23, 26-32; 9, 11 single, 26-32; 8-11, 14, 21, 23, 26-33; 9-12, 14, 16, 26-32: 28-31; 9, 11, 12 , 23-32; 8-11, 23-32; 24-32; 9, 12, 26-32; 9-12; 9-12, 26-32; 26-31; 22 (RHS), 23 (LHS), 27-33; 8-12, 26-32; 22-33. On segments before 15 they usually occur between the c and d setal lines whereas on segments after 15 they occur between the a and b setal lines. Setae: closely paired, aa = be. Behaviour when handled: wriggles and squirms excitedly. A yellow foul smelling fluid is also ejected. Distribution Collected at 38 localities (Figure 3), in contrast to Jackson ( 1931 ) who noted it at only one locality. Recorded only from man-made habitats such as gardens, stables and chicken runs (Table 1). Range outside metropolitan area: Very localized, see Figure 9. Range outside Australia: cosmopolitan (Michaelsen 1907). Eiseniella tetraedra (Savigny, 1826) Enterion tetraedrum Savigny, 1826: 184. Eiseniella tetraedra Michaelsen 1900: 473; —Michaelsen 1907: 228; —Michaelsen 1911: 142; —Jackson 1931: 123. Eiserdedla intennedius ]'Ac\:.so\\, 1931: 123; —Michaelsen 1935:40. Diagnosis (external) Length; to 45 mm. Colour: reddish-purple except fawn ventrally. Male pore: on segment 13, between a and c setal lines. Clitellum: on segments 22-27. Tubercula pubertatis: on segments 23-26 or 23-i426 between h and c setal lines but closer to c. Genital tumescences: obvious only on segments 23-25. Setae: closely paired, with aa = be. Note: Segments anterior to 22 are of circular cross-section in contrast to those after the clitellum which are square in transverse section. Distribution This species was collected from only one locality, much disturbed, in South Perth close to the left bank of the Swan River (Figure 4). In the 1920s, it was collected only near 19 Earthworms in the Perth Metropolitan Area Figure 5 Known distribution in 1977-79 of Amynthas corticus in Perth metropolitan area. 20 Ian Abbott Lake Monger (Jackson 1931). Range outside metropolitan area: the only record is near Albany (Michaelsen 1907). Range outside Australia: cosmopolitan (Michaelsen 1907). This species would be expected to occur only in or close to water. Family Megascolecidae Amynthas corticus (Kinberg, 1867) Perichaeta corticis Kinberg, 1867: 102. Amynthas corticus Sims & Easton 1972: 235. Pheretima heterochaeta Michaelsen 1903: 96; —Michaelsen 1907: 226; —Michaelsen 1911: 142; Jackson 1931: 119. Diagnosis (external) Length: to 110 mm. Colour; pale brown-yellow dorsally and fawn ventrally; fawn; pale yellow-fawn; pale brown. Male pore: paired, eyelike, on segment 18, rarely segment 17. The slit may be centred on setal line f, g, h, orj, and usually extends to one or two setal lines on either side. Clitellum: chocolate coloured or light brown, usually covering segments 14-16, occasionally 14-1416 or '/ 2 I 3 -I 6 . Female pore; visible as a central white circle on ventral part of segment 14. Spermathecal pores: usually paired. Variants recorded are; single on 8 (d line); 5/6, 6/7, 7/8, 8/9 as small white bulbs on h line; 5/6, 6/7, 7/8; 5-8; 7-9 on d line; 7-8 on d line; 8-9 on d line; 8, single (RHS) on d line; 7 (paired) and 8 (single) on c line; 8 on c line. On the basis of such variation in position of these pores, E.G. Easton (pers. comm.) suggests that two species may be involved. However, there is still considerable variation in this feature in earthworms collected at the same locality. Setae: perichaetine with about 40-50 per segment, pointing forward, with the setal rings unbroken dorsally and ventrally. aa slightly > ab. Dorsal blood vessel not obvious. Behaviour when handled: snake-like in its movements. Distribution Rare (Figure 5), recorded in gardens only. In the 1920s, this species was first recorded in what are now Supreme Court Gardens in Perth (Jackson 1931). Range outside metropolitan area: Boyanup (Michaelsen 1907) and banks of Gooralong Brook, Jarrahdale. Range outside Australia: cosmopolitan (Michaelsen, 1907). Megascolex imparicystis Michaelsen, 1907 Megascolex imparicystis Michaelsen, 1907: 209; —Michaelsen 1911: 141: —Jackson 1931: 109; -Michaelsen 1935: 39. Diagnosis (external) Length: mean 215 mm. Range 150-410 mm based on 27 specimens held in Western Australian Museum (mode of preservation not known). Colour: fawn with grey dorsum; very pale yellow with purplish dorsum; light grey but darker dorsally; pale yellow-fawn all over. Male pore: unpaired on segment 18 between a setal lines. Clitellum: rarely distinct, 14-19, with sides not meeting ventrally. Accessory glands; arrangement 21 Earthworms in the Perth Metropolitan Area Figure 6 Known distribution of coded by decades: 1 = 1920s, 2 = 193()s, 3 = 195()s.4 = 1960s, 5 = 1970s. (Most records from specimens in Western Australian Museum.) 22 Ian Abbott Figure 7 Known distribution in 1977-79 Megascolex sp. indet A in Perth metropolitan area. 23 Earthworms in the Perth Metropolitan Area Figure 8 Known distribution in 1977-79 of Microscolex dubius in Perth metropolitan area. 24 Ian Abbott quite variable, but always occurring singly and ventrally in furrows between segments- 17-18, 18/19 (twice); 17/18, 18/; 15/16, 16/17-23/24; 14/15-23/24; 15/16-22/23; 4/5-8/9, 17/18-25/26; 14/15-19/20; 16/17-19/20; 16/17-18/19; 6/7, 7/8, 15/16-21/22; 6/7-8/9, 15/16-23/24; 4/5-7/8, 15/16-18/; 16/17-20/21; 15/16-22/23; /16, 16/17, 18/19-23/24; 4/5-7/8, 16/17-25/26. Those before segment 10 lie between a setal lines and are not always present. Those accessory glands between segments 15 and 24 are large, but vary in size, lying between c setal lines to g setal lines. Setae: perichaetine, with 30-50 per segment, aa ^ 1.5 ab. Behaviour when handled: sluggish. Distribution I did not hnd this species in the metropolitan area in 1977-1980, although specimens from three metropolitan localities were brought into the Western Australian Museum. The date of collection (if available) has been coded by decades in Figure 6. This species has been collected as new suburbs were created from dune scrub or woodland and also localized specimens held in the Western Australian Museum, and the author’s collection, now in the Western Australian Museum). 25 Earthworms in the Perth Metropolitan Area Figure 10 Known distribution oi' Aporrectodea trapezoides in temperate Western Australia (based on all records available). from reserves and agricultural land remaining. Both Michaelsen (1935) and Jamieson (1971) characterized this indigenous species as peregrine, on the basis that the species has been found in gardens. Megascolex imparicystis has not been collected south of the Swan River or in the Darling Range. Range outside metropolitan area: north of the metropolitan area this species still survives on the coastal plain (Figure 9). I have collected specimens from pasture near Lancelin and Dandarragan. During cultivation large numbers become caught in the tynes near Lancelin (personal observation) and near L^pper Swan (R. Barrett-Lennard, pers. comm.). Megascolex longicystis Nicholls and Jackson, 1926 Megascolex longicystis N'lchoWs 2 Lnd Jackson, 1926: 142; — Jackson 1931: 116. Diagnosis (external) The following description is based on the one specimen collected. Length: 55 mm (to 80 mm, according to Nicholls and Jackson). Colour; purplish-brown dorsally, fawn 26 Ian Abbott ventrally, and anterior and posterior extremities tipped white. Male pore: paired on segment 18, centred on b setal line and extending from a to c setal lines. Clitellum, not distinct, H-VilV. Accessory glands, not visible but according to Nicholls and Jackson there are two pairs on anterior margins of 8 and 9 in r setal lines. Setae: perichaetine, with about 20-25 per segment. Setal ring is broken ventrally and dorsally, with (m ~ 1.5 ab. Distribution Found only at one locality (Figure 4), under eucalypts with a completely weedy undergrowth. Range outside metropolitan area: the two 1920s records, at Wungong and Armadale (Jackson 1931) are still the only non-metropolitan records. Megascolex sp. indet. A Diagnosis (external) Length: to 105 mm. Colour: first 3-12 segments reddish or pinkish-fawn; rest of body grey or pale yellow. Male pore: paired on segment 18, rarely on 17; eyelike, centred on b setal line and extending from a-c lines or centred on d line and extending from c~e lines. Female pore: median and unpaired on ventral part of segment 14. Clitellum: chocolate coloured, on segments 14-16, occasionally 14-15, rarely 13-16. Accessory glands: nil. Setae: Perichaetine, with 20-30 per segment. aa> ab> be > cd ....xy < zz. Dorsal blood vessel is prominent. Behaviour when handled: wriggles excitedly and will often break in two. Distribution See Figure 7. Recorded only in gardens and a plant nursery (Table 1). There are no records from outside the metropolitan area. Note: The material has been registered in the Western Australian Museum collection as WAM 149-81 to 164-81 inclusive. Megascolex sp. indet. B Diagnosis (external) Length: to 55 mm. Colour: dorsally dark red to purple, ventrally fawn; Male pore: paired on prominent papillae on segment 18, between a and b setal lines. Clitellum: covering segments 13 or 14-17. Genital markings: paired, lying in the furrow of 17/18 between b and c setal lines. In one specimen there is a single genital marking in the furrow of 19/20 between the LHS a and b setal lines. Setae: perichaetine, about 20 per segment, with setal ring broken dorsally and ventrally. aa= 1.5 ab. Dorsal blood vessel is visible through preclitellar segments only. Distribution Found only at two localities, one in black sand under Melaleuca at the edge of a fresh- water lake, the other in sand under Banksia woodland (Figure 4). This species is not known to occur outside the metropolitan area. Note: This material has been registered in the Western Australian Museum collection as WAM 165-81 and 166-81. 27 Earthworms in the Perth Metropolitan Area Microscolex dubius (Fletcher, 1887) Eudrilus duhius Fletcher, 1887: 378. Microscolex dubhis-RosR 1890: 511; -Michaelsen 1907: 146; -Michaelsen 1911: 140; - Tackson 1931:85. Microscolex phosphoretLs ] 2 Lm\esox\. 1974: 201. Diagnosis (external) Length: to 110 mm. Colour: pinkish or pinkish-white to about segment 12, otherwise pale grey, pale yellow; fawn-yellow to about segment 12, otherwise grey-yellow. Male pore: paired on a setal lines of segment 17 or 18 and surrounded by white papillae. Female pore: median ventral on segment 13 or 14, in clitellate specimens only though not alw'ays visible. Clitellum: yellow-fawn or light brown, usually on segments 13-16, or 14-17, occasionally on 13-17, 13-i/2l7, !/2l3-!/2l7, VilS-lG, 14-16, 14-/217, or 14-18. Setae: widely spaced; aa = 2ab except on segments 17-20 where b is close to a. The a and b setae on 17 are about twice the length of those on 18. ah < cd < be. Behaviour when handled: sluggish. Jamieson (1974b) provides a very detailed description of both external and internal features. Figure 11 Known distribution of Microscolex didhixs in temperate Western Australia (based on all records available). 28 Ian Abbott Distribution See Figure 8. Recorded from man-disturbed sites (Table 1). This species was collected from Green Island, a small stack adjacent to Rottnest Island. It is possible that nesting Silver Gulls Larns novaehollandiae have introduced it from nearby Rottnest Island, which is much disturbed by man. Microcolex dubius has a wider distribution on the coastal plain than Aporrectodea trapezoides. In 1905 this species was recorded at Rottnest Island and six metropolitan sites, and in the 1920s at Rottnest Island, Wungong and Cottesloe. Range outside metropolitan area: Extensive (Figure 11). This species is more widespread than stated by Jamieson (1971: 503). Range outside Australia: cosmopolitan (Michaelsen 1907). Woodwardiella libferti (Michaelsen, 1907) Woodwardia libfertiMichdiChen, 1907: 193 Woodwardia lipferti—M\c\\ 2 iGhcn 1911: 141. Woodwardiella libferti—Siephenson 1925: 888; —Jamieson 1970: 105. Woodwardiella lipfer-ti— Jackson 1931: 103. Diagnosis (external) Length: to 50 mm. Colour: Segments 10-50 are yellow, otherwise pale grey; Male pore: paired on papillae on segment 18, between a and b setal lines; Female pore: small, paired on segment 14, between a setal lines. Clitellum: not obvious in specimens examined. Genital markings: a pad lies in the ventral furrow of segments 11 and 12, and extends to between the a and b setal lines. Setae: aa = 2ab. Behaviour when handled: sluggish. Distribution Recorded at four places (Figure 4), all in sand under Banksia woodland. This is the only species collected in the Banksia woodland of Kings Park and is not known to occur outside the metropolitan area. Remarks: Michaelsen 's text figures for W. libferti and W. molaeleonis (given on pp. 194 and 195 respectively) have inadvertently been reversed as they do not match his descriptions. This apparently has led to confusion by Jamieson (1971: 487). Sp. indet. C Diagnosis (external) Length: to 55 mm. Colour: Segments 1-15 pale yellow, remainder grey: Male pores: paired on 18, between a setal lines; Clitellum: covering 14-17 or not obvious; Genital markings: single, in furrow 19/20 between b setal lines. Setae: four pairs, aa = 2-3 ab. Distribution Found in only one site in the metropolitan area (Figure 4). This species has been found at only one location outside the metropolitan area, under undisturbed woodland at Wilbinga Grove north of Perth. Note: This material has been registered in the Western Australian Museum collection WAM 167-81 and 168-81. 29 Earthworms in the Perth Metropolitan Area Sp. indet. D Diagnosis (external) Length: to 90 mm. Colour: Segments 1-16 fawn, remainder grey; Male pores: paired on 18 on a setal lines: Clitellum: not obvious. Genital markings: paired on a setal lines of segments 8 and 9, and paired on large papillae in furrows of 15/16, 16/17, 19/20 on a setal lines. (Papillae lie between h-c setal lines). The only other variants noted were single (LHS) 14/15 {a line), and paired 15/16, 16/17, 19/20; and 16/17, 19/20, 20/21. Distribution Found only in Gnangara Pine Plantation in bank of a creek (Figure 4). Note: This material has been registered in the Western Australian Museum collection as WAM 169-81. Discussion Causes of Differences in Distribution of Native and Introduced Earthworms A clear-cut difference was not found between the distribution of native and introduced earthworm species in the Perth metropolitan area. All introduced species were found only in sites disturbed by settlement, but only a minority of native species were found only in uncleared sites. The hrst hnding was expected, but the second was a surprise. The literature on Southern Hemisphere earthworm faunas has often noted that native and introduced species have exclusive distributions, and there has been controversy about the relative roles of interspecihc competition between native and introduced species and the replacement of native habitats by ones suitable for introduced species (Stephenson 1930, Barley 1959, Satchell 1967, Lee 1961, Ljungstrdm 1972). What is particularly curious is that of the seven native species collected only two (Meffoscolex sp. indet. B and Woodwardiella libferti) were found only in uncleared sites. It is not known whether the other five species are native to the Perth region and have adapted to land-clearing, or if they have been introduced from another part of south- western Australia. In either case, the possibility of interspecihc competition between these hve species and the introduced species cannot be ruled out. I constructed from Michaelsen (1907) a table similar to my Table 1. Unfortunately the number of cases tor 1905 is so small that no useful conclusions can be drawn from a comparison with the 1977-9 data. Introduced Earthworm Fauna Round Perth Compared with Other Australian Regions Comparisons of the earthworm fauna of metropolitan Perth with that of other capital cities in Australia is difficult for several reasons. There has been inadequate collecting, particularly of introduced species, and many of the older specimens held in museums are in a poor state of preservation, often being impossible to identify. I have examined the earthworm collection in the South Australian Museum, the catalogued part of the collection in the National Museum of Victoria, and the card indexes of specimens held in the Australian Museum. 30 Ian Abbott Gates (1972(2, 197S) provides authoritative information on the occurrence of certain lumbricid species in Australia. The following lumbricid ' species (with their known occurrence elsewhere in Australia) have not been recorded from the Perth metropolitan area: Aporrectodea longa (Tasmania), A. tnberadata (New South W^\es),Bimastos parous (New South Wales), Devdrodriliis nibidus (New South Wales, Victoria), Lumbricus nibeUiLs (New South Wales, Victoria), Octolasion tyrtaemn (New South Wales). Both Bimmtos pannis and Dendrodrilus nibidus in the above list were recorded in Western Australia in 1905 but outside the metropolitan area (Michaelsen 1907). In addition, Eisenia rosea and Octolasion cyaneum have recently been recorded in south-western Australia (Abbott 1981), and' would seem to be good candidates for colonizing and establishing in the Perth metropolitan area. Why some lumbricid species and not others have flourished in the Perth metropolitan area is a subject meriting more detailed analysis. At present it is not possible to state whether either opportunity of introduction to the area or difficulties with establishment because of unsuitable climate, soil type etc. is involved. Parthenogenesis is usually stated to be an advantage in a newly colonized environment if mates are scarce. Of the six introduced species present in metropolitan Perth, Aporrectodea trapezoides, Eiseniella tetraedra, Amynthas corticus and Microscolex dubius are parthenogenetic, Aporrectodea caliginosa and Eisenia fetida are amphimictic (Gates 19726, Martin 1977, Reynolds et ai 1974). There thus seems to be no close association between breeding system and distribution of species in the Perth metropolitan area. Similarly for the eight lumbricid species recorded elsewhere in Australia but not in the Perth region, hve species (Bimastos parvuSy Dendrodrilus nibidus, Eisenia rosea, Octolasion cyaneum, O. tyrtaeum) are partheno- genetic and the remaining three (Aporrectodea longa, A. tuberculata, Lumbricus rubelhis) are amphimictic. These comparisons suggest that breeding system is not a particularly important characteristic for peregrine earthworm species. Jaenike and Selander (1979) have proposed that ‘parthenogenetic earthworms commonly occur in ephemeral or unstable habitats, in which r-selection may be expected, whereas sexual species tend to inhabit more stable environmental situations, where K-selection may be more important’. When this reasoning is applied to Western Australia, we should expect all earthworm species to be parthenogenetic. It seems more likely that the frequency of human-assisted movements and the ecology of the soil of suburban areas will be of more relevance to understanding the distribution of peregrine species. Acknowledgements I thank many friends for collecting specimens: J. Conacher kindly provided a dispro- portionate share of these. L. Marsh gave me access to the earthworm collection in the Western Australian Museum, and P. Hutchings (Australian Museum), C. Lu (National Museum of \ ictoria) and W. Zcidler (South Australian Museum) loaned specimens or provided other help. I also thank J.D. Plisko-Winkworth for the method of preserving specimens. 4 he suggestions of two referees substantially improved the manuscript. ^All introduced to Australia. 31 Earthworms in the Perth Metropolitan Area Key to Identification of Species {Diplotrema cornigravei and Graliophilus levis, collected by Michaelsen [1907], were not sighted by me and are omitted.) 1 Male pores on segment 13 or 15 2 Male pores on segment 17 or 18 3 2 Clitellum terminating on or before segment 33 4 Clitellum terminating on segment 34 or 35 5 3(1) Setae, 8 on each segment 6 Setae, more than 8 on each segment 7 4(2) Clitellum terminating on or before segment 27; tubercula pubertatis terminating on segment 26 Eiseniella tetraedra Clitellum terminating on or before segment 33; tubercula pubertatis terminating on segment 30, 31 or 32 Eisenia fetida 5(2) Tubercula pubertatis rod-like Aporrectodea trapezoides Tubercula pubertatis distinctly bilobed, the lobes joining on segment 32 Aporrectodea caligiriosa 6(3) Accessory genital markings present 8 Accessory genital markings absent Microscolex dubius 7(3) Male pore single Megascolex imparicysth Male pores paired 9 8(6) Genital markings single 10 Genital markings paired Sp. indet. D 9(7) Setae^40 per segment; setal ring unbroken dorsally and ventrally Amynthas corticus Setae<30 per segment; setal ring broken dorsally and ventrally 1 1 10(8) Genital marking in furrow 11/12 Woodwardiella Ubferti Genital marking in furrow 19/20 Sp. indet. C 1 1(9) Dorsal surface purplish or purplish-brown, ventral surface fawn or grey 12 Dorsal and ventral surface fawn or grey Megascolex sp. indet. A 12(1 1) Both ends of body tipped white (genital markings never paired in furrow 17/18) Megascolex longicystis Ends of body not white (genital markings often paired in furrow 17/18) Megascolex sp. indet. B. 32 Ian Abbott References Abbott, I. (1981). Two species of lumbricid earthworm newly recorded from Western Australia. Rec. W. Aust. Mus. 9: 273-277. Barley, K.P. (1959). The influence of earthworms on soil fertility. I. Earthworm populations found in agricultural land near Adelaide. Aust. J. Agr. Res. 10: 171-178. Duges, A. (1828). Recherches sur la circulation, la respiration et la reproduction des Annelides Abranches. Annls Sc. nat. 15: 284-337. Eisen, G. (1873). Om Skandinaviens Lumbricider. OfversK. VetenskAcad. Fork. Stockholm. 30: 43-56. Fletcher, J.J. (1887). Notes on Australian earthworms. Part III. Proc. Linn. Soc. N.S.W. ser. 2, 2: 376-402. Gates, G.E. ( 1972a). Contributions to North American earthworms (Annelida: Oligochaeta) No. 3. Toward a revision of the earthworm family Lumbricidae IV. The trapezoides species group. Bull. Tall Timbers Res. Stn. No. 12: 1-146. Gates, G.E. (1972/d- Contributions to North American earthworms (Annelida). No. 5. On variation in another anthropochorous species of the oriental earthworm genus Pheretima Kinberg 1866 (Megascolecidae). Bull. Tall Timbers Res. Stn. No. 13; 18-44. Gates, G.E. (1973). Contributions to North American earthworms (Annelida) No. 8. The earth- worm genus Octolasion in America. Bull. Tall Timbers Res. Stn. No. 14: 29-50. Jackson, A. (1931). The Oligochaeta of South-Western Australia./. R. Soc. W. Aust. 17: 71-136. Jaenike, J. and Selander R.K. (1979). Evolution and ecology of parthenogenesis in earthworms. Amer. Zool. 19: 729-737. Jamieson, B.G.M. (1970). A revision of the Australian earthworm genus Woodwardiella with descriptions of two new genera (Megascolecidae: Oligochaeta)./. Zool, Lond. 162: 99-144. Jamieson, B.G.M. (1971). Earthworms (Megascolecidae: Oligochaeta) from Western Australia and their zoogeography./. Zool. Lond. 165: 471-504. Jamieson, B.G.M. (1974a). The zoogeography and evolution of Tasmanian Oligochaeta. In Bio- geography and Ecology in Tasmania. (Ed., W.D. Williams): 195-228. ^unk: The Hague.) Jamieson, B.G.M. (1974/d. Earthworms (Oligochaeta: Megascolecidae) from South Australia. Trans. R. Soc. S. Aust. 98: 79-1 12. Jamieson, B.G.M. (1981). Historical biogeography of Australian Oligochaeta. In Ecological Biogeography of Australia, (Ed., A. Keast): 887-921. (Junk: The Hague.) Kinberg, J.G.H. Annulata nova. Ofvers. K. VetenskAcad. Fork. Stockholm. 23: 97-103, 356-357. Lee, K.E. (1961). Interactions between native and introduced earthworms. Proc. Ecol. Soc. New Zealand 8: 60-62. Ljungstrom, P.-O. (1972). Introduced earthworms of South Africa. On their taxonomy, distribution, history of introduction and on the extermination of endemic earthworms. Zool. Jb, Syst. Bd. 99: 1-81. Malm, A.W. (1877). Om daggmasker, Lumbricina. Ofvers. Sdllsk. Hort. Vann. Goteborgs F'drh. 1: 34-47. Martin, N.A. (1977). Guide to the lumbricid earthworms of New Zealand pastures. N. Z. J. Exp. Agric. 5: 301-309. Michaelsen, W. (1900). Oligochaeta. Das Tierreich 10: 1-575. (R. Friedlander 8c Sohn: Berlin.) Michaelsen, W. (VHYd). Die geographische VerbreitungderOligochaten. (R. Friedlander & Sohn: Berlin.) Michaelsen, W. (1907). Oligochaeta. In Die Fauna Sudwest-Australiens, 1: 117-232. Michaelsen, W. (1911). Second abstract of the reports of the (ierman expedition of 1905 to South- Western Australia. Part II. Oligochaeta. /. Nat. Hist. Sci. Soc. VV' Aust. 3: 138-142. Michaelsen, W. (1935). Earthworms from South-Western Australia././?. Soc. West. Aust. 21: 39-43. Morison, M.P. (1979). Settlement and development. The historical context. In Western Towns and Buildings (Eds., M.P. Morison and J. White): 1-73. (Western Australian Government, Perth.) Nicholls, G.E. and Jackson, A. A. (1926). Some new species of Megascolex from South-Western Australia././?. Soc. W. Aust. 12: 141-147. 33 Earthworms in the Perth Metropolitan Area Orley, L. (1885). A Palaearktikus ovben elo terrikolaknak revizioja es elterjedese. Magy. ticdorn. Akad. Ert. a tenrihzettud. Kor'ebol. 15(18): 1-34. Reynolds, J.W., Clebsch. E.E.C. and Reynolds, W.M. (1974). Contributions to North American earthworms (Oligochaeta). No. 12. The earthworms of' Tennessee (Oligochaeta), I. Lumbricidae. Bull. Tall Timbers Res. Stn. No. 17: 1-107. Rosa. D. (1890). Terricoli Argentinl raccolti dal Dott. Carlo Spegazzini. Annali Mus. civ. Stor. rmt. Giacomo Doria 29: 509-521. Satchell, J.E. (1967). Lumbricidae. In Soil Biology.{Eds., A. Burges and E Raw): 259-322. (Academic Press: London.) Savigny. [.-C. (1826). Analyse d’un Memoire sur les Lombrics par Cuvier. Mem. Acad. Sci. Inst. France. 5: 176-184. Sims, R.W. and Easton, E.G. (1972). A numerical revision of the earthworm genus Pheretima auct. (Megascolecidae: Oligochaeta) with the recognition of new genera and an appendix on the earthworms collected by the Royal Society North Borneo expedition. Biol. J. Linn. Soc. 4: 169-268. Stephenson. J. (1925). Oligochaeta from various regions, including those collected by the Mount Everest Expedition, 1924. Proc. Zool. Soc. Lond. 1925: 879-907. Stephenson, |. (1930). The Oligochaeta. (Clarendon Press: Oxford.) Glossary The following terms used in this paper are standard ones used in the literature of earthworms. Stephenson (1930) and most elementary books on invertebrates provide explanatory diagrams. Accessory glands Clitellum Female pore Genital markings Genital tumescence Male pore Papillae Peregrine Perichaetine Segments Setal lines Spermathecal pores Tubercula pubertatis Glands (sometimes called prostates) associated with the male reproductive organ. A glandular thickening of the body wall associated with cocoon production . Externa! opening of the female reproductive organ; the pore through which eggs leave the body. Pads or ridges often associated with the male reproductive organs. Papillae or ridges bearing modified setae, probably to facilitate mating. External opening of the male reproductive organ; the pore through which sperm leave the body. White lips surrounding the male pores. Earthworm species easily transported by Man and which live in areas disturbed by Man e.g. gardens and pasture. More than 8 setae per segment. Figures quoted of number of setae per segment refer to segments 20-25. These are numbered from the anterior backwards. The prostomium (a lobe over the mouth) is not numbered. The next segment = 1. Setae are arranged in regular longitudinal lines around the body. The most ventral setae are each labelled a, the next b, etc. The most dorsal setae are each labelled z, the nexty, etc. Intersetal distances are then quoted as (for example) aa — ah, or aa = 1 .5 ah. External openings of the spermathecae (sperm-storage organs); the pores through which sperm leave the body after eggs are produced. In this paper used to designate genital markings found as ventrolatei al swellings on the clitellum of lumbricid earthworms. Received 24 June 1981 Accepted 12 March 1982 Published 30 June 1982 34 Rec. West Aust. Mus. 1982 , 10 ( 1 ): 35-45 Variation in Pseudechis australis (Serpentes: Elapidae) in Western Australia and Description of a New Species of Pseudechis L.A. Smith* Abstract Pseudechis australis (Gray) and P. butleri sp. nov. from Western Australia are described and their distributions are mapped. The substantial variation in P. australis is analysed and the relationships of P. butleri are discussed. Introduction Some of Boulenger’s generic concepts for Australian elapid snakes have stood the test of time better than others. There is general agreement amongst herpetologists today that his concept of Denisonia (1896) was too broad, although there is still disagreement about the generic position of species removed from Denisonia (sensu Boulenger) (Storr 1981). On the other hand Boulenger’s concept of Pseudechis (1896) has changed little over the last 85 years. In fact much of the literature pertaining to Pseudechis stems from authors with short series of specimens at their disposal describing variants of species, particularly the widely distributed P. australis. Boulenger (1896: 328) recognized eight species oi Pseudechis: australis (Gray, 1842), cupreus sp. nov., darwiniensis Macleay, 1878, fei'ox (Macleay, 1881) microlepidotus (McCoy, 1879), papuanus Peters and Doria, 1878, porphyriacus (Shaw, 1794) and scutellatus Veters. 1867. Pseudechis colletti Boulenger, 1902, R guttatus De Vis, 1905, R denisonioides Werner, 1909, P. mortonensis De Vis, 1911, R platycephalus Thompson, 1933 and P. wilesmithii De Vis, 1911 have subsequently been described. Kinghorn (1923) removed R scutellatus from Pseudechis and erected Oxyuranus for it. Thompson (1930) synonymized cupreus and darwiniensis with australis and later described platycephalus. Mack and Gunn (1953) synonymized mortonensis with guttatus and R. wilesmithii De Vis with O. scutellatus, the latter being foreshadowed by Longman (1913). McKay (1955) transferred P. platycephalus and P. denisonioides to the synonomy of australis, the latter first being suggested by Glauert (see Loveridge 1934: 282). Kinghorn (1955) merged ferox with microlepidotus and erected Parademansia for it. Until recently O. scutellatus and P. microlepidota were treated conspecifically, the latter being considered an inland form of the other. Covacevich and Wombey (1976) have shown them to be distinct species. Covacevich et al. (1981) treat the two species as congeners. * Department of Ornithology and Herpetology, Western Australian Museum, Francis Street, Perth, Western Australia 6000. 35 Variation in Pseudechis australis Figure 1 Top: A Pseudechis butleri from near Sandstone. Photographed by S. Wilson. Bottom: A Pseudechis australis from Nita Downs. Photographed by R.E. Johnstone. 36 L.A. Smith The unique specimen oi Denisoyiia brunnea Mitchell (1951) is considered here to be a P. mistralis possessing two traits unusual for that species. I he lower primary temporal is contacting the lower postocular and all subcaudals are divided. Data for eastern Australian species oi' Pseudechis are taken from McKay (1955). The 213 specimens of T. australis and 21 specimens of P. hutleri examined are lodged in the Western Australian Museum herpetological collections (R Series). Scale rows at neck and tail were counted at the first and last ventral respectively. Bilateral characters such as temporals were counted on both sides of the head. Systematics Pseudechis australis (Gray, 1842) Naja australis Gray, 1842. The zoological miscellany: 55. Type locality NF, Australia. Pseudechis darwiniensis Macleay, 1878, Proc. Linn. Soc. N.S.W. 2: 221. Type locality Port Darwin, Northern Territory. Pseudechis cnpreus Boulenger, 1896, Catalogue of the snakes of the British Museum (Natural History) 3: 329. Type locality Murray River. Pseudechis denisonioides Werner, 1909, Die Fauna Sudwest-Australiens 2 (16): 258. Type locality Eradu, Western Australia. Pseudechis platycephalus Thompson, 1933. Proc. zool. Soc. Loud. 1933: 859. Type locality East Alligator River, Northern Territory. Deuisonia brunnea Mitchell, 1951, Rec. S. Aust. Mus. 9: 551. Type locality Mt Wedge, Eyre Peninsula, South Australia. Diagnosis Distinguished from P btitleri by its colour, particularly the ventrals which are cream with a reddish-brown base (yellow with a black base in R butleri). Pseudechis australis from the same latitudes as P. butleri usually have fewer ventrals (189-207 v. 204-216). Lack of crimson or pink pigment on ventrals distinguishes P. australis from P. porphyriacus (the only other Pseudechis species with 17 midbody scale rows). Description A very large snake (up to 201 cm total length). Tail 13.5-22.2% of SVL (N 47, mean 17.8). Head slender and neck indistinct in small specimens; head broad and neck moderately distinct in large specimens. Canthus rostralis prominent. Rostral 1.1-2. 1 times as wide as high (N 147, mean 1.5). Frontal L1-L9 times as long as wide (N 160, mean 1.5). Nasal completely divided; in contact with the preocular (97% of specimens). One preocular, two postoculars. Primary temporals 2, the lower wedged deeply between the fifth and sixth labials and separated from the lower postocular (66% of specimens). Secondary temporals 2. Upper labials 6, third and fourth entering orbit. Lower labials 6, second smallest, fourth largest. Two pairs of chin shields, anterior pair always in contact, postocular separated (97% of specimens). Ventrals 185-220, lowest in south, highest in north (N 170, mean 201.6). Anal rarely single. Subcaudals 50-78 (N 158, mean 58.6), percentage of undivided .subcaudals 35.7-100 (N 158, mean 69.0). Sum of ventrals plus subcaudals 236-295 (N 144, mean 258.4). Midbody scale rows 17; scale rows at neck 17-24 (mostly 19); scale rows at tail 13-18 (mostly 17). 37 Variation in Pseudechis australis Head and neck black, blackish-brown or brown. Dorsal scales black, blackish-brown or brown, rarely without a cream or brown anterior spot. Scales on lower flanks with more pale pigment than those dorsally. Chin, throat and belly cream, base of each ventral reddish-brown (Figure 1). Iris brick red. Mouth pink or pinkish-grey. Distribution In Western Australia from the far north south to Upper Swan, Yornaning, Marvel Loch, Kalgoorlie, Naretha, Haig and Forrest. Also Sir Graham Moore, Cockatoo, Koolan, Rosemary, Barrow, Bernier, Dorre and Dirk Hartog Islands (Figure 2). Geographic Variation The sample was divided into 10 classes of latitude (one class for every 2°), and ventrals, subcaudals, and percentage of undivided subcaudals were analysed (see Table 1). South of 19°S ventrals and sum of ventrals plus subcaudals vary clinally with latitude. North of \9°S the dine is interrupted. Kimberley specimens have on average more undivided subcaudals and slightly longer tails (Kimberley specimens 14.9-22.2% SVL, mean 19.1, south of Kimberley 13.5-20.1% SVL, mean 18.3). The nasal is separated from the preocular only in the Kimberley (16.6% of specimens). At first these differences seemed to correlate with the Kimberley colour form (see below). However, some of the Kimberley and North-West Division specimens with relatively short tails and few undivided subcaudals have some markings on the head. The lower primary temporal contacts the lower postocular most frequently in the south (43.3% in the South-West Division, 35.4% in the Eastern Division, 33.6% in the North-West Division and 10.3% in the Kimberley Division). Overall dorsal colour of a specimen varies with three factors: (a) colour of pale anterior spot on each scale which varies from cream to brown (spot absent when scales are wholly black); (b) colour of apex of each scale which varies from brown to black; (c) size of pale anterior spot relative to the whole scale which varies from zero (when scale is wholly black) to more than half the scale. An attempt at analysing this variation was made by scoring, for individuals, the colour of their mid-dorsal scales and placing the evaluation into one of three categories for (a) and (b) above and expressing the proportion of pale anterior spot relative to a whole scale as one of three conditions for (c) above. Thus: Score Colour of anterior spot Colour of scale apex Area of pale spot 1 cream brown >!/2 scale 2 brown blackish-brown V 2 scale 3 spot absent (scales wholly black) black <‘/2 scale 38 Variation in Pseudechis australis Results indicated that some colour variations had fairly precise geographic boundaries, otherwise colour types at best can only be described as conhned to) particular areas. Large anterior spots tend to be cream and cream anterior spots generally precede brown apices (giving an overall impression of a pale snake). Brown anterior spots are usually small and precede blackish-brown or black apices (giving an overall impression of a dark snake). Darkest specimens occur south of a line joining Jurien Bay, Badgingarra, New Norcia and Quairading. Here the anterior spot is brown or absent (not cream). Only a few specimens have all black dorsals. Colour of Kimberley specimens is distinctive for two reasons. Firstly, contrast between anterior spot and darker apex of a scale is greatly reduced. Scales are a fairly even brown or slaty gcey. Secondly, sutures of head shields, particularly those between parietals, are dark brown. There are also dark spots and flecks on head shields themselves. Juveniles also have three longitudinal stripes (vertebral and dorsolaterals) which begin on the nape and extend a short distance posteriorly. Head and nape markings reduce with age and are usually absent on large adults, the parietal suture mark persisting longest. Specimens from the vicinity of North West Cape tend to have scales with large cream anterior spots and blackish-brown or black apices, giving specimens a reticulated pattern. Specimens from eastern areas (Gibson and Great Victoria Deserts and Nullarbor Plain) tend to have scales with a small cream anterior spot and the remainder of scale very dark brown, giving specimens a freckled appearance. Two of these specimens had dark freckling on the belly. Apices of scales on Cockatoo Island specimens are a distinctive rich reddish-brown and the anterior spot is cream. predominating in (but not being Individual Variation So far meristic data have been arranged to elucidate geographic variation which masks individual variation. The large sample examined included several series from single localities which indicate the extent of individual variation. Range for characters in Table I for a series of 5 from llgararie Creek was: ventrals 195-212, subcaudals 53-60, sum of ventrals plus subcaudals 248-272 and percentage of undivided subcaudals 62-77%. Series from Kalumburu (Kimberley Division), Warburton Range Mission (Eastern Division) and Kellerberrin (South-West Division) show similar variation. Greatest variation in percentage of undivided subcaudals was at Kalumburu (49-100%). The lower primary temporal was fused to the last labial on one side in two specimens. The third upper labial was divided into 3 on one side of one specimen. Last two upper labials were fused on one side of one specimen, and the upper primary and secondary temporal were fused to the parietal on one side of another specimen. One or three secondary temporals are rare. Material Examined Kimberley Divmon Sir Graham Moore I. (44127); Kalumburu (21853, 28080, 42794); 14 km SSE of Walsh Point (61729); Mitchell Plateau (58318); Drysdale River National Park in 15°02°S. 126°55'E, 40 L.A. Smith \5°i}S'S. 127°()6'E and I5°03'S. 126°44'E (503‘M, 50459 and 50542 respectively); Forest River Mission (12496); Prince Regent River Reserve in 15^07'S, I25°04'E and 15°48'S. 125°20'E ( 47 () 41_42 and 46896 respectively); Kuri Bay (22778, 22925); Kimberley Research Station (9994, 144 ^ 7 ); Kununurra (20569-70); 18 km SE of Kununurra (28106); Cockatoo I. (14073, 14141. 15880, 81959); Kooian I. (29140); VVotjulum (11248, 11722); Lake Argyle (44787, 52666, 70717); 10 km W of One Arm Point (60909); Beagle Bay (8802); Inglis Gap (28077); Veeda (29722); Broome-Derby road (18846); Caniballin (26780); Mt Anderson (58616-17); McHugh Bore (53882); Dampier Downs (58552); LaGrange (28075). North-West Division Rosemai'y 1. (41001); DeGrey Station (5106); Carlindi (15067); Barrow L (28696-98, 48958); Pyramid (25669-70); Marble Bar (18688); 1 12 km VV of Woodstock (28076); 22 km SE of Onslow (22957); Vlaming Head (14012, 19678); Vardie Creek (56129); presumably Exmouth (81420); Wittenoom (11445, 18492); 4 km Sof Mt Bruce (69569); 82 km Eof Mt Bruce (69571); near Tom Price (31019-20); Xingaloo (82039); 82 km E of Point Cloates (25658); Weeli Wolli (22728); Mt Newman (26488): 14 km SE of South Hill (6381 1); Turee Creek (17698, 22785); Mundiwindi (19205, 45090); Ilgararie Creek (22721-25); 10 km N of Kumarina (22731-82); 16 km SE of Kumarina (25151); Carnarvon (18472. 82028); Carnarvon area (25880); Meeragoolia (64702); Bernier 1. (84092, 64483); Dorre I. (13475, 64875); 37 km W of Neds Creek HS (22727, 25208); 5 km SW of Mt Leake (25209); Dirk Hartog 1. (10294, 42397); 3 km SW of False Entrance Well. Carrarang (54730); 19 km E of Hamelin HS (14935); 9 km N of Coburn (66294); Tainala (6519, 6521). Easterji Division 3 km SW of Lens Bore (63356-57); McGuire Gap (63271); Rudall River (40701); 13 km N of Well 17, Canning Stock Route (31323); 16 km X of Beyonde (21527); Carnarvon Range (51917); south end of Carnarvon Range (53647); 30 km E of Everard Junction (22815); 52 km WSW of Everard Junction (60097); Warburton Mission (22073, 22178, 31350-53); presumably Warburton Mission (22003); 48 km SW of Warburton Missioti (28989); .\lbion Downs (30981); Yakabindie (24691); Booylgoo Spring (1179); Cosmo Newbery (13851. 13853); 30 km N of Neale Junction (48755); 27 km^E of Point Sunday (53543); 8 km W of Yainarna MS (28789); 26 km N of Kalgoorlie ( 15608); Kalgoorlie (61623); Burbidge (29489); Marvel Lock (23329); Naretha (14936); 59 km of WNW' of Rawlinna (41228). E ucla Division 19 km SE of Lake Gidgie (39067); 176 km N of Haig (43903); 58 km N of Forrest (14104). South-West Division 30 km E of Kalbarri (36465); Balia (26011); presumably Binnu (24856); 11 km E of Mutt River mouth (28079); East Chapman (4449. 4523); Tenindewa (44551); Geraldton (8594); Bowgada (6325); Bunjil (52107); 10 km SW of Eneabba (30306); 16 km E of Green Head (28331); Wubin (12928); 19 km Eof Wubin (26202); Gunyidi (6141); 16 km ENEofJurien Bay (46133); 10 km E of [urien Bav (60516); Jurien Bay (42380); Namban (21571); 7 km NW of Badgingarra (32082); Badgingarra (17113-14); presumably Badgingarra (21833); presumably Goomberdale (24846); Kulja (5169, 6276); Bindi Bindi (284(M)); Moora (24991); Moora Shire (34346); presumably Moora area (26080-81 ); 8 km S of Moora (43922); 20 km W of Dandaragan (59973); 45 km X of Beacon (48390); Bencubbin (5089); 15 km W of Koorda (37886); Wongan Hills (9763, 59973); Mukinbudin area (34702); Moonijin (.5520); New X'orcia (59979); Bulfsbrook (7 1868); Mangowine, via Xungarin (7852, 7859); Konongorring (12487); Trayning (45975); 48 km E of Dowerin (29490);"l6 km S of Calingiri (14662); Goomalling (10767); 8 km Sof Bolgart( 17089); Bejoording (9775, 10053); Merredin (18502, 22297); 25 km N of Kellerberrin (56542); Baandee (3365, 9988); Kellerberrin and vicinity (22728, 26549-50. 24878, 25973); 19 km NW of Nortbam (20583); Cunderdin (32000); 10 km Eof Northam (24089); Upper Swan (5163); 13 km W of York (22900); 41 Variation in Pseudechis australis Table 1 Geographic variation in ventrals, subcaudals, ventrals plus subcaudals and percentage of undivided subcaudals in Pseudechis australis. Ventrals Percentage Latitude South Ventrals Subcaudals plus subcaudals subcaudals undivided N 5 4 4 4 13-15*=’ range 200-217 63-71 263-279 49.2-100 mean 207.0 66.5 272.0 82.3 N 21 16 15 16 15-17° range 196-220 53-78 259-295 54.2-100 mean 206.7 65.5 274.9 81.7 N 7 6 6 6 17-19° range 196-201 57-63 255-262 57.8-91 mean 198.0 59.1 257.1 70.2 N 7 8 7 8 19-21° range 204-216 54-65 258-279 45.4-76.9 mean 209.0 58.7 268.4 66.0 N 17 17 17 17 21-23° range 198-213 53-64 258-279 50-85.7 mean 208.5 59.1 267.5 69.1 N 22 20 19 19 23-25° range 201-213 53-72 248-276 46.2-100 mean 206.3 57.7 264.1 67.8 N 19 17 16 17 25-27° range 196-209 50-69 254-274 50.7-93.2 mean 203.0 59.1 262.5 67.9 N 13 9 8 9 27-29° range 189-204 51-61 234-267 54-100 mean 197.4 57.0 251.9 68.0 N 30 29 25 29 29-31° range 185-206 51-59 238-262 35.7-96.2 mean 194.5 55.4 250.2 66.5 N 29 32 27 33 31-33° range 186-205 50-61 236-263 45.8-81.8 mean 196.2 57.7 252.3 65.4 1 km SW of York (23341); York area (10559); Quairading area (31321); 13 km E of Quairading (52271); Beverley (68999); presumably Beverley area (23825); Corrigin (64625); 25 km E of Yornaning (50161, 56887). Northern Territory Yirrkala Mission (13525); Adelaide River (9986); Katherine (13982, 16505, 21590, 24930, 26348); 34 km SW of Borroloola (32063); 96 km W of Roper River mouth (32064); Kildurk (36333);3kmSofElliott(47689);64 km N of Tennant Creek (34011); ‘Tennant Creek’ (21514-16). 42 L.A. Smith Pseudechis butleri sp. nov. Holotype R22345 a gravid female (SVL 93 cm) collected 19 km SE of Yalgoo, Western Australia in 28°29'S, 116M9^E, by LC. Carnaby on 15 October 1963. Paratypes North-West Division New Springs (13703); Wurarga (5372, 7395); Yalgoo (7472); 3 km S of Yalgoo (29232); 48 km S of Yalgoo (12919); 56 km Sof Yalgoo (25815); Muralgarra (8177); Barnong (25978); Thundelarra (34694); Fields Find (R13555, 13667); Warriedar (1329). Eastern Division Wonganoo (51101); Boovlgoo Spring (1519, 1380, 7627); 8 km W of Laverton (44550); Laverton (22395); Mt Malcolm (10701). Diagnosis Distinguished from P. australis by its colour, particularly the ventrals which are black- based and bright yellow in butleri (cream with a reddish-brown base in australis). Further distinguished from P. australis (from the same latitudes) by usually having more ventrals (204-216 V. 189-207). Pseudechis porphyriacus of eastern Australia has fewer ventrals (175-210, mean 186.3) and black-based crimson or pink ventrals. Description A large snake (up to 156 cm in total length). Tail 13.9-16.4% of SVL (N 4, mean 15.2). Head broad and moderately distinct in large specimens. Canthus rostralis prominent. Rostral 1. 3-2.0 times as wide as high (N 20, mean 1.6). Frontal 1.1-1. 5 times as long as wide (N 19, mean 1.2). Nasal completely divided; in contact with the preocular (98% of specimens). One preocular, two postoculars. Primary temporals 2, the lower wedged deeply between the fifth and sixth labials and separated from the lower postocular (95% of specimens). Secondary temporals 2. Upper labials 6, third and fourth entering orbit. Lower labials 6, second smallest, fourth largest. Two pairs of chin shields, anterior pair always in contact, postocular pair separated (95%< of specimens). Ventrals 204-216 (N 15, mean 211. 4). Anal divided. Subcaudals 55-65 (N 16, mean 58.6), percentage of undivided subcaudals 35-76 (N 15, mean 59.6). Sum of ventrals plus subcaudals 268-279 (N 15, mean 270). Midbody scale rows 17; scale rows at neck 16-23 (mostly 19); scale rows at tail 15-18 (mostly 17). Rostral, nasals, preoculars, labials (except for a short black subocular streak bordering orbit), chin shields and gulars reddish-brown. Remainder of head and nape black with a reddish-brown tinge. Reddish-brown head and neck most prominent in juveniles. Back with irregular groups of all-black scales. Remainder of dorsals black with yellow (rarely brownish) centres. Most yellow is present on scales of lower flanks (black apices), least mid-dorsally (small yellow spots with broad black margins). Ventral surface bright yellow, base of each ventral always unevenly edged black, remainder of ventrals sometimes flecked black (Figure 1). R7627 arrived at the Western Australian Museum freshly dead. Glauert (1957: 32) described its lighter parts as ‘primrose yellow’. 43 Variation in Pseudechis australis Distribution Arid mid-west of Western Australia north to New Springs, south and west to Barnong and east to vicinity of Laverton (Figure 2). Remarks Glauert (1957: 31) suspected that P. butleri was a new taxon but was loath to place a name on a small series. He was not aware that much of the extensive variation inP. australis was clinal which prevented him from discovering meristic differences between the two species. That P. australis can be very dark (sometimes black) dorsally in the extreme south-w^est of its range further obscured the situation. The distinctiveness of P butleri becomes apparent when its meristics are compared with those of P. australis from similar latitudes (25°-28°S): Ventrals Percentage Ventrals Subcaudals plus subcaudals subcaudals undivided N 32 26 24 26 P. australis range 189-207 50-69 238-267 52.2-100 mean 201.0 58.3 260.3 64.7 N 15 16 15 15 P butleri range 204-216 55-65 268-279 35-76 mean 211.4 58.6 270.0 59.6 Pseudechis australis with all-black scales do not occur in these latitudes. Of the other species oi Pseudechis (australis, colletti, guttatus, papuanus dead porphyriacus) butleri is most like porphyriacus in having 17 midbody scale rows and a similar colour pattern. Cogger (1979: 396) says porphyriarus often has a light brown snout, while McKay (1955: 18) mentions specimens... ‘with a few crimson scales scattered on the dorsal surface...' In butleri the rostral and lateral head shields are reddish-brown and yellow pigment is always present on most dorsal scales. Thus despite obvious differences in the colour of non-black areas of scales, the two species’ patterns are similar, the difference depending on the extent of black pigment. Like other Australian ‘black snakes’ (P. colletti, guttatus and porphyria c^ls) butlerixs not widely distributed compared with P. australis, which is found in all but extreme southern parts of Australia (see Cogger 1979: 395). These relatively localised, disparate distri- butions of the black snakes suggest that P. australis may have expanded its range at their expense. This species is named after Mr W.H. Butler CBE in recognition of his efforts for the cause of conservation in Australia. Acknowledgements 1 am grateful to Dr G.M. Storr. Head, Department of Ornithology and Herpetology, for comments on the manuscript, to Mr S. Wilson for the photograph of P butlen and to Mr R.E. Johnstone for the photograph of P australis. 44 L.A. Smith References Boulenger, G. A. ( 1896). Catalogue of the Snakes in the British Mtiseum (Natural History) 3. (Trustees of the British Museum (Natural History): London.) Cogger, H.G. (1979). Reptiles and Amphibians of Australia. (A.H. 8c A.W. Reed; Sydney.) Covacevich, J. 8c Wombey, J. ( 1976). Recognition of Parademansia microlepidotus (McCoy) (Elapidae), a dangerous Australian snake. Proc. R. Soc. Qd 87: 29-32. Covacevich, J., McDowell, S.B., Tanner, C. and Mengden, G.A. (1981). The relationship of the Taipan Oxyuranus scutellatus and the Small-scaled Snake Oxyuranus microlepidotus (Serpentes: Elapidae). In Proc. Melbourne Herpetological Symposium 19-21 May, 1980. (Eds Banks, C.B. and Martin, A. A.) Zoological Board, Melbourne. Glauert, L. (1957). Handbook of the Snakes of Western Australia. Western Australian Naturalists’ Club. Handbook No. 1, Perth. Kinghorn, J.R. (1923). A new genus of elapine snake from north Australia. Aust. Mus. 14: 42-45. pi. 7. Kinghorn, J.R. (1955). Herpetological notes No. 5. Rec. Aust. Mm. 23: 283-286. Longman, H.A. (1913). Herpetological notes. Part 1 Systematic. Including the description of one new species. Mem. QdMus. 2: 39-42. Loveridge, A. (1934). Australian reptiles in the Museum of Comparative Zoology, Cambridge, Massachusetts. Bull. Mus, Comp. Zool. 77 (6). Mack, G. 8c Gunn, S.B. (1953). De Vis’ types of Australian snakes. Mem. Qd Mus. 13: 58-70. McKay, R.D. (1955). A revision of the genus Pseudechis. Proc. Roy. Zool. Soc. N.S.W. 1953-1954: 15-23. Mitchell, FJ. (1951). The South Australian reptile fauna. Part 1. Ophidia. Rec. S. Aust. Mus. 9: 545-557. Storr, G.M. (1981). The Denisonia goiddii species-group (Serpentes, Elapidae) in Western Australia. Rec. West. Aust. Mus. 8: 501-515. Thompson, D.F. (1930). Observations on venom of large Australian snake, Pseudechis australis (Gray): 1, Synonomy. Aust. J. Exp. Biol. & Med. Sci. 7: 125-133. Received 31 July 1981 Accepted 22 February 1982 Published 30 June 1982 45 Rec. West Aust Mus. 1982, 10 (1): 47-52 The Habitat and Life History of the Pilbara Ningaui Ningaui timealeyi J.N. Dunlop and Maryanne Sawle* Abstract The preferred habitat of the Pilbara ningaui is dense to mid-dense hummock grass- land especially with an upper stratum of open mallee or scrub. Males are larger than females. Litters of 4-6 pouch young or females with distended mammae were observed from September to March. The species is short-lived, with few individuals surviving into a second breeding season. Introduction The ecology of the recently described marsupial genus Ningaui is as yet little known. Two species have been described, Ningaui ridei from central Western Australia and N. timealeyi from the Pilbara (Archer 1975). More species almost certainly exist as specimens of undescribed forms have recently been collected in other arid parts of the continent (DJ. Kitchener, pers. comm.). This paper presents information on the habitat preferences and life history of N. timealeyi based on data collected on a number of biological surveys conducted in the eastern Pilbara between March 1979 and March 1981. Six distinct localities w^ere trapped; three during biological/environmental surveys accompanying mineral exploration and three during a privately organized survey of the Hamersley Range National Park. I'hese localities are shown on Figure 1. Methods Ningauis were captured using pitfall and drift fence traplines. These consisted of 10-12 lined pits (0. 18 m diameter x 0.43 m deep) placed at 4 m intervals along, and on alternate sides of a 35-50 m long flywire fence, 0. 15 m high. Each pit had a minimal catching area of about 12 m of fence and, during the study period, trapping efforts totalled 6333 pit trap days. The vegetation structure of each trapping locality was described using the life-form/density classes of Muir (1977). Although most ningauis captured were released alive after being toe-clipped for subsequent recognition, some v'oucher specimens were collected from each locality and lodged at the Western Australian Museum. Animals were weighed to 0. 1 g using a Pesola spring balance and the maximal scrotum width of males was measured to O.I mm using vernier calipers. Females were examined for pouch and mammae development and for the presence of pouch young. * School of Environmental and Life Sciences, Murdoch University, Western Australia 6150. 47 The Pilbara Ningaui Results and Discussion Habitat Preferences During the study period, 156 ningauis were trapped (excluding recaptures), making them the most common mammal in the study area. They were trapped in all six localities surveyed. The data from all three years have been combined in the analysis of habitat preferences. The habitat (vegetation structure) preferences of N. timealeyi were e\'aluated in terms of relative abundance from trapping data. Capture rates from each of the six localities sampled were based on different trapping efforts and population sizes and are therefore not directly comparable. However, using an abundance index (AI) similar to that devised by Kitchener (1981), results fiom different localities can be standardised for effort and population size giving importance values for each habitat type; n p- (AI) = 10^' 2 Y[ i=l where Pi = proportion of captures in each locality in the ith habitat type Ti = number of trap-days in each habitat of the ith type n = number of ith habitat types sampled. 48 J.N. Dunlop and Maryanne Sawle Although trapping success varied seasonally, with the lowest numbers of captures in winter (June/ July), ningauis were captured in reasonable numbers throughout the year. Data from all sampling periods have been used in this analysis. Abundance indices were calculated for each vegetation stratum sampled using the life-form/height and canopy cover density (LFD) classes of Muir (1977). I’hese values are presented in Muir’s LFD matrix (Table I) along with the total trapping effort for each stratum (in parenthesis). The total A1 values in the LFD matrix indicate a preference in N. timealeyi for open tree or very open shrub mallee or scrub over dense or mid-dense hummock grass. The open tree and shrub mallee strata almost always comprised the same stands and the A I values for these classes could be lumped together. The scrub stratum in the area usually consisted of Mulga Acacia aneina and associated Acacia spp. Both mallee and scrub strata were usually over mid-dense hummock grass. In Mulga low woodland (sparse trees 5-15 m) the hummock grass stratum was absent. No ningauis were trapped in this LFD. The species was also scarce in sparse or very sparse stands of hummock grass including areas which were regenerating after fire. It would appear that a dense or mid-dense hummock grass stratum is an essential component of the habitat of ningauis. Optimal conditions exist where the hummock grassland has an upper stratum of open mallee or scrub. This may be because the animal is to some degree scansorial. One ningaui was observed at night climbing the stem of a low branching shrub. Life History Scrotal size may be used as an indicator of spermatogenesis in dasyurid marsupials (Woolley 1966). The measurements of maximal scrotum width for males trapped in each of five sampling periods (Table 2) suggest that males reached peak sexual development in spring (September/October). Thereafter, those animals continuing in the population showed a decline in gonadal size. Young were produced over the spring May i9ap 17 Survey periods June / Ju ly 1979 I 2 3 4 12 3 4 Weight classes Sept / Oct 1979/80 12 3 4 December 1979/80 12 3 4 Figure 2 The weight distributions iANingaui timealeyi captured at four times of year. 49 The Pilbara Ningaui Table 1 Abundance I ndex values for Ningaui thnealeyi for all habitats (strata) trapped. The life- form/height and canopy cover matrix is that of Muir (1977). Values in parenthesis indicate trapping effort, i.e. total trap-days for each habitat. Life-form/ Height Class Canopy Cover Dense 70-100%d Mid-Dense 30-70%c Sparse 10-30%i Very Sparse 2-10%r T Trees >30 m — M Trees 15-30 m Not sampled Not sampled — — LA Trees 5-15 m - - 0 (100) Not sampled LB Trees <5 in — - 2.0 (346) 6.4 (633) KT Mallee tree form KS Mallee shrub form - 12.9 (174) 3.6 (629) 26.4 (3168) S Shrubs >2 m — 0 (297) 28.4 (1845) 9.8 (1835) SA Shrubs 1. 5-2.0 m - - 5.4 (308) Not sampled SB Shrubs 1.0-1. 5 m - - 4.2 (2558) 0 (179) SC Shrubs 0.5-10. m - - 8.3 (12) — SD Shrubs 0. 0-0.5 m - - 3.6 (140) - P Mat plants — — — — H Hummock grass 17.9 (659) 24.9 (4520) 10.3 (537) 0.5 (316) GT Bunch grass >0.5 m - Not sampled — — GL Bunch grass <0.5 m 0 (40) - - - J Herbaceous spp. - - 0 (100) - VT Sedges >0.5 m — — — — VL Sedges <0.5 m - - - - X Ferns - - - - Mosses, liverworts - - - - * Indicates that IT D is not present in the study area. and summer months (Table 2) with pouch young as early as September in 1979 and females still lactating in late March 1980. During 1978, however, pouch young were recorded only in December, suggesting a shorter breeding period in poorer seasons. Litters consisted of four to six pouch young. Data for March were collected in 1981, for May in 1980 and for June/July during 1979. Animals trapped during September/October and December were of similar sizes in both 1979 and 1980 (Table 3), and data from both years have therefore been combined. Figure 2 presents the weight distributions of all ningauis at Bve different times of year. It is assumed that seasonal weight distributions reflect the age structures of the population at these times. During the breeding period the mean weight of males trapped (7.0 g ± s.e. 0.2) was significantly greater than the mean weight of females (5.8 g ± s.e. 0,3) = +3.67; P < 0.001). Females as small as 5.5 g were recorded with pouch young 50 J.N. Dunlop and Maryanne Sawle Table 2 Breeding data for Pilbara Ningaui, Ningaui timealeyi. Males Females Trapping period Number measured Scrotum Mean width (mm) Range Number recorded Number with unfurred pouch Number with pouch voung (N: CRL)* Number with distended mammae in pouch March 10 4.6 2.0-6.3 7 - - 1 May 4 6.3 5.0-7.0 13 - - - June/July 5 7.0 5.0-8.8 7 - — — September/October 17 9.0 7.1-10.3 6 1 1 (6: 7mm) — December 18 7.7 6.7-9.2 5 2 (5: 11 mm) (6: 6 mm) 4 * (Number of pouch young: Crown-rump length [mm] of pouch young.) indicating maturity at around 5 grams. Males weighing 7.0 g or over were considered to be adults since scrotal width did not increase after this weight. Animals caught in March consisted predominantly of immatures and there were few surviving adults from the previous year. This annual cohort of young moved progress- ively into the larger weight classes in May and June/July (Figure 2). Only four adults were trapped in May and none in June/July. This suggests that adults from the previous year probably disappeared from the population by mid-winter and would not have reproduced a second time. In September/October all animals trapped were adults; males had reached peak breeding condition and some females had developed pouches or were carrying pouch young. The histogram for December was markedly bi-modal with a new cohort of young evident as class 1 and the large breeding and post-breeding adults making up classes 3 and 4. Animals as small as 2.1 g were trapped in December and March. Table 3 The live weights (g) of male and female Ningaui timealeyi captured during September, October and December during both 1979 and 1980. 1979 1980 Trapping period Males Females Males Females September/October 4.6.6.7,7.0, 7.4.7.9.9.4 4.3,6.0,6.5 5. 0. 5. 0.5. 5. 6.0, 6. 0. 6. 5. 6. 5. 7.0, 7.0. 7.5.7.9.8.0, 8.4.8.5.8.5 4. 5, 5. 5* December 5.9, 7.5, 8.4, 8.5 6.5* 3.0. 3.9.6.6.6.7, 6.7.6.7.7.0. 7.0, 7.2,7.3,7.5,7.6, 7.9,7.9,8.8 2. 1,2.5, 2.5, 3.0. 3. 1.3. 3, 3.5.5.9.6.0, 6. 5, 6. 5, 7. 5* Indicates a female with pouch young. 51 The Pilbara Ningaui Conclusion These results suggest that N. timealeyi is a short-lived seasonally breeding species. Probably few, if any, individuals survive into a second breeding season. Populations are therefore dependent on the progress of a single, annual cohort of young. This characteristic could make the species vulnerable to local extinctions if populations were isolated in small pockets of habitat by burning patterns or development. However, the habitat of this ningaui is so widespread in the region that the species is unlikely to be endangered in the foreseeable future. Acknowledgements We gratefully acknowledge the assistance of Cliffs International, Inc., Texasgulf (Australia) Ltd, CRA Ltd, CSR Ltd, Hamersley Iron Pty Ltd, and the Western Australian National Parks Authority. Dr R.D. Wooller kindly commented on the manuscript. References Archer, M. (1975). Ningaui, a new genus of tiny dasyurids (Marsupialia) and two new species, N. timealeyi and N. ridei from arid Western Australia. Mem. Qld. Mm. 17: 237-249. Kitchener, D.J. (1981). Breeding, diet and habitat preference oi Phascogale calura (Gould, 1844) (Marsupialia; Dasyuridae) in the southern wheat belt. Western Australia. Rec. West. Amt. Mm. 9: 173-186. Muir, B.G. (1977). Vegetation and habitat of Bendering Reserve. In Biological Sumey of the Western Australian Wheatbelt, Part l.Rec. West. Amt. Mm. Suppl. No. 3: 7142. Woolley, P. (1966). Reproduction in Antechinus s,pp. and other dasyurid marsupials. In Comparative Biology of Reproduction in Mammals (Ed. I.W. Rowlands): 281-294. (Academic Press: Aberdeen.) Received 3 September 1981 Accepted 7 January 1982 Published 30 June 1982 52 Rec. West. Aust Mus. 1982 , 10 ( 1 ): 53-59 Two New Gehyra (Lacertilia: Gekkonidae) from Australia G.M. Storr* Abstract The new species are G. purpurascens from the arid interior of the continent and G. montium from the Central Australian highlands; both are closely related to G. variegata (Dumeril and Bibron). Gehyra purictata (Fry) is redescribed. Introduction A few of the Australian species of Gehyra are easily identified on peculiarities of their habit, subdigital scales or chin-shields. The remainder, including the widespread and abundant G. variegata, can often be distinguished only after the consideration of several characters, viz. rostral shape, size of postnasal (relative to posterior supranasal), height of first upper labial (relative to second), number of upper labials, subdigital lamellae and pre-anal pores, and certain details of coloration. The present paper adds two more species to this difficult section of the genus. One of the new species ( G. montium) has been confused with G. punctata (Fry); the latter is therefore redescribed. This study is based on material in the Western Australian Museum (R series). Subdigital lamellar counts do not include the distal azygous scale of the pad; upper labials are counted back to the level of centre of eye; and the only internasals counted are those in contact with the rostral. Systematics Gehyra purpurascens sp. nov. Holotype R72660 in Western Australian Museum, collected by A.V. Milewski on 6 October 1980 at 3.5 km NE of Comet Vale, Western Australia, in 29°55'S, 121°08'E. Paratypes One hundred and eighteen specimens in Western Australian Museum from Western Australia, Northern Territory and South Australia. For details see Material. Diagnosis A medium-sized arboreal Gehyra, most like G. variegata but larger (SVL up to 64, v. up to 54 mm) and with ground colour greyish (rather than brownish), dorsal pattern less * Department of Ornithology and Herpetology, Western Australian Museum, fYancis Street, Perth, Western Australia 6000. 53 Two New Gehyra from Australia prominent and lacking white spots or short white bars, top of rostral almost horizontal (more acutely gable-shaped in variegata) and fewer pre-anal pores (often less than II, V. rarely less than 11 in variegata). Distinguishable from G. pilbara Mitchell and G. montium sp. nov. by absence of pale spots and by postnasal not much larger than posterior supranasal, and additionally from pilbara by hrst upper labial not higher than second. Description Snout-vent length (mm): 34-64 (N 119, mean 51.6). Length of tail (% SVL): 87-119 (N 24, mean 104.9). Rostral about half as high as wide; top horizontal or sloping slightly downwards on each side, usually with a small median notch from which a groove descends to about centre of scale. Nostril surrounded by rostral, first labial, postnasal (about same size as posterior supranasal) and two supranasals (anterior much the larger). Internasals 0 (N 13), 1 (95), 2 (6) or 3 (1). Upper labials 7 (N 61), 8 (53) or 9 (4). Anterior chin-shields not in contact with second lower labial. Lamellae under pad of fourth toe in 6 (N 2), 7 (68), 8 (44) or 9 (3) pairs. Pre-anal pores 8-11 (N 58, mean 9.8) in males; pore-bearing scales contiguous and arranged in a chevron, i.e. median pore is anteriormost. Upper surfaces pale purplish-grey, occasionally without pattern but usually with sparse to moderately dense, brownish-grey to blackish-grey markings in form of irregular spots or short streaks. Distribution Arid interior of western two-thirds of Australia: Western Australia from far north of Great Sandy Desert and Tanami Desert south nearly to Beacon, Comet Vale and Queen Victoria Spring (i.e. south to about the mulga-eucalypt line but excluding the Nullarbor Plain); central and southern Northern Territory, north to Elliott; and northern South Australia. See Figure 1. Remarks The name purpurascens is Latin for ‘purplish’. Material Kimberley Division (WA.) 26 km NE McLarty Hills (46049) and 21 km NNE (46087-8). North-West Division (WA.) 33 km NE Bulgamulgardy Soak (63209). Eastern Division (WA.) 92 km S Balgo Mission (47679); Swindell Field (29382); Well 35, Canning Stock Route (40157); 14 km SE Miles Hill (22°30'S, 122°23'E) (63767); 8 km NW Gary lunction (26961); 15 km W Dakota Hills (57253); 72 km W Terry Range (45115) and 7 km E (45221-2); 36 km E Jupiter Well (45210-3) and 65 km E (40152); Pollock Hills (40175, 45173-4, 45176, 57068); 37 km SE Gargoonvah WH (22^52'S, 12U58'E) (63844); Well 24. CSR (63905-6, 63916-7); 60 km N Windy Corner ('45225-6); Durba Springs (51935); Well 11. CSR (51950); Pass of the Abencerrages (20753); 207 km ENE Carnegie (28867) and 65 km NE (40598) and 56 km NE (26883-4); 12 km- SE Mt Beadell (21039); Winburn Rocks (20992-8); Mt Eveline (15702); Warburton Range (16480-1. 21015, 21018, 21020, 22029-31); 64 km E Skipper Knob (37498); 96 km N Neale 54 G.M. Storr • G. purpurascens o G. montium • • Figure 1 Map of western half of Australia showing location of specimens of Gehyra purpurascens and G. montium. 55 Two New Gehyra from Australia Junction (41579) and 70 km N (31909, 41578) and 35 km N (48777); 23 km ENE Cosmo Newbery (73912); 27 km E Point Sunday (53536-9); 138 km NE Laverton (13103a); White Cliffs (21195-7) and 27 km NE (53336-9); 9 km SSE Banjawarn (66039) and 13 km SE (69245-7, 69249-53, 69313-4, 74756, 74792); 3.5 km NE Comet Vale (65799) and 50 km E (72595); Queen Victoria Spring (58728). South-West Division (W.A.) 48 km N Beacon (48342, 48359). Northern Territory Elliott (24185); Renner Springs (74019); Tennant Creek (21381) and 10 km E (21394-5); 23 km N Wauchope (24293-4) and 10 km N (34635-6); 26 km NE Teatree (24375-7); 8 km SW Deep Weil RS (24466); Docker River (20771); 35 km W Victory Downs (20927). South Australia Mt Davies Camp (31702); 10 km E Vokes Hill (36606). Gehyra montium sp. nov. Holotype R31732 in Western Australian Museum, collected by G.M. Storr, J.R. Ford and RJ. Fuller on 29 August 1968 at Mt Lindsay, South Australia, in 27°02'S, 129°53'E. Paratypes Seventy-six specimens in Western Australian Museum from Western Australia, Northern Territory and South Australia. For details see Material. Diagnosis A small rock-inhabiting Gehyra, most like G. variegata but slightly smaller (SVL up to 50, V. up to 54 mm) and with ground colour reddish (rather than brownish), pale dorsal spots usually detached from dark markings (in variegata the white spots or short white transverse bars are contiguous to posterior edge of dark transverse bars and are partly enclosed by these bars when they curve concavely backwards), and top of rostral less acutely gable-shaped. Further distinguishable from G. pilbara by first upper labial not higher than second, and from G. purpurascens by larger postnasal and presence of pale spots on head and back. Description Snout-vent length (mm): 28-50 (N 77, mean 40.7). Length of tail (% SVL): 87-124 (N 29, mean 107.5). Rostral about half as wide as high; top horizontal or sloping slightly downwards on each side, usually with a small median notch from which a groove descends to about centre of scale. Nostril surrounded by rostral, first labial, postnasal (usually much larger than posterior supranasal) and two supranasals (anterior much the larger). Internasals 0 (N 15), 1 (50), 2 (7) or 3 (1). Upper labials 6 (N 20), 7 (42), 8 (1 1) or 9 (2). Anterior chin- shields not in contact with second lower labial. Lamellae under pad of fourth toe in 6 (N 13), 7 (40) or 8 (22 pairs). Pre-anal pores 9-15 (N 30, mean 11.5) in males; pore-bearing scales contiguous and arranged in a chevron, i.e. median pore is anteriormost. 56 G.M. Storr Upper and lateral surfaces pale reddish-brown (more yellowish on head), marked with short blackish-brown streaks, mostly oblique on head (except for two longitudinal stripes on side of head), mostly transverse on back and tail; oblique and transverse streaks sometimes short enough to call spots; transverse streaks sometimes long enough to call cross-bands. Pale spots (especially yellowish spots on head) usually discernible and tending to alternate longitudinally with dark markings on back and tail. Distribution Rocky hills and granite outcrops of Central Australia: central and southern Northern Territory, north to Devils Marbles; far east of Western Australia, west to the Warburton Range; far north-west of South Australia, south to the Birksgate Range. See Figure 1. Geographic Variation Throughout most of its range G. montiurn has mostly 6 or 7 (rarely 8) subdigital lamellae. At Mt Lindsay counts of 8 exceed those of 7, and 6 is unknown. Upper labials are also more numerous at Mt Lindsay than elsewhere. Remarks This species is the chromosome race 2n = 38 of King (1979: 381). The name montium is Latin for ‘of the mountains’. Material Eastern Division (W.A.) Warburton Range (22028, 22032); Barrow Range (20715-23); Cavenagh Range (20730-3); Blackstone Pass (20981-7, 34149); Hinckley Range (31692). Northern Territory Devils Marbles (12 km N Wauchope) (24296-9, 24301); 2 km E Emily Gap (54289, formerly M. King no. 534). South Australia Cave Hill, Musgrave Ranges (20944); Piltadi Rockhole, Mann Ranges (20967); 8 km NW Mt Davies Camp (31698-700) and 18 km S (31712); Krewinkel Hill (31721-3); Mt Lindsay (31732-66, 44364-5) and 28 km NW (31728). Gehyra punctata (Fry, 1914) Peropus variegatus var. punctatus Ery 1914, Rec. West. Aust. Mus. 1: 178. Strelley River, W.A. Gehyra fenestra Mitchell 1965, Senck. biol. 46: 307. Mt Herbert, W.A. Diagnosis A medium-sized rock-inhabiting Gehyra with depressed head and body, swollen nostril region, boldly patterned upper surfaces (large dark and pale spots arranged in alternating transverse rows on back and tail), and large chin-shields, the anterior pair usually in contact with second lower labial. 57 Two New Gehyra from Australia Description Snout-vent length (mm): 23-65 (N 246, mean 44.1). Length of tail (% SVL): 90-133 (N 74, mean 111.7); original tails slender and circular in section. Rostral a little more than half as high as wide; top horizontal or sloping downwards on each side; rarely a small notch at top of groove that descends to about centre of scale. Nostril surrounded by rostral, two supranasals (anterior much the larger), postnasal (very much larger than posterior supranasal and often precluding first labial from nostril), and often first labial. Internasals 0 (N 67), 1 (109), 2 (6) or 3 (2). Upper labials 6 (N 3), 7 (68), 8 (94) or 9 (14), first usually a little higher and considerably narrower than second. Anterior chin-shields (at least on one side) in contact with second lower labial (N 92) or narrowly separated (25). Lamellae under pad of fourth toe in 6 (N 3), 7 (20), 8 (86), 9 (51) or l6 (5) pairs. Pre-anal pores 7-18 (N 53, mean 11.8) in males; pore- bearing scales contiguous and arranged in a chevron, i.e. median pore is anteriormost. Upper and lateral surfaces brown, heavily spotted with blackish-brown and yellow: spots on back and tail arranged in transverse rows, a row of pale spots alternating with a row of dark spots. Lips and two stripes on side of head dark brown. Distribution Rocky hills and granite outcrops in arid north-west of Western Australia from the Pilbara (including several continental islands but not Barrow 1.) south to the Yalgoo district. Also semi-arid south-west Kimberley (Napier Downs). Geographic Variation The isolated Kimberley population, judging from our single male specimen, is notable for the low number (7) of pre-anal pores. Otherwise variation is clinal. From north to south there is a decrease in number of pre-anal pores (9-18 in the Pilbara, v. 8-13 south of the Tropic), a decrease in number of upper labials (mostly 8 north of the Tropic, mostly 7 further south), a decrease in number of subdigital lamellae (9 almost as frequent as 8 north of the Tropic, much less frequent than 8 further south), a decrease in frequency of contact between first upper labial and nostril (usually in narrow contact north of the Tropic, usually not in contact further south), an increase in number of internasals (none as frequent as one north of Tropic, none much less frequent than one further south), and an increase in frequency of contact between anterior chin-shields and second lower labial (83% north of the Tropic, v, 94% further south). From north to south the colour pattern becomes more conspicuous: in the south the spots tend to be larger and less circular (i.e. elliptic with longer axis longitudinal); the pale spots are more richly yellow (golden yellow v. yellowish-white), and the dark spots may be edged with pale yellowish-brown. Remarks Mitchell (1965) divided this species into a larger G. fenestra sp. nov. restricted to the Pilbara and a smaller G. punctata (Fry) widespread in Australia between latitudes 20 and 25 °S. This division was made largely on the number of mesosternal ribs (two in the former, and one in the latter). I have not checked this character for constancy, but 58 G.M. Storr I suspect that Mitchell’s concept of G. punctata was partly based on Pilbara specimens of G. variegata and/or G. pilbara. At any rate I believe that only three species of Gehyra occur in the Pilbara. Material Kimberley Division (W.A.) 9 km SSE Mt Amy (70039, 70554, 70663). North-West Division (W.A.) Strelley River (holotype); 22 km S Port Hedland (52122-4); Depuch I. (14553); Dolphin I. (14245-6, 14284-7, 14292, 37268-72); Angel I. (37250-4); Rosemary I. (14522-9, 37366-70); Enderby L (37344-6); West Lewis L (37329-30); Trimouille 1. (37450); Point Samson (14576-7); Eramurra Creek (16084-6); 10 km NW Yandevarra (25373); Doolcna Pool, Goongan River (63603-5); Marble Bar (16079-80, 51718) and jo km E (16081, 58976-9); Mt Edgar (16034, 16036-65, 16095); Woodie Woodie (63149); 34 km N Nullagine (37014); Abydos (10814); Mt Herbert (16083, 20199, 20203-4); Tambrey (20201); Muiron Is (37231-3); 19 km SW Peedamulla (52121, 52130); 10 km SW Pannawonica (68320-8); Hooley (10821-2); Cockeraga River (39742-3); ‘Fortescue flats’ (presumably N of Wittenoom] (37077); 15 km S Wittenoom (37080-1); Hancock Gorge (69807); Coppin Pool (69800-4, 69808-9); Paraburdoo (56138); Barradale (69997); Lyndon River (8211); 37 km NW Mt Vernon HS (25241); 18 km S Moogooree (62415); Mt Phillips (40635); Yinnietharra (40619-31, 53010, 56853) and 40 km S (53011); Mooka (47845); Landor (52131); 16 km W Dairy Creek HS (24830-4); Coordewandv (28367, 31079, 51692); 23 km N Meekatharra (16087-99); Mileura (15810, 40221-2, 45720-1) and 7 km W (28332-3); Beebyn (54311-2); Afghan Rock (34721-2); Big Bell (31508); Meka (29243-64); Billabalong(51184-5): near Tallering Peak (47713-4): 43 km W Mt Magnet (16100-6); ‘between Yalgoo and Mt Magnet’ (22785); 32 km E Yalgoo (45902-6) and 13 km SE (45902-6) and 30 km W (50062-9, 60502-4); ‘Wurarga granite outcrop’ (30214-25); Muralgarra (7509-10). Eastern Division (W.A.) 32 km E Jiggalong (25207); Mt Davis, Canning Stock Route (28601). References King, M. (1979). Karyotypic evolution in Gehyra (Gekkonidae: Reptilia) 1. The Gehyra variegata- punctata complex. Aust.J. Zool. 27: 373-393. Mitchell, F.J. (1965). Australian geckos assigned to the genus Gehyra Gray (Reptilia: Gekkonidae). Senck. bioL 46: 287-319. Received 30 November 1981 Accepted 22 February 1982 Published 30 June 1982 59 Rec. West Aust Mus. 1982 , 10 ( 1 ): 61-66 Taxonomic Notes on the Genus Tympanocryptis Peters (Lacertilia: Agamidae) G.M. Storr* Abstract Two new subspecies of Tympanocryptis lineata Peters are described: T. 1. houstoni from the Nullarbor Plain and T L rnacra from the Kimberley and neighbouring part of Northern Territory. Tymparwcryptis tetraporophora Lucas and Frost and T. uniformis Mitchell are removed from the Western Australian list, and T. cephala gigas Mitchell is merged in T. cephala Gunther. Introduction Since my revision of Tympanocryptis (Storr 1964) the concept of the genus has been amended. Tympanocryptis parviceps was transferred to Amphibolurus, even though it lacked an external ear opening (Storr 1977). On the other hand aurita was placed in Tympanocryptis despite its exposed tympanum and numerous femoral and pre-anal pores (Storr 1981). Four other matters in my earlier paper require amendment. First, the identification of Kimberley specimens as T. tetraporophora is now seen to be wrong, and the population from which they came is described as a new subspecies of T. lineata. Also wrong was my assumption that the Nullarbor population of T lineata belonged to the nominate race; it too is described as a new subspecies. Recently acquired specimens of T. cephala from the Pilbara are compared with material from the remainder of the species’ range. Finally the specimen of 'Tympanocryptis uniformis’ from the Kimberley is re-identified. With one exception, all the specimens cited in this study are lodged in the Western Australian Museum (R series). Systematics Tympanocryptis lineata macra subsp. nov. Holotype R44553 in Western Australian Museum, collected by L.A. Smith and R.E. Johnstone on 20 January 1972 at 16 km S of main dam at Lake Argyle, Western Australia, in 16°15'S, 128°40'E. * Department of Ornithology and Herpetology, Western Australian Museum, Erancis Street, Perth, Western Australia 6000. 61 Genus Tympanocryptis Peters Paratypes Kimberley Division (W.A.) Ord River below main dam, Lake Argyle (11752); Argyle Downs (42728-9, 42734-5, 44552); Old Lissadell (42672-5); King Sound (Macleay Mus. 930); 5 km NNW Mt Percy (70682); Fitzroy Crossing (75123) and 50 km SE (36164). Northern Territory 40 km SSW Bullo River HS (60330). Diagnosis A moderately large, relatively slender subspecies of T. lineata Peters, most like T. 1. centralis Sternfeld but larger and having longer limbs and tail and more subdigital lamellae. Description Snout-vent length (mm): 36-64 (N 16, mean 51.8; v. 23-61, 41, 44.8 in centralis). Length of appendages (% SVL): foreleg 42-50 (N 14, mean 46.1; v. 36-46, 39, 40.8); hindleg 64-81 (N 14, mean 74.2; v. 52-69, 39, 61.6); tail 137-185 (N 14, mean 166.1; v. 120-175, 37, 149.3). A pre-anal pore discernible in most specimens. Usually no femoral pore (one in one specimen). Lamellae under fourth toe 17-22 (N 14, mean 19.5; v. 15-20, 40, 17.5). Scales on head strongly keeled. Scales on back varying much in size, the largest being spinose and more strongly keeled than others. No midlatera! fold. Gulars weakly keeled and mucronate. Dorsal and lateral ground colour pale reddish-brown to greyish-brown. A pale grey vertebral stripe and a brownish-white to greyish-white dorsolateral stripe occasionally discernible; vertebral stripe no wider than dorsolateral. Reddish-brown to greyish- brown cross-bands on body, limbs and tail, interrupted by the longitudinal stripes and sometimes barely discernible on body. No pattern on head or indication of midlateral stripe. Distribution Semi-arid zone of south and east Kimberley and adjacent part of Northern Territory (see Figure 1). Remarks Previously (Storr 1964), this population was confused with T. tetraporophora Lucas and Frost, a species (or subspecies of T. lineata) confined to the Lake Eyre drainage and characterized by having a femoral pore. The name macra is Latin for ‘lean’. Tympanocryptis lineata houstoni subsp. nov. Figure 2 Holotype R53427 in Western Australian Museum, collected by G. Harold, C?. Barron and M. Peterson on 25 April 1976 at 10 km SSL of Cocklebiddy, Western Australia, in 32°07'S, 126°06'E. 62 G.M. Storr Figure 1 Map of Western Australia showing location of specimens of Tympanocryptis lineata macra, T i centralis and T. 1. hovstoni. 63 Genus Tympanocryptis Peters Figure 2 Holotype of Tympanocryptis lineata houstoni, photographed in life by G. Harold. Paratypes For details of 61 specimens in Western Australian Museum from Western Australia and South Australia, see Material. Diagnosis A large subspecies of T. lineata Peters, distinguishable from all others by its broad vertebral stripe. Further differing from T 1. centralis Sternfeld in its stouter habit, thicker neck and more strongly developed colour pattern (especially on head). Description Snout-vent length (mm): 22-68 (N 62, mean 46.5). Length of appendages (% SVL): foreleg 33-48 (N 59, mean 41.9); hindleg 56-80 (N 59, mean 67.0); tail 134-196 (N 62, mean 156.3). A pre-anal pore discernible in most specimens; 17% of latter also having a femoral pore. Lamellae under fourth toe 18-22 (N 55, mean 19.7). Scales on head strongly keeled. Scales along vertebral stripe weakly keeled; remaining dorsals strongly keeled, spinose scales higher but not much larger than ordinary dorsals. 64 G.M. Storr A slight midlateral fold on body, coincident with white stripe. Gulars smooth, not mucronate. Dorsal and lateral ground colour reddish-brown to greyish-brown. Head variegated with pale and dark brown, including a brownish-white band from orbit to orbit. Broad greyish-white vertebral stripe, 2-4 times as wide as dorsolateral stripe. Narrow creamy- white dorsolateral stripe, sometimes discernible only where crossing dark bands. Chocolate-brown or blackish-brown bands across body and base of tail, interrupted by vertebral stripe, extending down to narrow white midlateral stripe, and widest at contact with vertebral stripe. Throat and venter white, except occasionally for irregular grey streaks or vermiculations. See also description and figure in Houston (1978: 46-47). Distribution Nullarbor Plain of Western Australia and South Australia (see Figure 1). Remarks Previously (Storr 1964), I included this population in the nominate race, but, as Houston (1978) was first to appreciate, it is distinct from that subspecies. Material Eastern Division (W.A.) 16 km NE Fraser Range (14184); Kanandah (39711, 41225); Naretha (19101-4, 51804-6) and 3 km W (29656); 95-115 km NNE Rawlinna (33399, 34022, 36475, 37053-4, 41216, 45358). Eucla Division (W.A.} 20 km E Naretha (12222) and 32 km E (25866); 70 km NNE Rawlinna (36475, 41646, 43592-4); Seymour Downs (19105-10); Rawlinna (15209) and 10 km N (53756-7); l.oongana (29174-5) and 5 km S (28706) and 18 km E (41603); Forrest (16502, 29335); 23 km S Reid (37674); 7 km NNW Eucla (66499-500); Mundrabilla (67261); Madura (24649); Cocklebiddy (67249-53) and 10 km SSE (53428-30); 41 km SW Caiguna (66762); Toolinna Cave (56883); Toolinna Rockhole (45646-7, 66789-92); 20 km SW Balladonia HS (17418). South Australia 77 km S Cook (361 19). Tympanocryptis cephala Gunther, 1867 Tympanocryptis cephalus Gunther 1867, Ann. Mag. nai. Hist. (3) 20: 52. Nickol Bay, W.A. Tympanocryptis cephalus gigas Mitchell 1948, Rec. S. Aust. Mus. 9: 65. Between Ashburton and Gascoyne Rivers, W.A. Remarks For my earlier paper (Storr 1964) I had only a single specimen (12495) from the vicinity of the type locality of T. cephala. As it and the syntypes differed slightly in coloration and scalation from the populations further south and east, the latter were tentatively treated as a distinct subspecies, T. c. gigas. Subsequently the Western Australian Museum has accessed several adult specimens from the Pilbara coastal plain. These specimens do not differ substantially from those from elsewhere; the attempt to divide cephala into subspecies is therefore abandoned. 65 Genus Tympanocryptis Peters Tympanocryptis uniformis Mitchell, 1948 Tympanocryptis uniformis Mitchell 1948. Rec. S. Aust. Mus. 9: 76. Near Darwin, N T. Remarks In my revision (Storr 1964) a juvenile specimen (13638) from 32 km SE of Luluigui in arid south-west Kimberley was identified as T uniformis. I now find that this specimen agrees in coloration and scalation with northern juveniles of Tympanocryptis lineata centralis Sternfeld, e.g. 51275-6 from 50 km SE of Christmas Creek HS. From the latter, 13638 only differs in its shorter snout and more obtrusive eyes; these were the vei^ characters that formerly induced me to identify 13638 as uniformis. However, the original description oinnifoimis indicates a very different lizard, Mitchell particularly drawing attention to its extremely stout body and almost uniform dorsal scalation. References Houston, T. E (1978). Dragon lizards and goannas of South Australia. (South Australian Museum: Adelaide.) Mitchell, E J. (1948). .A revision of the lacertilian genus Tympaiiooyptis. Rec. S. Aust. Mus. 9: 57-86. Storr, G.M. (1964). The agamid lizards of the genus Tympanocryptis inMosicrw Australia./. Proc. R. Soc. West. Am.s/. 47: 43-50. Storr, G.M. (1977). The Amphibolurus adelaidensis group (Lacertilia: Agamidae) in Western Australia. Rec. West. Aust. Mus. 5: 73-81. Storr, G.M. (1981). Three new agamid lizards from Western Australia. Rec. West. Aust. Mm. 8: 599-607. Received 3 December 1981 Accepted 15 February 1982 Published 30 June 1982 66 GUIDE TO AUTHORS Subject Matter Reviews and papers reporting results of research in all branches of natural science and human studies wiU be considered for publication. However, emphasis is placed on studies pertaining to Western Australia. Material must be original and not have been published elsewhere. Authors are advised to follow the layout and style in the most recent issue of the Rec. West. Aust. Mus. including headings, tables, illustrations and references. The title should be concise, informative and contain key words necessary for retrieval by modem searching techniques. Names of new taxa must not be included. An abridged title (not exceeding 50 letter spaces) should be included for use as a running head. An abstract must be given, summarizing the scope of the work and principal findings. It should normaUy not exceed 2% of the paper and should be suitable for reprinting in reference periodicals. Contrary to Recommendation 23 of the International Code of Zoological Nomen- clature it may include names of new taxa. Footnotes are to be avoided, except in papers dealing with historical subjects. The International System of units should be used. Numbers should be spelled out from one to nine in descriptive text; figures used for 10 or more. For associated groups, figures should be used consistently, e.g. 5 to 10, not five to 10. Spelling should follow the Concise Oxford Dictionary. Systematic papers must conform with the International Codes of Botanical and Zoological Nomenclature and, as far as possible, with their recommendations. Synonymies should be given in the short form (taxon, author, date, page) and the full reference cited at the end of the paper. The original and two copies of manuscripts should be submitted to the Publications Officer, Western Australian Museum, Francis Street, Perth, Western Australia 6000. They must be in double-spaced typescript on A4 sheets. All margins should be at least 30 mm wide. Tables plus headings and legends to illustrations should be typed on separate pages. The desired positions for insertion of tables and illustrations in the text should be indicated in pencil. Tables should be numbered consecutively, have headings which make them understandable without reference to the text, and be referred to in the text. Drawings must be suitable for direct photographic reproduction. Photographs must be sub- mitted as high quality black and white prints. The original and two copies (the latter at desired reproduction size) of all iUustrations are required with figure numbers lightly pencUled on the back. Lettering on them must be of appropriate size for reduction, if this will be necessary. Scale must be indicated, preferably on the illustration. All illustrations, whether line drawings or photographs, should be numbered in sequence and referred to as ‘Figure/s’ in the text. Each must have a brief, fully self-explanatory caption. In papers dealing with historical subjects references may be cited as footnotes. In all other papers references must be cited in the text by author and date and all must be listed alpha- betically at the end of the paper. The names of journals are abbreviated according to World List of Scientific Periodicals. The use of ‘unpublished data’ or ‘personal communication’ is dis- couraged. Processing . j j u Papers are reviewed by at least two referees and acceptance or rejection is then decided by an editorial committee. The senior author is sent two sets of galley proofs (one to be retained) and one set ot page proofs which must be returned promptly to the Publications Officer after correction. The senior author will receive fifty free offprints of the paper. Additional offprints can be purchased and should be ordered in advance through the Publications Officer. ;; Contents ^ ■■'■‘x.-'i Storr, G.M. Four New Leris ta (Lacertilia: Scincidae) from Western and South Australia Abbott, Ian The Distribution of Earthworms in the Perth Metropolitan Area Smith, L.A. Variation in Pseudechis australis (Seipentes; Elapidae) in Western Australia and Description of a New Species of Pseudechis Dunlop, J.N. and Sawle, Maryanne The Habitat and Life History of the Pilbara Ningaui Ningaui timealeyi Storr, G,M, Two New Gehyra (Lacertilia: Gekkonidae) from Australia Storr, G.M, Taxonomic Notes on the Genus Tympanocryptis Peters {Lacertilia: Agamidae) - ■ n lte::'V ' : ■ ' !> 1 . ^ 1 . ’ ’ ilk',,. - TERN ^USTRyglA iii |Tr-.-‘ * v. - ■'• •2^;r5?" K''* Wi: .. . ' i, ,- |l-'rk-^5v;tn^^y '7. ■»5 Volume M Records of the Western Australian Museum Editorial Committee Chairman P.F. Berry Natural Science Human Studies G.R. Allen KJ. McNamara LM. Crawford GJ. Henderson M.E. Lofgren Publications Officer A.N. Browne The Records of the Western Australian Museum (Rec. West. Aust. Mus.) is published irregularly with up to four parts appearing each year. A series of supplements, usually devoted to recording the basic data and results of specific Museum studies, is also produced. The journal and supple- ments are available for sale or exchange, the current price per part being $5.00 plus postage. All but a few back issues are available and orders and enquiries should be addressed to the Book- shop, Western Australian Museum, Francis Street, Perth, W.A. 6000, Australia. Cover Spears from the Pilbara Region, Western Australia. Illustrations by Ben Jackson. © Western Australian Museum, 1982 ISSN 0312 3162 Published by the Western Australian Museum, Francis Street, Perth, Western Australia 6000. Printed in Western Australia by Advance Press Pty Ltd. Rec. West Aust Mus. 1982 , 10 ( 2 ): 67-103 A Collection of Freshwater Fishes from Western New Guinea with Descriptions of Two New Species Gerald R. Allen* and M. Boesemanf (Gobiidae and Eleotridae) Abstract Collections of freshwater fishes from western New Guinea (Irian Jaya) are reported. They were procured mainly during two expeditions from the Rijksmuseum van Natuurlijke Historic (Leiden) during 1954-55 and 1959. Collections were made at or in the vicinity of Ajamaru Lakes, Jamur Lake, Wissel Lakes, Digul River at Tanah Merah, Merauke, Japen Island, and the vicinity of Jayapura including Lake Sentani and the Tami River. The material includes 77 species representing 50 genera and 30 families. Two new species are described. Glossogobius hoesei sp. nov. (Gobiidae) from the Ajamaru Lakes region is characterized by a truncate tongue, the absence of a branched pit organ canal below the eye, and a relatively short head (about 26 to 28% of the standard length). Oxyeleotris wisselensis sp. nov. (Eleotridae) from the mountainous Wissel Lakes region is related to the widely distributed lowland species O. fimbriata. It differs, however, with regard to head shape, coloration, and maximum size. A brief diagnosis, illustrations and table of proportional measure- ments are presented for the new species. Other species are treated in an annotated checklist. In addition, a list of the 158 species thus far recorded from fresh waters of New Guinea is appended. A brief zoogeographical discussion of the New Guinea fauna is also included. Introduction The freshwater fish fauna of New Guinea is relatively impoverished compared with the rich cypriniform-dominated fauna lying to the west. The New Guinea species, with the exception of Scleropages jardinii, are secondary freshwater forms having evolved in relatively recent times from marine ancestors. Most of the species thus far documented were collected by Dutch expeditions between 1903 and 1920. Major collectors during this period included de Beaufort (1903 and 1910), Gjellerup (1910-1911), Gooszen (1909), van Heum (1920), van Kampen (1910-1911), Koch (1904), and Lorentz (1907 and 1909). The majority of these collections were summarized by Weber (1908 and 1913). The only major collections since 1920 are those from the Fly, Purari, and Laloki Rivers reported by Roberts (1978), Berra et al. (1975), and Haines (1979) * Department of Ichthyology, Western Australian Museum, Francis Street, Perth, Western Australia 6000. t Rijksmuseum van Natuurlijke Historic, Raamsteeg 2, Leiden, Nederland. 67 Freshwater Fishes from Western New Guinea respectively, and those from western New Guinea reported in the present paper. Most of the latter collections were procured between October 1954 and May 1955 by the second author on an expedition from the Rijksmuseum van Natuurlijke Historie (Leiden) at the request of the prevailing government of western New Guinea. Additional specimens were obtained in the vicinity of Tanah Merah during August and September 1959. These were collected at the conclusion of an RMNH expedition to the Star Mountains under the direction of Dr L.D. Brongersma. As a result of personal contact made during the RMNH expeditions a number of complementary collections were subsequently sent to Leiden, mainly by government civil and naval personnel. Our present collection includes 268 lots containing 2,124 specimens. Seventy- seven species are represented belonging to 50 genera and 30 families. Two species, Figure 1 Map of western New Guinea (Irian Jaya). Numbers denote principle collection areas: (1) Ajamaru Lakes; (2) Jamur Lake; (3) Wissel Lakes; (4) Japen Island; (5) Jayapura; (6) Lake Sentani; (7) Tami River; (8) Tanah Merah (9) Merauke. 68 Gerald R. Allen and M. Boeseman a gobiid and eleotrid, are herein described as new. In addition, Allen and Cross (1980) published descriptions of four new rainbowfishes (Melanotaeniidae) result- ing from these collections. The principal collection areas are indicated in Figure 1. Boeseman (1963) presented detailed information regarding the physiography and climate of the various sites visited by the 1954-55 expedition. A detailed itinerary was also presented with a series of maps and index of geographic place names. The 1959 fishes were taken from the Digul River near the settlement of Tanah Merah mainly between 5 and 13 September. These specimens are indicated by an asterisk (*) in the species section which follows. Families are arranged in phylogenetic sequence following Greenwood et aL (1966). An abbreviated reference is given for the original description of each species. The complete reference appears in the bibliography. Under each species a list of specimen lots is given with an abbreviated locality reference (see below) and the museum registration number followed by the number and size range of the specimens. All lengths are standard length unless indicated otherwise. Annota- tions are included for each species which contain information on the distribution, and in some cases comments on the current status of problematical taxa. Counts and proportions which appear in parentheses in the descriptions of the two new species refer to the range for paratypes.if different from the holotype. Abbreviations Institutions — Lembaga Biologi Nasional, Bogor, Indonesia (LBM); Rijks- museum van Natuurlijke Histoire, Leiden (RAINH); National Museum of Natural History, Washington, D.C. (USNM); and Western Australian Museum, Perth (WAM). Collection localities (see also Boeseman 1963) — AJ — vicinity of Ajamaru on Jow Lake, Vogelkop Peninsula, 3-7 March 1955. AT — vicinity of Aitinjo on Aitinjo Lake, Vogelkop Peninsula, 11-14 March 1955. DA — vicinity of Dimija Village on Dimija River between Paniai and Tage Lakes (Wissel Lakes), 3-9 January 1955. DR — Digul River at Tanah Merah, March and June 1956. DU — vicinity of Djitmau, about 12 km east of Ajamaru, Vogelkop Peninsula, 8-9 March 1955. IR — Ibaru River on the Nimboran Plain about 60 km west of Jayapura near the native villages of Nangkuku and Benjom, 3 November 1954. JI ~ Japen Island near village of Serui, 1954, collected by D.L. Leiker. JL - Jamur Lakes, along lake shore and in small streams in vicinity of Gariau Village, 7-13 December 1954. JR ~ Jawej River at Keniapi Village, Wissel Lakes region, 28-30 December 1954. 69 Freshwater Fishes from Western New Guinea JV — small streams and ponds in vicinity of Jayapura (formerly Hollandia), November 1954. MN — Wosi River, west of Manokwari, 9 March 1955, collected by L.B. Holthuis. MV — small streams in vicinity of Merauke, 4-10 April 1955. PL — streams in vicinity of Paniai Lake, Wissel Lakes, 26-28 December 1954. SL - Sentani Lake near Jayapura, most specimens collected between 20-26 October 1954, but a few taken in September and November 1954 and November 1960. TA — streams in vicinity of Tage Lake, Wissel Lakes, 30 December 1954 to 3 January 1955. TGG— Digul River in vicinity of Tanahtinggi, 10-11 March 1956, collected by Lt Romer, Royal Netherlands Navy. TI — streams in vicinity of Tigi Lake, Wissel Lakes, 11-17 January 1955. TM -- Digul River in vicinity of Tanah Merah, 14-17 April 1955 unless indicated by an asterisk (*) in which case specimens collected 5-13 September 1959. TR — Tami River, cut-off arm (oxbow lake) and main channel, about 22 km SE of Jayapura, 18-21 November 1955. Systematics Family Carcharhinidae . . . Sharks Carcharhinus leucas (Muller and Henle) Carcharias leucas Muller and Henle, 1841: 42 (Antilles). JL (FMNH 24698), male, 148 cm XL; JL (RMNH 24699), male, 146 cm XL; JL (RMNH 24611), female, 73 cm XL; JL (RMNH 24697), female, 125 cm XL. World-wide circumtropical distribution, frequently entering and sometimes breeding in fresh water. The occurrence of this shark in Lake Jamur was reported by Boeseman (1964). Family Pristidae . . . Saw Sharks Pristis microdon Latham Pristis microdon Latham, 1794: 280 (locality unknown). SL (RMNH 28608), 241 cm; SL (RMNH 28609), 284 cm;SL (RMNH 28659), 61.5 cm;XM* (RMNH 28430), 10: 71-79 cm. In addition, the RMNH collection contains 8 saws measuring 200-690 mm taken from specimens collected at Sentani Lake, Moif River near Genjem (see Boeseman 1963), and Digul River near Xanah Merah (RMNH reg. nos D3051-52, D3054-58 and D3026). Widespread in the tropical Indo-Pacific region frequently found in fresh water, particularly in large rivers. 70 Gerald R. Allen and M. Boeseman Family Megalopidae . . . Oxeye Herrings Megalops cyprinoides (Broussonet) Clupea cyprinoides Broussonet, 1782: plate ix (Jamaica, Antigua, Brazil, and New Hebrides). TM (RMNH28431),2: 190 and 225 mm; MV (RMNH 24527), 2: 118 and 128 mm. World-wide circumtropical distribution, occurring in the sea and estuaries, but frequently entering fresh water. Family Clupeidae . . . Herrings Nematalosa erebi (Gunther) Chatoessus erebi Gunther, 1868 (see Gunther 1866): 407 (Mary River, Queensland). JL (RMNH 28432), 23: 27-92 mm; JL (RMNH 28433), 33: 31-83 mm; TM* (RMNH 28434), 6: 79-105 mm;TM (RMNH 28435), 11: 106-148 mm;TM (RMNH 25078), 4: 270-275 mm; TM (RMNH 28436), 2: 203 and 220 mm; TM (RMNH 28437), 5: 74-127 mm; TM* (RMNH 28438), 3: 92-108 mm;JL (RMNH 28439), 10: 33-97 mm. We provisionally follow Nelson and Rothman (1973) in identifying our south- ern New Guinea material as Nematalosa erebi, Roberts (1978) reported two different forms of Nematalosa from the Fly River and stressed the need for a re- evaluation of the New Guinea populations. Distributed in fresh waters of Australia and southern New Guinea. Family Engraulidae . . , Anchovies Thryssa scratchleyi (Ramsay and Ogilby) Engraulis scratchleyi Ramsay and Ogilby, 1887: 18 (Strickland R., New Guinea). TM* (RMNH 28440), 2: 135 and 138 mm. Fresh and brackish waters of northern Australia and southern New Guinea. Family Osteoglossidae . . . Bony Tongues Scleropages jardinii (Kent) Figure 2 Osteoglossum jardinii Kent, 1892: 105 (Batavia River, Cape York, Australia). TM (RMNH 23976), 2: 172 and 182 mm; DR (RMNH 25928), 560 mm; DR (RMNH 25929), 485 mm; TM (RMNH 28441), 2: 470 and 550 mm. Some confusion exists regarding the identity and nomenclature of Scleropages from the New Guinea-Australia region. Most works, for example Fowler (1941), Munro (1956), and Lake (1978) recognized the existence of two species, one from the Fitzroy River system of central-eastern Queensland, and another from a 71 Freshwater Fishes from Western New Guinea few rivers in far northern Australia and central southern New Guinea. Inadequate comparison of the two forms in the literature has caused some speculation that only a single species may exist. Even among authors who recognize two species there is not universal agreement on nomenclature. Fowler (1941) used the names S. leichardti Gunther and S. guntheri Castelnau for the northern and southern forms respectively. Munro (1956) recognized these forms as being distinct only at the subspecific level, assigning the names S. leichardti leichardti and S. leichardti guntheri. Lake (1978) MscdS. jardinii (Kent) for the northern fish and S. leichardti for the southern one. Weber and de Beaufort (1913) recognized only one species from Queensland and New' Guinea, S. leichardti, placing S.jardinii in its synonymy and failed to mention S. guntheri. We have made direct comparisons at RMNH of similar-sized specimens (Figures 2 and 3) belonging to the two forms. They exhibit significant differences related to a number of features. Our findings are summarized in Table 1. The northern form is characterized by a sloping nape profile, a longer more gradually sloping jaw, a more forward directed mouth, and a much larger head. By contrast, the southern or Fitzroy system fish has a straight, non-sloping profile from the dorsal fin origin to the snout tip and a short jaw inclined upward at a very steep angle with the mouth directed more dorsally. Figure 2 Scleropages jardinii (RMNH 28441), 55 cm SL, Digul River near Tanah Merah, Irian Jaya. Figure 3 Scleropages leichardti (RMNH 28613), 56 cm SL, Fitzroy River system, Queens- land, Australia. 72 Gerald R. Allen and M. Boeseman Table 1 Comparison of certain characters of Scleropages jardinii and S. leichardti (based on two specimens of S. jardinii, 455 and 550 mm, Digul River, New Guinea and two specimens of S. leichardti, 470 and 553 mm, Fitzroy River system, Queensland. Character S. jardinii S. leichardti Body depth — % SL 26.8-27.5 24.0-24.2 Head length — % SL 28.7-30.5 22.3-24.4 Angle of mouth (in relation to horizontal axis of body) 41-45° 24-25° Extent of maxillary to well beyond not beyond rear eye of eye Dorsal profile nape arched flat Colour pattern cresentic mark 1-2 spots at at rear of most centre of most scales scales Dorsal rays 20-21 15-16 Anal rays 28-29 25 The only name available for the northern form is S. jardinii described by Kent (1892) from the Batavia (now the Wenlock) and Gregory Rivers of far northern Queensland. This species is presently known from the northernmost section of Cape York Peninsula, the Gregory River flowing into the Gulf of Carpentaria, and the East Alligator system of the Northern Territory, all in the far north of Australia. New Guinea localities include the Fly and Digul Rivers. The oldest name for the southern form is S. leichardti described by Gunther (1864) from the Burdekin and Fitzroy Rivers. Castelnau (1876) failed to give exact locality data in his description of S. guntheri, but from the fin ray counts (D. 17; A. 26) and proportions which are given it is clearly referable to the synonymy of S. leichardti. Castelnau’s type appears to be missing. It is neither in Paris or in the collections of Australian museums. Family Ariidae . . . Fork-tail Catfishes Arius carinatus Weber Arius carinatus Weber, 1913: 537 (Lorentz River, West New Guinea). TM (RMNH 28007), 210 mm SL. Rivers of central-southern New Guinea; thus far recorded from the Purari, Digul, Lorentz, and Fly Rivers. Arius leptaspis (Bleeker) Hexanematichthys leptaspis Bleeker, 1862: 27 (south-east New Guinea). TM (RMNH 28442), 4: 275-370 mm; TM* (RMNH 28385), 335 mm; TM (RMNH 28008), 2: 360 and 380 mm; 73 Freshwater Fishes from Western New Guinea TM (RMNH 28009), 370 mm; TM (RMNH 28810), 2: 360 and 390 mm; JL (RMNH 28443), 7: 45-175 mm. Widely distributed in fresh waters of northern Australia and New Guinea (both north and south of the central dividing range). The identification of this species is provisional pending further studies by P. Kailola. Anus sp. A JL (RMNH 28811), 2: 255 and 350 mm. This species is presently under study by P. Kailola, who is revising the Ariidae of the Australia-New Guinea region. It is very similar to A. australis Gunther, which is widely distributed in northern Australia. Anus sp. B TGG (RMNH 28812), 400 mm; TM (RMNH 28813), 320 mm. According to Kailola, this species which is characterized by a strongly depressed, spatulate head, is closely related to a similar species inhabiting fresh waters of northern Australia. Anus sp. C TR (RMNH 28814), 89 mm. According to Kailola this species, which occurs on the northern side of New Guinea’s central dividing range, has been erroneously referred to as Arius leptaspis by previous authors. The northern population appears to represent a distinct species which is presently under study by Kailola. Arius sp. D SL (RMNH 28815), 6: 225-345 mm. This is a new species which will be described by P. Kailola. It is characterized by a single ovate patch of teeth on each side of the palate. It also occurs in the Ramu and Sepik Rivers of northern Papua New Guinea. Cinetodus froggatti (Ramsay and Ogilby) Arius froggatti Ramsay and Ogilby, 1887: 14 (Strickland River, Nevr Guinea). TM (RMNH 28816), 380 mm. 74 Gerald R. Allen and M. Boeseman CinetoduSy a monotypic genus, was previously thought to be endemic to rivers of central-southern New Guinea. However, according to Kailola (pers. comm.) specimens have been collected in the Roper River, Northern Territory, Australia. Cochief elis spatula (Ramsay and Ogilby) Arius spatula Ramsay and Ogilby, 1887; 15 (Strickland River, New Guinea). TGG (RMNH 28817), 540 mm. Rivers of central-southern New Guinea; known thus far only from the Fly and Digul systems. Hemipimelodus macrorhynchus Weber Hemipimelodus macrorhynchus Weber, 1913. 549 (Lorentz River, New Guinea). TM (RMNH 28818), 6: 179-295 mm. Rivers of central-southern New Guinea; known thus far only from the Purari, Fly, Digul, and Lorentz systems. Hemipimelodus taylori Roberts Hemipimelodus taylori Roberts, 1978:40 (Fly River, Papual New Guinea). TM (RMNH 28444), 59 mm. Recently described from the Fly River of Papua New Guinea. Hemipimelodus velutinus Weber Hemipimelodus velutinus Weber, 1908: 125 (northern New Guinea). SL (RMNH 28445), 250 mm. Fresh water of northern New Guinea between the Tami and Tawarin Rivers. Hemipimelodus sp. TR (RMNH 28819), 3; 246-331 mm. According to Kailola this species is possible undescribed. It is most closely related to H, papillifer of northern New Guinea but differs with regard to anal and pectoral ray counts, and proportional measurements related to the length of the 75 Freshwater Fishes from Western New Guinea adipose fin base and eye diameter. The species is characterized by the following combination of characters (counts and measurements for//, papillifer indicated in parentheses): anal rays 20 to 21 (19); pectoral rays I, 11 (I, 10); adipose fin base in interdorsal space 1.6 to 2.0 (3.4 to 4.0); eye diameter in head length 8.6 to 9.4 (7.0 to 7.6). Thus far known only on the basis of the three specimens collected in the Tami River near Jayapura. Nedystoma dayi (Ramsay and Ogilby) Hemipimelodus dayi Ramsay and Ogilby, 1886: 16 (Strickland River, New Guinea). TM (RMNH 28446), 9; 91-164 mm; TM (RMNH 28820), 5: 135-170 mm. Rivers of central-southern New Guinea. Thus far reported from the Purari, Fly, Digul, and Lorentz systems. Family Plotosidae . . . Eel-tail Catfishes Neosilurus ater (Perugia) Lamhertia atra Perugia, 1894: 551 (Inawi, Papua). TM (RMNH 28447), 3: 308-375 mm; TM (RMNH 28138), 2: 335 and 395 mm; JL (RMNH 28139), 3: 345-375 mm; JL (RMNH 28821), 3: 315-333 mm. Fresh waters of central southern New Guinea and northern Australia. Neosilurus brevidorsalis (Gunther) Copidoglanis brevidorsalis Gunther, 1867: 22 (Cape York, Australia). AJ (RMNH 28448), 2: 105 and 121 mm; JL (RMNH 28134), 2: 60 and 79mm;AJ (RMNH 28135), 3: 81-115 mm; AT (RMNH 28136), 5: 68-110 mm; DU (RMNH 28137), 7: 62-94 mm. Fresh waters of central southern New Guinea and northern Cape York Peninsula, Australia. Family Belonidae . . . Needlefishes Strongylura kreffti (Gunther) Belone kreffti Gunther, 1866: 250 (Australia). JL (RMNH 24674), 180 mm; TM* (RMNH 28449), 6: 395-575 mm. Freshwater streams of New Guinea and northern Australia. 76 Gerald R. Allen and M. Boeseman Family Hemirhamphidae . . . Halfbeaks Zenarchopterus novaeguineae (Weber) Hemiramphus (Zenarchopterus) novae-guineae Weber, 1913: 553 (Lorentz River, southern New Guinea). TGG (RMNH 27386), 163 mm. Freshwater streams and estuaries of central southern New Guinea. Family Melantoaeniidae . . . Rainbowfishes Chilatherina crassispinosa (Weber) Rhombatractus crassispinosa Weber, 1913: 567 (Buarin River, Sepik River, Begowre River, Sermowai River and Tawarin River, northern New Guinea). IR (RMNH 28427), 32 mm. Streams of northern New Guinea between the Markham and Tawarin Rivers. Chilatherina fasciata (Weber) Rhombatractus fasciatus Weber, 1913: 565 (Sermowai River, river near Njao, and tributary of Sepik River, northern New Guinea). JV (RMNH 28424), 62 mm. Streams of northern New Guinea between the Markham and Mamberamo Rivers. Chilatherina sentaniensis (Weber) Rhombatractus sentaniensis Weber, 1908: 235 (Lake Sentani, northern New Guinea). SL (RMNH 28426), 51: 43-78 mm; SL (RMNH 27809), 36: 48-72 mm; SL (RMNH 27810), 15; 59-86 mm; SL (RMNH 27814), 59 mm; SL (RMNH 27815), 3: 72-80 mm. Lake Sentani and nearby Sekanto River, northern New Guinea. Glossolepis incisus Weber Glossolepis incisus Weber, 1908: (Lake Sentani, northern New Guinea). SL (RMNH 28410), 2: 41 and 49 mm; SL (RMNH 28428), 11: 50-78 mm; SL (RMNH 27812), 16: 39-56 mm; SL (RMNH 27813), 21: 38-56 mm; SL (RMNH 27816), 2: 78 and 86 mm; SL (RMNH 27818), 54 mm. Lake Sentani, northern New Guinea. Glossolepis pseudoincisus Allen and Cross Glossolepis pseudoincisus Allen and Cross, 1980: 392 (Tami River, northern New Guinea). TR (RMNH 28072, holotype), 76 mm; TR (LBN 2489, paratypes), 4: 41-64 mm; TR 77 Freshwater Fishes from Western New Guinea (RMNH 28073, paratypes), 41: 33-79 mm; TR (USNM 220907, paratypes), 4: 48-67 mm; TR (WAM P26793-001), 5: 60-77 mm. Known only from an ox-bow lake next to the Tami River, 23 km SE of Jaya- pura, northern New Guinea. Melanotaenis affinis (Weber) Rhombatractus affinis Weber, 1908: 234 (Lake Sentani, Sekanto River, and Wagani Rivers, northern New Guinea). JV (RMNH 28412), 3: 25-37 mm; IR (RMNH 28413), 6: 9-18 mm; IR (RMNH 28420), 13: 25-75 mm; SL (RMNH 28411), 46 mm. Streams and lakes of northern New Guinea between the Markham and Sermowai Rivers. Melanotaenia ajamaruensis Allen and Cross Melanotaenia ajamaruensis Allen and Cross, 1980: 348 (Ajamaru Lakes, western New Guinea). AJ (RMNH 28068, holotype), 78 mm; AJ (LBN 2488, paratypes), 4: 37-45 mm; AJ (RMNH 28069, paratypes), 46: 26-65 mm; AJ (RMNH 28070, paratypes), 6: 32-62 mm; AJ (RMNH 28071, paratype), 57 mm; AJ (USNM 220905, paratypes), 3: 43-59 mm; AJ (WAM P26792- 001, paratypes), 6: 44-68 mm. Ajamaru Lakes System in the central Vogelkop Peninsula of Western New Guinea. Melanotaenia boesemani Allen and Cross Melanotaenia boesemani Allen and Cross, 1980: 379 (Ajamaru Lakes, western New Guinea). AJ (RMNH 28061, holotype), 66 mm; AJ (LBN 2487, paratypes), 3: 50-63 mm; AJ (RMNH 28062, paratypes), 27: 35-63; AJ (RMNH 28063, paratypes), 9: 36-61 mm; DU (RMNH 28064, paratypes), 7: 32-53 mm; AJ (RMNH 28065, paratypes), 3: 27-49 mm; AT (RMNH 28066, paratypes), 3: 54-56 mm; AJ (RMNH 28067, paratypes), 6: 42-87 mm; AJ (USNM 220904, paratypes), 3: 47-53 mm; AJ (WAM P26 79 1-001, paratypes), 3: 50-63 mm. Ajamaru Lakes system and nearby Aitinjo Lake in the central Vogelkop Penin- sula of western New Guinea. Melanotaenia goldiei (Macleay) Aristeus goldiei Macleay, 1883: 269 (Goldie River, southern New Guinea). TM (RMNH 28419), 77 mm; JL (RMNH 28414), 47 mm; JL (RMNH 28422), 41: 16-75; TM (RMNH 24561), 3: 88-97; TGG (RMNH 25263), 100 m; TM* (RMNH 28450), 10: 21-99 mm. Lowland and foothill streams of southern New Guinea and the Aru Islands. 78 Gerald R. Allen and M. Boeseman Melanotaenia japenensis Allen and Cross Melanotaenia japenensis Allen and Cross, 1980: 387 (Japen Island, off northern New Guinea). JI (RMNH 28140, holotype), 77 mm; JI (RMNH 28141, paratypes), 2:57 and 60 mm. Japen Island in the vicinity of Serui, off northern New Guinea. Melanotaenia splendida nibrostriata (Ramsay and Ogilby) Nematocentrus rubrostriatus Ramsay and Ogilby, 1887: 14 (Strickland River, southern New Guinea), MV (RMNH 28415), 9: 14-20 mm; TM (RMNH 28421), 6: 78-103 mm; TM (RMNH 24561), 3: 88-97 mm;TM (RMNH 25263), 100 mm. Lowland streams of central southern New Guinea and the Aru Islands. Family Atherinidae . . . Hardyheads Craterocephalus randi Nichols and Raven Craterocephalus randi Nichols and Raven, 1934: 3 (Kubuna, southern New Guinea). JL (RMNH 28451), 6: 16-60 mm. Our specimens possess about seven longitudinal rows of distinctive dark spots on the sides and in this respect are very similar in appearance to C. stercusmuscar- um of northern Australia. The Craterocephalus of New Guinea require further studies. Known from streams and lakes of central southern New Guinea. Family Syngathidae , . . Pipefishes Doryichthys retzti (Bleeker) Syngnathus retzii Bleeker, 1856: 76 (Celebes). JI (RMNH 25235), 4: 68-102 mm; JI (RMNH 27531), 2: 53 and 59 mm; MN (RMNH 27591), 66 mm. Freshwater streams, tidal creeks and brackish estuaries of the Indo-Australian Archipelago and Philippine Islands. Family Ambassidae . . . Glassfishes Ambassis macleayi (Castelnau) Pseudoambassis macleayi Castelnau, 1878: 43 (Norman River, Queensland). TM (RMNH 28452), 3: 38-56 mm. Fresh waters of central southern New Guinea and northern Australia. 79 Freshwater Fishes from Western New Guinea Ambassis reticulata Weber Ambassis interruptus var. reticulatus Weber, 1913: 574 (Merauke and Lorentz Rivers, southern New Guinea). JL (RMNH 25244), 7: 38-47 mm; JL (RMNH 28453), 4: 34-44 mm. Munro (1967) placed A. reticulata in the synonymy of A, macleayi. We have examined the types of both species and find them to be distinct differing in gill rakers and fin-ray counts and coloration. Ambassis reticulata generally has 17-20 rakers on the lower limb of the first gill arch (25 or more in macleayi), 9V2 soft rays in the second dorsal and anal fins (10^2 in macleayi), and lacks a distinct black marking on the pectoral base which is characteristic for macleayu There is a possibility that our specimens from Lake Majur represent an undescribed species. They have a higher lateral scale count (28-34, usually 30-31 v. 25-28) and a taller first dorsal fin (> than head length vs. < than head) than syntypes of A. reticulata examined at RMNH and ZMA. Fresh waters of central southern New Guinea. Parambassis gullivera (Castelnau) Acanthoperca gulliveri Castelnau, 1878: 45 (Norman River, Queensland). TM (RMNH 25876), 203 mm; TM (RMNH 25879), 4: 78-1 74 mm. Fresh waters of central southern New Guinea and northern Australia. Family Centropomidae . . . Barramundi Lates calcarifer (Bloch) Holocentrus calcarifer Bloch, 1790: 100 (Japan). MV (RMNH 24691), 2: 69 and 75 mm;TM (RMNH 28454), 2: 420 and 460 mm. Widely distributed in estuaries and freshwater streams from the Persian Gulf eastward to southern China and the Indo-Australian Archipelago. Family Lobotidae . . , Tripletails Lobotes surinamensis (Bloch) Holocentrus surinamensis Bloch, 1790: 98 (Surinam), TR (RMNH 28455), 150 mm. Circumtropical distribution in coastal and brackish water, occasionally in lower reaches of freshwater streams. 80 Gerald R. Allen and M. Boeseman Family Teraponidae . . Grunters Amniataba affinis (Mees and Kailola) Therapon affinis Mees and Kailola, 1977; 72 (Morehead and Fly River systems, southern New Guinea). TM (RMNH 28456), 4: 70-142 mm. We follow the recent family revision of Vari (1978) in placing this species in the genus Amniataba, Known from the Fly, Morehead, and Digul River systems of central southern New Guinea. Hephaestus roemeri (Weber) Therapon romeri Weber, 1910: 233 (Lorentz River). TM (RMNH 24936), 2: 117 and 143 mm; TM (RMNH 24941), 123 mm; TM* (RMNH 25909), 88 mm; TM (RMNH 28457), 155 mm. Known only from the Lorentz and Digul Rivers of central southern New Guinea. Pingalla lorentzi (Weber) Helotes lorentzi Weber, 1910: 236 (Lorentz River, southern New Guinea). TM* (RMNH 25390), 2: 50 and 53 mm. Streams of central southern New Guinea and northern portion of Cape York Peninsula, Australia. Terapon jamoerensis (Mees) Therapon jamoerensis Mees, 1971: 214 (Lake Jamur, western New Guinea). JL (RMNH 25225, holotype), 82 mm;JL (RMNH 25224, paratypes), 4: 63-71 mm. Known only from Lake Jamur, western New Guinea. Family Apogonidae . . . Cardinalfishes Glossamia aprion (Richardson) Apogon aprion Richardson, 1842: 16 (near Darwin, Australia). TM (RMNH 24562), 2: 112 and 120 mm; TM (RMNH 28386), 180 mm; TM* (RMNH 28465), 2: 69 and 72 mm; TM (RMNH 28466), 5: 14-35 mm; JL (RMNH 28458), 12 mm; JL (RMNH 28459), 29 mm; JL (RMNH 28460), 54 mm. Coastal streams of central southern New Guinea and northern and eastern Australia. 81 Freshwater Fishes from Western New Guinea Glossamia beauforti (Weber) Apogon beauforti Weber, 1908: 246 (Lake Sentani, northern New Guinea). SL (RMNH 28461) 24 mm. ” Northern New Guinea between Lake Sentani and the Mamberamo River. Glossamia wichmanni (Weber) Apogon wichmanni Weber, 1908: 248 (Lake Sentani, Tawarin River, Moso River, Sekanto River - northern New Guinea). IR (RMNH 28462), 40 mm; SL (RMNH 28463), 2: 25 and 26 mm; SL (RMNH 28464), 81 mm. Northern New Guinea between the Ramu and Tawarin Rivers. Family Silliginidae . . . Sand Whitings Sillago sihama (Forssk^) Atherina sihama Forsskal, 1775: 70 (Arabia). TR (RMNH 24547), 150 mm. Widely distributed in the tropical Indo-west Pacific. Generally a marine or estuarine fish, but occasionally entering the lower reaches of freshwater streams. Family Sciaenidae . . . Croakers Johnius belengerii (Cuvier) Corvina belengerii Cuvier (in Cuvier and Valenciennes), 1830 (Malabar). TM (RMNH 28467), 147 mm; TGG (RMNH 27128), 325 mm; TM (RMNH 27129), 4: 135-203 mm. Widely distributed in shallow seas of the Indo-west Pacific region, frequently entering estuaries and freshwater streams. Family Toxotidae . . . Archerfishes Toxotes chatareus (Hamilton) Coins chatareus Hamilton, 1822: 101 and 370 (Ganges River, India). JL (RMNH 27808), 2: 50 and 71 mm; JL (RMNH 27811), 7: 23-108 mm;JL (RMNH 28468), 3; 20-24 mm; JL (RMNH 28469), 8; 10-24 mm; TM (RMNH 28384), 230 mm. Estuaries and freshwater streams of South-East Asia (India to China), Malaysia, Indonesia, New Guinea, and northern Australia. 82 Gerald R. Allen and M. Boeseman Toxotes jaculatrix (Pallas) Sciaenia jaculatrix Pallas, 1767: 186 (Batavia Jakarta, Java), JL (RMNH 28470), 21 mm. Widely distributed between India and the New Hebrides. Usually found in salt or brackish conditions, seldom entering pure fresh water. Family Scatophagidae . . . Scats Scatophagus argus (Linnaeus) Chaetodon argus Linnaeus, 1766: 464 (Indies), MV (RMNH 28471), 11 mm. Widely distributed in coastal seas of the Indo-west Pacific region, frequently in fresh or brackish water, particularly juveniles. Family Mugilidae . . . Mullets Liza dussumieri (Valenciennes) Mugil dussumieri Valenciennes (in Cuvier and Valenciennes), 1836: 147 (Bombay and Coro- mandel), SL (RMNH 28472), 146 mm. Widely distributed in the tropical Indo-west Pacific, frequently found in estuaries and fresh water. Liza macrolepis (Smith) Mugil macrolepis Smith, 1849: pi. 28 (South Africa). TM (RMNH 28473), 2: 158 and 162 mm; TM (RMNH 28474), 4: 355-390 mm. Widely distributed in the tropical Indo-west Pacific, frequently found in estuaries and rivers. Valamugil seheli (Forsskal) Mugil seheli Forsskal, 1775: 73 (Lohajae, Red Sea), SL (RMNH 28475), 3: 227-240 mm. Widely distributed in the tropical Indo-west Pacific, frequently found in estuaries and fresh water. Family Gobiidae . . . Gobies Glossogobius aureus Akihito and Meguro Glossogohius aureus Akihito and Meguro, 1975: 128 (Okinawa, Japan). JL (RMNH 28476), 4: 35-106 mm; SL (RMNH 28477), 127 mm; SL (RMNH 28478), 5: 31-69 mm; SL (RMNH 83 Freshwater Fishes from Western New Guinea 28479), 14: 26-41 mm; SL (RMNH 28480), 103 mm; JL (RMNH 28481), 68 mm; TR (RMNH 28482), 142 mm; TM (RMNH 28483), 149 mm; SL (RMNH 25233), 5: 35-74 mm; JL (RMNH 25240), 73 mm; SL (RMNH 28485), 2: 87 and 150 mm; SL (RMNH 28486), 6: 24-43 mm; TM* (RMNH 28487), 2: 74 and 82 mm. Widely distributed in the western tropical Pacific including Okinawa, Taiwan, Philippine Islands, Thailand, Malaysia, Singapore, Indonesia, New Guinea, and northern Australia. Usually found in fresh water. Glossogobius celebius (Valenciennes) Gohius celebius Valenciennes (in Cuvier and Valenciennes, 1837: 69 (Celebes). TM* (RMNH 28488), 2: 61 and 71 mm. Widely distributed in the western tropical Pacific including Okinawa, Philippine Islands, Indonesia, New Guinea and northern Australia. Usually found in fresh or brackish water. Glossogobius hoesei sp. nov. Figures 4 and 5; Table 2 Holotype RMNH 28560, 63.2 mm, Jow Lake in vicinity of Ajamaru, Vogelkop Peninsula, Irian Jaya, Indonesia (approximately 1°2US, 132®16'E), M. Boeseman, 3-7 March 1955. Paratypes AJ (RMNH 28489), 20: 36-65 mm; AJ (RMNH 28490), 41: 29-71 mm; AJ (RMNH 28491), 54: 36-70 mm; DU (RMNH 28492), 16: 28-50 mm; AJ (RMNH 28493), 35: 43-67 mm; AJ (WAM P27387-001), 10: 50-65 mm. Diagnosis A species of Glossogobius which closely resembles G. concavifrons, but differs in lacking a branched pit-organ canal below the eye, in possessing a slightly shorter head (about 26 to 28% of standard length v. 28 to 32%), and usually 16 pectoral rays (17 or 18 in concavifrons). Description Dorsal rays VI-1,11 (9 to 12); anal rays, 1,9 (9 or 10); pectoral rays 16 (14 to 17); gill rakers poorly developed, about 5 or 6 low rudiments on lower limb of first branchial arch; scales in lateral series 31 (31 or 32); horizontal scale rows between anal fin origin and dorsal fin base 10; predorsal scales 15 (15 to 17). Body elongate, laterally compressed in posterior portion, more cylindrical anteriorly; maximum depth 5.2 (4.8 to 5.4) in standard length. Head blunt with moderately produced, rounded snout; maximum depth of head about equal to its maximum width or 1.9 (1.7 to 2.0) in standard length; length of head 3.6 84 Gerald R. Allen and M. Boeseman (3.5 to 3.7) in standard length. Interorbital narrow, its width 11.6 (9.8 to 11.2) in head length. Snout 3.6 (3.4 to 3.9), eye 4.4 (4.2 to 5.2), both in head length. Lower jaw slightly produced; teeth canine-like arranged in several rows in both upper and lower Jaws, those of outer row enlarged; palate endentulous; tongue notched (sometimes not apparent if tongue folded); maxillary extends to level of front of eye or slightly anterior to this point. Gill opening extends to level of posterior preopercular margin. A series of six horizontal pit organ canals on cheek, each canal composed of a single row of pit organs (see Figure 5 for arrange- ment of cephalic sensory canals). A prominent, rounded bony protrusion on lower jaw on side of isthmus. Figure 4 Glossogobius hoesei, holotype, 63.2 mm, Ajamaru Lakes. Figure 5 Cephalic sensory canals and pores of Glossogobius hoesei. The numbering system follows that of Akihito and Meguro (1975). 85 Freshwater Fishes from Western New Guinea Table 2 Proportional measurements of selected type specimens of Glossogobius hoesei (expressed as a percentage of the standard length). Holotype Paratypes RMNH RMNH 28489 28560 Standard length (mm) 63.2 67.9 60.3 58.5 58.0 50.0 Body depth 19.0 18.4 19.9 20.6 19.1 19.1 Head length 27.6 28.1 28.2 28.5 27.3 27.5 Head depth 14.7 13.8 16.6 14.9 15.9 16.0 Head width 15.8 16.5 15.9 14.5 16.4 14.8 Snout length 7.6 7.2 8.3 7.9 7.6 7.6 Eye diameter 6.3 6.2 6.6 5.5 6.2 6.4 Interorbital width 2.4 2.5 2.7 2.9 2.8 2.6 Pectoral fin length 23.7 22.8 25.7 23.1 26.2 24.6 Pelvic fin length 21.4 23.0 24.2 22.6 23.3 23.8 Caudal fin length 24.5* 28.7 31.5 28.0 29.3 30.7 fin damaged Scales of head and body cycloid or finely ctenoid. Preopercle, opercle, inter- orbital, snout, lips, chin, and lower jaw scaleless. Pectoral and pelvic fins relatively elongate, their lengths 1.2 (1.0 to 1.2) and 1.3 (1.2 to 1,3) respectively in head length. Pelvic fins united. Caudal fin oblong, its posterior margin rounded, its length 1.1 (0.9 to 1.1) in head length. Colour in alcohol; generally light tan, scales on upper back dusky brown; a series of 5 or 6 diffuse brown blotches, about twice size of eye on middle of side; a distinct, circular brown spot, about eye size or slightly larger at base of caudal fin; head with dusky brown band from lower, anterior corner of eye to pre- maxillary; a large brown spot on lower half of opercle; first dorsal fin pale tan basally with broad dark brown to blackish band across middle portion, outer margin of fin pale tan; second dorsal fin with faint alternating dark and light stripes; caudal fin with series of faint, vertical brown bands; anal fin whitish on basal half, brownish distally; pelvic fins mainly dark brown; pectoral fins pale tan with brown bar across base of uppermost rays. Remarks Glossogobius hoesei is thus far known only from the vicinity of the Ajamaru Lakes which are located near the centre of the Vogelkop Peninsula at the western extremity of Irian Jaya. The lakes are situated at the headwaters of the Ajamaru River which drain into the Kais River, eventually flowing southward to the Ceram Sea. Boeseman (1963) gave further details for the area in which the type specimens were collected. He recorded an elevation of about 250 m for the main system of lakes and a pH of 6.4. Two species of rainbowfishes (Melanotaeniidae), 86 Gerald R. Allen and M. Boeseman which are possibly endemic to the Ajamaru Lakes system, have been described by Allen and Cross (1980). Glossogobius hoesei is very similar in many respects to G. concavifrons (Ramsay and Ogilby) previously known only from the Fly River system of south- ern New Guinea, but recently collected by D. Hoese and the senior author from several streams near the northern extremity of Cape York Peninsula, Australia. The important difference between these species are indicated in the diagnosis section above. Both G. concavifrons and G. hoesei share a modal count of 11 soft dorsal rays, which is relatively high for the genus. Akihito and Meguro (1975) presented brief diagnoses, cephalic sensory canal diagrams, and a key to the species of Glossogobius. The species is named hoesei in honour of Dr Douglass F. Hoese, Curator of Ichthyology at the Australian Museum, Sydney, in recognition of his contribu- tions to the knowledge of gobiid taxonomy. According to the second author’s field notes the local name for this species is ‘buseek’. Glossogobius koragensis Herre Glossogobius koragensis Herre, 1935: 419 (Sepik River, New Guinea). SL (RMNH 28484), 3: 96-142 mm. Known only from northern New Guinea between the Sepik River and Lake Sentani. Inhabits fresh water. Mugilogobius sp. TR (RMNH 28494), 15: 10-23 mm. Counts of 5 specimens as follows: dorsal rays VI-1,8; anal rays 1,8; pectoral rays 16; vertical scale rows from upper corner of gill cover to caudal fin base 33 or 34; horizontal scale rows from anal fin origin to base of dorsal fin 9 or 10; predorsal scales 14-16, extending to rear of interorbital. Colour pale tan with series of about 10 irregular branched brown bars along sides and a pair of distinctive dark brown spots at base of caudal fin. There is a black blotch on the distal portion of the first dorsal fin. Stiphodon elegans (Steindachner) Sicydium elegans Steindachner, 1880: 152 (Society Islands). JV (RMNH 28495), 31 mm. Widely distributed in freshwater streams of islands in the tropical western Pacific from Sumatra eastward to the Marquesas and Society Islands. 87 Freshwater Fishes from Western New Guinea Family Eleotridae . . . Gudgeons Bostrychus strigogenys Nichols Bostrychus strigogenys Nichols, 1937: 1 (Fly River, southern New Guinea). TM* (RMNH 28496), 12: 26-113 mm; (WAM P27388-001), 4: 35-90 mm. Freshwater streams of central southern New Guinea. Hypseleotris cyprinoides (Valenciennes) Eleotris cyprinoides Valenciennes (in Cuvier and Valenciennes), 1837: 248 (Mauritius). JV (RMNH 28497), 8: 17-38 mm. Generally similar to Hypseleotris guntheri (Bleeker) as described in Koumans (1953), but lacking a dark longitudinal band on the body and having about 19 predorsal scales instead of 15. Most of our specimens show distinct spotting on both dorsal fins, caudal fin, and anal fin. Other distinctive markings include a blackish eye-size spot at the middle of the caudal fin base and a narrow blackish bar on the pectoral fin base. Widely distributed in the Indo-West Pacific. Inhabits streams and estuaries. Mogurnda mogurnda (Richardson) Eleotris mogurnda Richardson, 1844: 4 (vicinity of Darwin, Australia). JL (RMNH 28498), 3: 55-72 mm; JL (RMNH 28499), 3: 35-50 mm; JL (RMNH 28500), 83 mm; JL (RMNH 25251), 4: 62-97 mm; TM* (RMNH 25908), 3: 78-97 mm; TM* (RMNH 28501), 242: 18-110 mm; TM (WAM P27388-002), 35: 32-115 mm. Freshwater streams and lakes of southern New Guinea and northern Australia. Ophieleotris aporos (Bleeker) Eleotris aporos Bleeker, 1854: 59 (Halmahera). JL (RMNH 28502), 31 mm; JL (RMNH 28503), 3: 69-97 mm; JL (RMNH 28504), 6: 58-93 mm; JL (RMNH 28505), 3: 49-109 mm; JL (RMNH 25212), 50 mm; JL (RMNH 25231), 2: 70 and 83 mm; JL (RMNH 28506), 3: 27-40 mm; SL (RMNH 28507), 7: 42-72 mm; SL (RMNH 28508), 156: 11-46 mm; SL (RMNH 28509), 17: 16-42 mm; SL (RMNH 24530), 3: 135-168 mm; SL (RMNH 25081), 2: 102 and 135 mm; SL (RMNH 28510), 4: 111-127 mm; SL (RMNH 28511), 30: 9-40 mm; TR (RMNH 28512), 3: 66-82 mm; TR (RMNH 25110), 140 mm; JI (RMNH 28513), 1 10 mm. Widely distributed in the tropical Indo-West Pacific from Madagascar eastward to Melanesia. Commonly found in brackish estuaries, streams, and lakes. 88 Gerald R. Allen and M. Boeseman Oxyeleotris fimbriata (Weber) Eleotris fimbriatus Weber, 1908: 254 (Etna Bay, southern New Guinea). AJ (RMNH 28514), 20: 57-113 mm; AJ (WAM P27387-002), 7: 78-93 mm; AJ (RMNH 28515), 11: 54-110 mm; AJ (RMNH 28516), 43 mm; AJ (RMNH 28517), 123 mm; AT (RMNH 28518), 3: 99-142 mm; AT (RMNH 24576), 3: 77-110 mm; AJ (RMNH 28519), 3: 41-45 mm; DU (RMNH 28520), 18: 31-84 mm; JL (RMNH 25219), 3: 73-114 mm; JL (RMNH 28521), 83: 24-112 mm; JL (WAM P27390-001), 10: 33-112 mm;JL (RMNH 28522), 2: 44 and 66 mm; JL (RMNH 28523), 23: 12-52 mm; JL (RMNH 24557), 7: 41-107 mm;TM* (RMNH 28528), 150 mm; TM*(RMNH 28524), 49: 24-143 mm; TM (WAMP P27388-003), 11: 30- 111 mm. We concur with Roberts (1978) who mentioned that the New Guinea Oxyeleo- tris are in need of systematic revision, particularly the members of the 'fimbriata complex’. The material we have identified as O, fimbriata is probably divisable into at least two species. Specimens from the Ajamaru Lakes region of the Vogelkop Peninsula possess smaller scales (about 70-80 in lateral series and 36-42 predorsal scales) than those from Lake Jamur and Tanah Merah on the Digul River. Specimens from the latter area generally have 55 to 65 scales in lateral series and about 28 to 35 predorsal scales. In addition, the colour pattern lacks the extensive blotching characteristic of much of our Ajamaru Lakes material. Reported from fresh waters of both northern and southern New Guinea by Koumans (1953), but further investigations may indicate that fimbriata is restrict- ed to southern drainages. Oxyeleotris lineolatus (Steindachner) Eleotris lineolatus Steindachner, 1867: 13 (Rockhampton, Queensland). SL (RMNH 28529), 114 mm; SL (RMNH 28530), 11: 32-57 mm; SL (RMNH 28531), 156 mm; JI (RMNH 28532), 2: 78 and 79 mm; JV (RMNH 28533), 2: 54 and 63 mm; JV (RMNH 28534), 5: 31-49 mm; JV (RMNH 25099), 105 mm; JL (RMNH 28525), 4. 19-51 mm; TM (RMNH 28526), 2: 30 and 33 mm; TM (RMNH 28527), 3: 283-390 mm. Freshwater streams and lakes of both northern and southern New Guinea. Oxyeleotris nullipora Roberts Oxyeleotris nullipora Roberts, 1978: 67 (Fly River, southern New Guinea). TM (RMNH 28535), 3: 23-27 mm; TM (WAM P27388-004), 2: 22 and 23 mm. Recently described from lacustrine and semi-lacustrine habitats of the Middle Fly River. Oxyeleotris paucipora Roberts Oxyeleotris paucipora Roberts, 1978: 67 (Fly River, southern New Guinea). TM (RMNH 28536), 10: 28-33 mm; TM (WAM P27388-005), 2: 25 and 41 mm. Recently described from riverine habitats of the Upper Fly River. 89 Freshwater Fishes from Western New Guinea Oxyeleotris wisselensis sp. nov. Figures 6 and 7 ; Table 3 Holotype RMNH 28541, 111.0 mm, small streams in vicinity of Tigi Lake, Wissel Lakes, Irian Jaya, Indonesia (approximately 4°09'S, 136°13T), M. Boeseman, 11-17 January 1955. Paratypes TA (RMNH 28537), 16: 18-92 mm; PL (RMNH 28532), 10: 27-91 mm; TI (RMNH 28539), 10: 32-68 mm; JR (RMNH 28540), 5: 85-96 mm; TI (RMNH 28541), 12: 73-109 mm; TA (RMNH 28542), 32: 17-85 mm; DA (RMNH 28543), 53: 20-90 mm; DA (RMNH 28544), 43: 54-95 mm; PL (RMNH 28545), 68 mm; JR (RMNH 28546), 51 mm; TA (WAM P27389- 001), 17: 26-82 mm. Diagnosis A species of Oxyeleotris which is allied to O. fimbriata, but differs from it on the basis of a less depressed head shape and a much shorter snout (Figure 7). In addition, O. wisselensis is considerably darker in overall coloration and lacks the 3-4 lines which radiate from the eye of O. fimbriata. The fins of the latter species are pale in vivid contrast to the dark fins of O. wisselensis. Moreover, O. fimbriata has distinct spotting which covers the entire second dorsal fin, whereas in O. wisselensis there are wavy brown lines interspersed with white and this feature is restricted to the basal half of the fin. Finally, the dark spot at the upper caudal fin base is more clearly evident in the adults of O. fimbriata. Description Dorsal rays VI-1,11 (10 to 12); anal rays 1,9 (8 to ll);pectoral rays 17 (16 to 18); gill rakers on first branchial arch 1 + 8 (1 or 2 + 7 to 9); scales in lateral series 66 (63 to 70); predorsal scales 26 (25 to 30). Body elongate, laterally compressed in posterior portion, more or less cylindri- cal anteriorly; maximum depth 6.0 (4.9 to 6.1) in standard length. Head blunt with rounded snout; maximum depth of head 1.9 (1.9 to 2.2), maximum width 1.9 (1.5 to 1.9), both in length of head, which is equal to 3.1 (3.1 to 3.3) of the standard length. Interorbital convex, its width 4.0 (3.6 to 4.6) in head length. Snout 4.3 (3.1 to 4.0), eye 5.3 (5.5 to 7.1), both in head length. Mouth terminal; edge of lips fimbriate; jaw teeth numerous in dense bands, teeth of outer row somewhat enlarged; palate endentulous; maxillary extends to level of below middle of eye. A series of 6 or 7 vertical pit organ canals below eye and area between anterior and posterior nares liberally covered with pit organs. Several conspicuous sensory pores on each side of head as follows. 2 nasal pores; 1 supraorbital pore; 5 preopercle pores; 2 pores slightly above and anterior to upper limit of opercle opening; and a single pore in middle of interorbital. Scales of head and body cycloid. Head entirely scaled except for lips, tip of snout, preorbital region, lower jaw, and chin. Scales of interorbital and cheeks generally smaller than body scales and tend to be embedded. 90 Gerald R. Allen and M. Boeseman Figure 6 Oxyeleotris wisselensis, paratype (RMNH 28541), 85.0 mm, Tigi Lake. Figure 7 Comparison of head shapes of Oxyeleotris wisselensis, 62 mm (left) and O. fimbriata, 67 mm. Drawings made at same magnification with aid of camera lucida. Table 3 Proportional measurements of selected type specimens of Oxyeleotris wisselensis (expressed as a percentage of the standard length). Holotype RMNH 28541 Paratypes Standard length (mm) 111.0 103.0 93.7 89.6 84.2 74.4 Body depth 20.4 18.7 20.4 16.7 20.1 18.8 Head length 31.9 32.3 31.5 32.5 32.1 31.6 Head depth 15.3 17.4 14.1 17.0 16.0 16.1 Head width 20.4 21.8 18.7 17.4 21.0 20.2 Snout length 8.4 10.3 8.2 7.5 8.3 7.7 Eye diameter 4.5 4.7 4.8 6.1 4.8 5.4 Interorbital width 9.0 8.9 7.5 8.0 8.6 6.9 Pectoral fin length 18.9 20.8 20.5 18.4 21.5 20.0 Pelvic fin length 16.0 15.8 16.0 16.7 15.2 17.5 Caudal fin length 20.7 21.2 21.3 20.8 20.3 20.8 91 Freshwater Fishes from Western New Guinea Pelvic fins relatively small compared to fan-like pectoral fins; pelvic fin length 1.9 (1.8 to 2.1) and pectoral fin length 1.8 (1.5 to 1.8), both in head length. Bases of pelvic fins widely spearated. Pectoral and caudal fins rounded, length of caudal fin 1.6 (1.5 to 1.6) in head length. Colour in alcohol: generally dark brown on upper half grading to light brown on ventral portion; dorsal fins dark brown, second dorsal with faint wavy brown lines interspersed with white on lower half; anal fin light brown basally, dark brown on outer half; second dorsal and anal fins with narrow white margin; remaining fins dark brown. Juvenile specimens generally light brown with dark brown head and series of dark chevron markings along side of body; a vague ocellus-like marking at base of upper caudal rays, this mark less evident with increasing size, but visible in adults. Colour in life: according to field notes the general colour is greyish-brown or occasionally yellowish with shades of green or gold. Chevron markings on young are brown to light brown or greyish. Remarks Oxyeleotris wisselensis appears to be restricted to the Wissel Lakes and their tributary streams. These lakes are situated in the central mountain chain of Irian Jaya at elevations ranging from 1640 to 1750 m. Paniai Lake is the largest with a length of 16 km and width fo 9 km, and a maximum depth of at least 50 m. The other two lakes, Tage and Tigi have a combined area about equal to half that of Paniai Lake. Additional details of the environment of this region were provided by Boeseman (1963). Oxyeleotris wisselensis is most closely allied and perhaps derived from O. fimbriata, a widely distributed species occurring in lowland areas of New Guinea on both sides of the central dividing range. Koumans (1949) discussed the great variability in fin ray and scale counts found in this species. He noted a tendency for Wissel Lakes specimens to have an additional ray in the second dorsal and anal fins compared with specimens of O. fimbriata from other localities. There is also a pronounced difference in the maximum size attained by the two species. The largest of our type series of 201 specimens is 111 mm SL compared with a maximum standard length of 143 mm for 258 specimens of O. fimbriata. Koumans (1949) recorded maximum standard lengths of 115 mm and 225 mm for the Wissel Lakes fish and lowland populations of O. fimbriata respectively. Additional studies are required to properly assess the taxonomic status of the many lowland populations of O. fimbriata. It is conceivable that this ‘species’ may be divisable into several distinct taxa (see Discussion section for this species). We have com- pared the specimens from the Wissel Lakes with 258 specimens of O. fimbriata from lowland streams of southern New Guinea. The species is named in reference to the type locality. 92 Gerald R. Allen and M. Boeseman Prionobutis microps (Weber) Pogoneleotris microps Weber, 1908: 258 (Tawarin River and Merauke River, New Guinea). TR (RMNH 24546), 150 mm;JV (RMNH 28547), 170 mm. Fresh and brackish waters of New Guinea (north and south), and northern Australia. Family Periophthalmidae . . . Mud Skippers Periophthalmus novaeguineaensis Eggert Periophthalmus cantonensis novaeguineaensis Eggert, 1935: 67 (Merauke River, southern New Guinea). MV (RMNH 28548), 2: both 32 mm. Fresh and brackish waters of central southern New Guinea. Family Kurtidae . . . Nurseryfishes Kurtus gulliveri Castelnau Kurtus gulliveri Castelnau, 1878: 233 (Norman River, Queensland). TM (RMNH 28549), 3: 260-305 mm. Fresh and brackish rivers of central southern New Guinea and far northern Australia. Family Anabantidae . . . Labyrinthfishes Trichogaster pectoralis (Regan) Trichopodus pectoralis Regan, 1910: 784 (Siam). DU (RMNH 28550), 2: 93 and 105 mm; AJ (RMNH 28551), 12: 84-118 mm; AJ (RMNH 28552), 30: 93-114 mm; SL (RMNH 28553), 4: 105-161 mm. An introduced species native to fresh waters of South-East Asia. Family Soleidae . . . Soles Aseraggodes klunzingeri (Weber) Pardachirus klunzingeri Weber, 1908: 250 (Merauke River, southern New Guinea). TM* (RMNH 28554), 104 mm;TM (RMNH 28555), 3: 75-102 mm. Fresh and brackish waters of central southern New Guinea and northern Aus- tralia. 93 Freshwater Fishes from Western New Guinea Brachirus villosa (Weber) Synaptura villosa Weber, 1908: 251 (Wagani River, western New Guinea). MV (RMNH 24695), 13 : 38-57 mm. Fresh and brackish waters of central southern New Guinea. Family Cynoglossidae . . . Tongue Soles Cynoglossus heterolepis Weber Cynoglossus heterolepis Weber, 1910: 237 (Lorentz River, southern New Guinea). TGG (RMNH 28556), 88 mm; TM* (RMNH 28557), 139 mm. Fresh and brackish waters of central southern New Guinea. Discussion The freshwater fishes of New Guinea continue to provide a fertile area for study. Recent collecting activity by Roberts (1978) and that of the senior author between 1977 and 1981 have revealed a wealth of both undescribed and poorly known species. Large tracts remain totally uncollected. For example, there have been no ichthyological explorations on the north coast between the Mamberamo River mouth and the western extremity of the island, an expanse of more than 800 km. Likewise, only a small number of collections have been made on the Vogelkop Peninsula, and in the central highlands of Irian Jaya. Only two previous authors, Weber (1913) and Munro (1964) have assembled comprehensive faunal lists. Both of these lists contain 145 species, but if fishes which are not strictly fresh water dwellers are eliminated, as well as various junior synonyms, the Weber list is reduced to 66 species, and that of Munro to 92 species. On the basis of our research for the present paper and also from consultation with various specialists (particularly for the Ariidae, Plotosidae, Gobiidae, and Eleotridae) we present a list of the freshwater fishes of New Guinea (Appendix Table 1). The list contains 158 species. We include only those fishes which appear to be restricted to fresh- water habitats. We have eliminated various widely distributed forms, which although frequently found in pure fresh water, have a marine stage for dispersal. Therefore we include only species whose distribution is restricted to the New Guinea-northern Australia region. Thus, we have excluded approximately 80 species which regularly penetrate fresh water. A number of the excluded fishes have their main populations in tidal creeks or estuaries, for example the archer- fish Toxotes jaculatrix (Toxotidae), the spotted scat Scatophagus argus (Scato- phagidae), and several species of gobiids and eleotrids. Others such as the eleotrid Ophieleotns aporos^ certain gobiids of the genus Glossogobius, and the eel genus Anguilla may spend the greater part of their life cycle in fresh water, but are evidently dependent on the sea for larval dispersal or as a breeding site. 94 Gerald R. Allen and M. Boeseman The central dividing range of New Guinea includes a number of peaks with elevations in excess of 4 000 m and represents a formidable faunal barrier. Only 11 species of the 158 purely freshwater forms occur on both sides of the central mountains. The southern and northern populations of some of these species will no doubt prove to be specifically distinct when studied in more detail. Kailola (pers. comm.) has recently verified this phenomenon for the ariid catfish Arius leptaspis. Aside from the realtively few shared species the respective fish faunas of the south and north are very distinctive. The southern fauna appears to be the richest with 106 species thus far recorded compared to 61 species from the north. However, part of this difference is no doubt related to the greater amount of collecting activity in the south. Twenty-eight species or approximately 26% of the southern fishes are also found in northern Australia, primarily Arnhem Land and Cape York Peninsula. The latter area was linked to southern New Guinea by a land bridge as recently as 6 500-8 000 years ago (Allen and Hoese 1980). Thus, the faunal similarity of these regions is not surprising. Allen and Hoese (1980) reported that at least 63% of the fishes collected in the Jardine River at the north- ern extremity of Cape York Peninsula are also found in southern New Guinea. Acknowledgements We thank Dr L.D. Brongersma, former Director and Curator of Herpetology of the Rijksmuseum van Natuurlijke Historie, Leiden, and Dr L.B. Holthuis, Curator of Carcinology of the same institution for their invaluable aid during the 1954-55 expedition. This trip was sponsored by the Government of the Netherlands and valuable logistic assistance was rendered by the Government of Netherlands New Guinea, the Royal Netherlands Navy, and missions of the Dutch Reformed Church, Roman Catholic Church, and the Christian and Missionary Alliance. Dr Brongersma was the leader of the 1959 Star Mountains Expedition, at the conclusion of which Dr W. Vervoort (present Director of RMNH) had the oppor- tunity to collect fishes from Tanah Merah. His efforts in this regard are gratefully acknowledged. Dr Vervoort kindly provided working facilities for the senior author during a visit to Leiden in April-May 1981. Gobiid and eleotrid identifications were assisted by Dr D.F. Hoese of the Australian Museum, Sydney and K. Meguro of the Crown Prince’s Palace, Tokyo. Teraponids were identified by Dr G.F. Mees of RMNH. Ariid and plotosid catfish identifications were assisted by Mrs P. Kailola of the University of Adelaide, Australia and Mrs M.N. Feinberg of the American Museum of Natural History, New York. Syngnathids were identified by Dr C.E. Dawson of Gulf Coast Marine Research Laboratory, Mississippi, U.S.A. Finally, we thank Mrs C.J. Allen for her careful preparation of the typescript. 95 Freshwater Fishes from Western New Guinea References Akihito, P. and Meguro, K. (1975). Description of anew gobiid fish, Glossogobius aureus, with notes on related species of the gGmxs. Jpn, J. IchthyoL, 22 (3): 127-142. Allen, G.R. and Cross, N.J. (1980). Descriptions of five new rainbowfishes (Melanotaeniidae) from New Guinea. Rec, West. Aust. Mus. 8 (3): 377-396. Allen, G.R. and Hoese, D.F. (1980). A collection of fishes from the Jardine River, Cape York Peninsula, Australia./. Roy. Soc. West. Aust. 63 (2): 53-61. Berra, T.M., Moore, R. and Reynolds, L.F, (1975). The freshwater fishes of the Laloki River system of New Guinea. Copeia (2): 316-326. Bleeker, P. (1854). Bijdrage tot de kennis der ichthyologische fauna van Halmaheira (Gilolo). Nat. Tijdschr. Ned. Ind. 6: 49-62. Bleeker, P. (1856). Beschrijvingen van nieuwe of weinig bekende vischsoorten van Menado en Makassar grootendeels verzameld op eene reis naar den Molukschen Archipel in het gevolg van den Gouvemeur-Generaal Duymaer van Twist. Act. Soc. Sc. Indo-Neerl. 1: 1-80. Bleeker, P. (1862). Atlas ichthyologique des Indes Orientales Neerlandaises, public sous les auspices du Gouvernement colonial neerlandais. 2. (Amsterdam.) Bloch, M.E. (1785-95). Naturgeschichte der auslzmdischen Fische. 3 vols. (Berlin.) Boeseman, M. (1963), Notes on the fishes of Western New Guinea I. Zool. Meded. 38 (14): 221-242. Boeseman, M. (1964). Notes on the fishes of Western New Guinea III. The freshwater shark of Jamoer Lake. Zool. Meded. 40 (3): 9-22. Broussonet, P.M.A. (1782). Ichthyologia, sistens piscium descriptions et iconis. Decas I. IV: 1-41. (London.) Castelnau, F. (1876). Memoire sur les poissons appeles barramundi par les Aborigenes du nord- est de TAustralie. /. Zool. (Gervais), 5: 129-136. Castelnau, F. (1878). Notes on the fishes of the Norman river. Proc. Linn. Soc. N.S.W. 3: 41-51. Cuvier, G. and Valenciennes, A. (1829-49). Histotre naturelle des poissons. 22 vols. (Paris.) Eggert, B. (1935). Beitrag zur Systemalik, Biologic und geographischen Verbreitung der Peri- ophthalminae. Ergebnisse einer durch die Notgiraeinschaft der Deutschen Wissinschaft ermoglichten Reise nach Neiderlandisch-Indien 1926-27 und der Sundaexpedition der Notgemeinschaft der Deutschen Wissenschaft 1929-30. Zool. Jahrb. Jena Syst. 67: 29-116. Forsskal, P. (1775). Descriptiones animalium avium, amphibiorum, piscium, insectorum, vemium; quae in itinere orientalie observavit. Post mortem auctoris edidit Carsten Niebukr, Havniae, 1-164. Fowler, H.W. (1941). Contributions to the biology of the Philippine Archipelago and adjacent regions. 5u//. U.S. nat. Mus. 13 (100): 1-879. Greenwood, P.H., Rosen, D.E., Weitzman, S.H. and Myers, G.S. (1966). Phyletic studies of teleostean fishes, with a provisional classification of living forms. Bull. Amer. Mus. nat. Hist. 131 (4): 339-456. Gunther, A. (1864). On a new generic type of fishes discovered by the late Dr Leichardt in Queensland. Ann. Mag. nat. Hist. 14: 195-197, Gunther, A. (1866). Catalogue of the fishes in the British Museum. 8 vols. (London.) Gunther, A. (1867). Additions to the knowledge of Australian reptiles and fishes. Ann. Mag. nat. Hist. 3, ser. 20: 45-68, Haines, A.K. (1979). An ecological survey of fish of the Lower Purari River System, Papua New Guinea. Purari River (Wabo) Hydroelectric Scheme. Environmental Studies, Vol. 6. Office of Environment and Conservation &: Dept of Minerals 8c Energy: 1-102 (Port Moresby.) Hamilton, F. Buchanan (1822). An account of the fishes found in the river Ganges and its branches. (Edinburgh and London.) 96 Gerald R. Allen and M. Boeseman Herre, A.W. (1935). New fishes obtained by the Crane Pacific Expedition. Field Mus. nat. Hist, ZooL Ser. 18 (12): 383-348. Kent, W. Saville (1892). Description of a new species of true Barrimundi, Osteoglossum jardinii from North Queensland. Proc. Roy, Soc. Qld 8: 105-108. Koumans, F.P. (1949). Zoological results of the Dutch New Guinea Expedition, 1939. No. II. The Fishes. Nova Guinea, new series, V: 284-288. Koumans, F.P. (1953). In: Weber and de Beaufort. The fishes of the Indo-Australian Archi- pelago. 10. (E.J. Brill: Leiden.) Lake,J.S. [191%), Australian freshwater fishes. Nelson: Melbourne, Latham, J. (1794). An essay on the various species of sawfish. Trans. Linn. Soc. London 2: 273-282. Linnaeus, C. (1766). Systema naturae, Ed. 12. Salvii, Holmiae. Macleay, W. (1883). Contribution to a knowledge of the fishes of New Guinea. Part 4. Proc. Linn. Soc. N.S.W. 8 (2): 252-280. Mees, G.F. (1971). Revisional notes on some species of the genus Therapon (Pisces, Theraponi- dae). Zool. Meded. 45: 197-224. Mees, G.F. and Kailola, P.J. (1977). The freshwater Therapontidae of New Guinea. Zool. Verh. 153: 1-89. Muller, J. and Henle, J. (1838-41). Systematische Beschreibung der Plagiostomen. (Verlag; Berlin.) Munro, I.S.R. (1956). Handbook of Australian fishes. No. 6. Fish. NewsL, Dept Primary Industry, Canberra. 15 (12): 15-18. Munro, I.S.R. (1964). Additions to the fish fauna of New Guinea. Papua New Guinea Agricul. J. 16 (4): 141-186. Munro. I.S.R. (1967). Fishes of New Gumea. Dept Agricul. Stock and Fish. (Port Moresby.) Nelson, G. and Rothman, M.N. (1973). The species of gizzard shads (Dorosomatinae) with particular reference to the Indo-Pacific region. Bull. Amer. Mus. nat. Hist. 150 (2); 131-206. Nichols, J.T. (1937). Results of the Archbold Expeditions, No. 15, A new fish ot the genus Bostrychus from New Guinea, ^mer. Mus. Novit. 922: 1-2. Nichols, J.T, and Raven, H.C. (1934). Two new fresh-water fishes (Percesoces) from New Guinea. Amer. Mus. Novit. 755: 1-4. Pallas, P.S. (1767). Spicilegia zoologica, quibus novae imprimis et obscurae animalium species iconibus, descriptionibus atque commentarius illustrantur. 2 vols., 1767-80. Perugia, A. (1894). Viaggio di Lamberto Loria nella Papuasia orientale xiii, Pisci d’acqua dolce. Ann. Mus. Civico Genova (2) 14: 546-553. Ramsay, E.P. and Ogilby, J.D, (1887). A contribution to the knowledge of the fish fauna of New Guinea. Proc. Linn. Soc, N.S.W. , (2) 1(1): 8-20. Regan, C.T, (1910). The Asiatic fishes of the family Anabantidae. Proc. zool. Soc. London: 767-787. Richardson, J. (1842). Contributions to the ichthyology of Australia. Ann. Mag. nat. Hist. 9: 15-31; 384-393. Richardson, J. (1844). Zoology. Voyage 'Erebus* and 'Terror*. (London.) Roberts, T.S. (1978), An ichthyological survey of the Fly River in Papua New Guinea with descriptions of new species. Smith. Contrib. zool. 281 : 1-72. Smith, A. (1849). Illustrations of the zoology of south Africa; consisting chiefly of figures and descriptions of the objects of natural history collected during an expedition into the interior of south Africa in 1834-36. 5 vols. (London.) Steindachner, F. (1867). Ueber einige Fische aus dem Fitzroy-Flusse bei Rockhampton in Ost- Australein. Sitzber. Akad. Wien, 55: 9-16. Steindachner, F. (1880). Ichthyologische Beitrage \l\\. Sitzber. Akad. Wm. Wien, 80, 1: 119-191. 97 Freshwater Fishes from Western New Guinea Vari, R.P. (1878). The Terapon perches (Percoidei; Teraponidae), a cladistic analysis and taxonomic revision. Bull. Am. Mus. nat. Hist. 159: 175-340, Weber, M. (1908). Susswasserfische von New-Guinea. Nova Guinea, V (2): 201-267. Weber, M. (1910). Neue Fische aus niederlandisch Sud-Neu-Guinea. Notes Leyden Mus. 32: 225-240. Weber, M. (1913). Susswasserfische aus Niederlandisch Sud-und Nord-New-Guinea. Nova Guinea. IX (4): 513-613. Weber, M. and de Beaufort, L.F. (1913). The fishes of the Indo-Australian Archipelago, Vol. 2. (E.J. Brill: Leiden.) Appendix Table List of the Freshwater Fishes of New Guinea Species Known distribution (SNG = S New Guinea NNG = N New Guinea) Family Clupeidae (3 spp.) 1 Clupeoides papuensis (Ramsay and Ogilby) 2 C. venulosus Weber and de Beaufort 3 Nematalosa erebi (Gunther) Digul and Fly Rivers (SNG) Lorentz and Fly Rivers (SNG) Jamur Lake, Digul and Fly Rivers (SNG); N Australia Family Engraulidae (2 spp.) 4 Thryssa rastrosa Roberts Fly River (SNG) 5 T. scratchleyi (Ramsay and Ogilby) Digul and Fly Rivers (SNG); N Australia Family Osteoglossidae (1 sp.) 6 Scleropages jardinii Gunther Family Ariidae (27 spp.) 7 Arius acrocephalus (Weber) 8 A. augustus Roberts 9 A. berneyi (Whitley) 10 A carinatus (Weber) 11 A. kanganamanensis (Herre) 12 A. latirostris (Macleay) Digul and Fly Rivers (SNG); N Australia Digul, Fly, Purari and Laloki Rivers (SNG) Fly River (SNG) Fly River (SNG); N Australia Lorentz, Digul, Fly and Purari Rivers (SNG) Ramu and Sepik Rivers (NNG) Lorentz, Purari and Laloki Rivers (SNG) 98 Gerald R. Allen and M. Boeseman 13 A. leptaspis Bleeker Widespread SNG and N Australia 14 A. solidus (Herre) Ramu and Sepik Rivers (NNG) 15 A. stirlingi (Ogilby) Widespread SNG and N Australia 16 A. sp. A Jamur Lake (SNG) 17 A. sp. B Digul River (SNG) 18 A. sp. C Widely distributed in NNG 19 A. sp. D Sentani Lake, Sepik and Ramu Rivers (NNG) 20 Brustiarius nox (Herre) Ramu and Sepik Rivers (NNG) 21 Cinetodus froggatti (Ramsay and Ogilby) Digul, Merauke, Fly, Kikori and Purari Rivers (SNG); N Australia 22 Cochlefelis danielsi (Regan) Lorentz, Fly and Kikori Rivers (SNG) 23 C. spatula (Ramsay and Ogilby) Digul and Fly Rivers (SNG) 24 Doiichthys novaeguineae Weber Lorentz River (SNG) 25 Hemipimelodus bernhardi Nichols Mamberamo River (NNG) 26 H. crassilabrus Ramsay and Ogilby Fly and Purari Rivers (SNG) 27 H. macrorhynchus Weber Lorentz, Digul, Fly and Purari Rivers (SNG) 28 H. papillifer Herre Ramu and Sepik Rivers (NNG) 29 Nedystoma dayi (Ramsay and Ogilby) Lorentz, Digul, Fly and Purari Rivers (SNG) 30 Netuvna microstoma (Nichols) Mamberamo River (NNG) 31 Tachysurus kanganamanensis (Herre) Ramu and Sepik Rivers (NNG) 32 T. solidus (Herre) Ramu and Sepik Rivers (NNG) 33 Tetranesodon conorhynchus Weber Lorentz River (SNG) Family Plotosidae (10 spp.) 34 Neosilurus ater ater (Perugia) Widespread SNG and NNG;N Australia 35 N. brevidorsalis (Gunther) Widespread SNG and N Australia 36 N. equinus (Weber) Widespread SNG and NNG 37 N. idenburgi (Nichols) Mamberamo, Sepik, Ramu and Markham Rivers (NNG) 38 N. meraukensis (Weber) Merauke, Fly and Nami Nami Rivers (SNG) 39 N. novaeguineae (Weber) Lake Sentani (NNG) 40 Oloplotosus luteus Gomon and Roberts Fly River (SNG) 41 0. marine Weber Lorentz River (SNG) 42 Plotosus papuensis Weber Lorentz and Fly Rivers (SNG) 43 Porochilus obbesi (Weber) Lorentz, Oriomo and Laloki Rivers (SNG); N Australia Family Anguillidae (1 sp.) 44 Anguilla interioris Whitley Widespread SNG and NNG Family Beloni Family Belonidae (2 spp.) 45 Strongylura kreftti (Gunther) Widespread SNG and N Australia 46 S. perornatus Whitley Sepik River (NNG) 99 Freshwater Fishes from Western New Guinea Family Hemirhamphidae (5 spp.) 47 Zenarchopterus alleni Collette Mamberamo River (NNG) 48 Z. caudovittatus (Weber) Merauke River (SNG) 49 Z. kampeni (Weber) Mamberamo, Sepik and Ramu Rivers 50 Z. novaeguineae (Weber) (NNG) Lorentz, Oriomo, Fly, Purari and 51 Z. robertsi Collette Laloki Rivers (SNG) Kumusi River (NNG) Family 52 Melanotaeniidae (41 spp.) Chilatherina axelrodi Allen Pual River (NNG) 53 C. campsi (Whitley) Upper Sepik and Purari systems, and 54 C. crassispinosa (Weber) Markham River (SNG and NNG) Mamberamo River to Markham River 55 C. fasciata Regan (NNG) Mamberamo River to Markham River 56 C. lorentzi (Weber) (NNG) Mamberamo River to Vanimo vicinity 57 C. sentaniensis (Weber) (NNG) Lake Sentani and Sekanto River (NNG) 58 Glossolepis incisus Weber Lake Sentani (NNG) 59 G. maculosus Allen Omsis River (NNG) 60 G. multisquamatus (Weber) Mamberamo and Sepik Rivers (NNG) 61 G. pseudoincisus Allen and Cross Tami River (NNG) 62 G. wanamensis Allen and Kailola Lake Wanam (NNG) 63 Triatherina werneri Meinken Merauke River to Fly River (SNG); 64 Melanotaenia affinis (Weber) N Australia Sermowai River to Markham River 65 M. ajamaruensis Allen and Cross (NNG) Ajamaru Lakes (SNG) 66 M. hoesemani Alien and Cross Ajamaru Lakes (SNG) 67 M. catherinae de Beaufort Waigeo Island (NNG) 68 M. corona Allen Sermowai River (NNG) 69 M. goldiei (Macleay) Widespread SNG and Aru Islands 70 M. herbertaxelrodi Allen Lake Tebera (SNG) 71 M. japenensis Allen and Cross Japen Island (NNG) 72 M. lacustris Munro Lake Kutubu (SNG) 73 M. maccullochi Ogilby Bensbach River to Fly River (SNG); 74 M. misoolensis Allen N Australia Misool Island (SNG) 75 M. monticola Allen Upper Purari system (SNG) 76 M. ogzlbyi Weber Lorentz River (SNG) 77 M. oktediensis Allen and Cross Oktedi River (SNG) 78 M. papuae Allen Port Moresby vicinity (SNG) 79 M. parkinsoni Allen Port Moresby to Alotau (SNG) 80 M. pimaensis Allen Pima River (SNG) 81 M. praecox Weber and de Beaufort Mamberamo River (NNG) 100 Gerald R. Allen and M. Boeseman 82 M, splendida ruhrostriata (Ramsay and Ogilby) 83 M. sexlineata (Munro) 84 M, vanheumi (Weber and de Beaufort) 85 Popondetta connieae Allen 86 P, furcatus (Nichols) 87 Pseudomugil gertrudae Weber 88 P. inconspicuus Roberts 89 P. novaeguineae Weber 90 P. paludicola Allen and Moore 91 P, sp. no. 1 92 P. sp. no. 2 Family Atherinidae (4 spp.) 93 Craterocephalus lacustris Trewavas 94 C. nouhuysi (Weber) 95 C. randi Nichols 96 C. sp. Family Ambassidae (12 spp.) 97 Ambassis agrammus Gunther 98 A, macleayi (Castelnau) 99 A. reticulata Weber 100 Denariusa bandata Whitley 101 Parambassis altipinnis Allen 102 P. confinis (Weber) 103 P, gulliveri (Castelnau) 104 Synechopterus caudovittatus Norman 105 Tetracentron apogonoides Macleay 106 Xenamhassis honessi Schultz 107 X. lalokiensis Munro 108 X. simoni Schultz Family Lobotidae (1 sp.) 109 Dantinoides campbelli Whitley Family Teraponidae (11 spp.) 110 Amniataba affinis (Mees and Kailola) 111 Hephaestus adamsoni (Trewavas) Digul River to Purari River and Aru Islands (SNG) Fly River (SNG) Mamberamo River (NNG) Vicinity of Popondetta (NNG) Musa River and Wanagela vicinity (NNG) Digul River to Fly River and Aru Islands (SNG); N Australia Fly River (SNG) Etna Bay to Fly River and Aru Islands (SNG) Morehead River to Binaturi River (SNG) Misool Island (SNG) Cape Ward Hunt (NNG) Lake Kutubu (SNG) Lorentz River (SNG) Jamur Lake to Balimo (SNG) Upper Purari system (SNG) Bensbach River to Fly River (SNG); N Australia Digul River to Balimo (SNG); N Australia Lake Jamur to Merauke River (SNG) Bensbach River to Fly River (SNG); N Australia Mamberamo River (NNG) Mamberamo and Sepik Rivers (NNG) Lorentz River to Purari River (SNG); N Australia Kokoda vicinity (NNG) Laloki and Kemp Welsh Rivers (SNG) Buna vicinity (NNG) Laloki River (SNG) Buna vicinity (NNG) Sepik River (NNG) Morehead and Fly Rivers (SNG) Lake Kutubu (SNG) 101 Freshwater Fishes from Western New Guinea 112 H. fuliginosus (Macleay) 113 H. ohtusifrons (Mees and Kailola) 114 H, raymondi (Mees and Kailola) 115 H. roemeri (Weber) 116 H. transmontanous (Mees and Kailola) 117 H. trimaculatus (Macleay) 118 Pingalla lorentzi (Weber) 119 Terapon jamoerensis (Mees) 120 T. lacustris (Mees and Kailola) Family Apogonidae (8 spp.) 121 Glossamia aprion (Richardson) 122 G. heauforti (Weber) 123 G. gjellerupi (Weber and de Beaufort) 124 G. heumi (Weber and de Beaufort) 125 G. narindica Roberts 126 G. sandei (Weber) 127 G. trifasciata (Weber) 128 G. wichmanni (Weber) Family Lutjanidae (1 sp.) 129 Lutjanus goldiei (Macleay) Family Toxotidae (2 spp.) 130 Toxotes lorentzi Weber 131 T. oligolepis Bleeker Family Gobiidae (8 spp.) 132 Acentrogobius bulmeri (Whitley) 133 Aloricatogobius asaro (Whitley) 134 Ctenogobius tigrellus (Nichols) 135 Glossogobius brunnoides (Nichols) 136 G. concavifrons (Ramsay and Ogilby) 137 G, hoesei Allen and Boeseman 138 G. koragensis Herre 139 Mugilo go bins f us cuius (Nichols) Family Eleotridae (16 spp.) 140 Bostrychus strigogenys (Nichols) 141 Hypseleotris moncktoni (Regan) Fly and Purari Rivers (SNG); N Australia Mamberamo and Sermowai Rivers (NNG) Morehead River (SNG) Lorentz and Digul Rivers (SNG) Sepik and Ramu Rivers (NNG) Mimika River to Laloki River (SNG) Lorentz, Digul, Morehead and Fly Rivers (SNG); N Australia Lake Jamur (SNG) Morehead River to Balimo (SNG) Oriomo River to Balimo (SNG) Mamberamo River, Lake Sentani (NNG) Mamberamo and Sepik Rivers (NNG) Mamberamo River (NNG) Fly River (SNG) Wagami River to Purari River (SNG) Lorentz and Fly Rivers (SNG) Tawarin River to Markham River (NNG) Fly, Purari and Laloki Rivers (SNG) Merauke River and Balimo vicinity (SNG); N Australia Jamur Lake (SNG); Molucca Islands and N Australia Upper Sepik system (NNG) Upper Purari system (SNG) Mamberamo River (NNG) Upper Kikori and Purari systems (SNG) Fly River (SNG) ; N Australia Ajamaru Lakes (SNG) Lake Sentani and Sepik River (NNG) Distribution unknown Digul River to Balimo (SNG) Agarambo vicinity (SNG) 102 Gerald R. Allen and M. Boeseman 142 Mogumda mogurnda (Richardson) Widespread SNG and NNG;N Australia 143 M. variegata Nichols Lake Kutubu (SNG) 144 M. sp. A Widespread SNG and NNG 145 M. sp. B Kemp Welsh River (SNG) 146 M. sp. C Bulolo River (NNG) 147 Odonteleotris nesolepis (Weber) Widespread NNG 148 Oxyeleotris fimbriata (Weber) Widespread SNG and NNG 149 O. herwerdeni (Weber) Widespread SNG and NNG;N Australia 150 O. lineolatus (Steindachner) Widespread SNG and NNG;N Australia 151 O. novaeguineae Koumans Widespread SNG and NNG 152 O. nullipora Roberts Digul and Fly Rivers (SNG) 153 O. paucipora Roberts Digul and Fly Rivers (SNG) 154 O. wisselensis Allen and Boeseman Wissel Lakes (SNG) 155 Tateumdina ocellicauda Nichols Popondetta vicinity and Musa River (NNG) Family Soleidae (2 spp.) 156 Aseraggodes klunzingeri (Weber) Widespread SNG; N Australia 157 Brachirus villosus (Weber) Wagani River to Fly River (SNG) Family Cynoglossidae 158 Cynoglossus heterolepis Weber Oetoemboewe, Lorentz, Digul and Fly Rivers (SNG); N Australia Received 9 March 1982 Accepted 17 May 1982 Published 13 December 1982 103 Rec. West. Aust. Mus. 1982 , 10 ( 2 ): 105-109 A New Species of Freshwater Rainbowfish (Melanotaeniidae) from Misool Island, Indonesia Gerald R. Allen* Abstract A new species of rainbowfish (Melanotaeniidae) belonging to the genus Melano- taenia is described from 23 specimens collected at Misool Island off the western extremity of Irian Jaya, Indonesia. Melanotaenia misoolensis sp. nov. is closely related to M. catherinae from Waigeo Island, but differs with regard to colour pattern and by having a greater number of soft anal rays. Introduction The freshwater rainbowfishes of the family Melanotaeniidae are small (usually under 12 cm) inhabitants of streams, lakes, and swamps in the Australian-New Guinea region. The group contains approximately 50 species which are assigned to eight genera (see Allen 1980). Melanotaenia is by far the largest genus with 25 known species, of which 21 are found in New Guinea (Allen and Cross 1982). This large island still remains relatively unexplored, and therefore can be expected to yield additional new species, particularly the poorly known western half (Irian Jaya). The present paper describes a new Melanotaenia which was located amongst unstudied New Guinea material at the Zoological Museum of the University of Amsterdam in the Netherlands. Melanotaenia misoolensis sp. nov. is described on the basis of 23 specimens collected at Misool Island in 1948. Misool is a relatively large (approximately 90 x 38 km) island lying just to the south of the western extremity of Irian jaya and separated from the mainland by a distance of 32 kilometres. The island has a maximum elevation of 990 m. Rainbowfishes of the genus Melanotaenia have been reported from several other islands off the coast of Irian Jaya. Melanotaenia catherinae (Beaufort) from Waigeo and M. japenensis Allen and Cross from Japen are endemic to these islands which are situated off the north coast. The Aru Islands off the south coast are inhabited by M. goldiei (Macleay) and M. splendida rubrostriata (Ramsay and Ogilby), both of which are widely distributed on the southern New Guinea mainland. All of these insular areas were formerly connected to the New Guinea land mass and are presently separated by shallow (less than 50 fathoms) seas. * Department of Ichthyology, Western Australian Museum, Francis Street, Perth, Western Australia 6000. 105 A New Freshwater Rainbowfish (Melanotaeniidae) from Indonesia Methods of counting and measuring follow those explained in Allen and Cross (1980). Counts and measurements are summarized in Tables 1 and 2. Data in parentheses indicate the range for paratypes when differing from the holotype. Proportional measurements are presented as percentage of the standard length. These data are based on the holotype and 12 paratypes, 42.0-56.8 mm SL. Type specimens are deposited at the National Museum of Natural History, Washington, D.C. (USNM), the Western Australian Museum, Perth (WAM) and the Zoologisch Museum, Amsterdam (ZMA). Table 1 Proportional measurements expressed in thousandths of the standard length for selected type specimens of Melanotaenia misoolensis sp. nov. Character Holotype ZMA 116.456 Paratypes ZMA 116.457 Standard length (mm) 56.8 58.5 52.9 51.0 50.0 Depth 335 345 342 350 336 Width 136 144 134 137 144 Head length 278 277 297 290 288 Snout to first dorsal fin origin 481 479 541 527 484 Snout to anal fin origin 502 509 493 480 504 Snout to pelvic fin origin 387 383 346 363 384 Length of second dorsal fin base 246 260 227 224 246 Length of anal fin base 396 421 384 363 404 Snout length 88 89 87 84 86 Orbit diameter 92 93 96 98 100 Bony interorbital width 106 104 112 102 104 Depth of caudal peduncle 109 115 121 122 112 Length of caudal peduncle 141 145 174 176 144 Length of pectoral fin 202 203 208 224 206 Length of pelvic fin 167 162 197 188 170 Longest ray of first dorsal fin 165 156 123 147 150 Longest ray of second dorsal fin 143 137 127 143 128 Longest anal ray 132 120 155 139 140 Length of caudal fin 218 239 246 231 270 Table 2 Fin ray counts for type specimens of Melanotaenia misoolensis. First dorsal fin spines Second dorsal fin soft rays IV V VI 12 13 14 2 21 2 10 13 2 Anal soft fin rays Pectortd fin rays 20 21 22 23 24 25 13 14 15 16 1 1 5 7 6 5 8 13 3 1 106 Gerald R. Allen Systematics Melanotaenia misoolensis sp. nov. Figure 1 Holotype ZMA 116.456 , male 58.5 mm SL, tributary of Wai Tama at Fakal, Misool Island, Indonesia (approximately 2®00'S, ISO^OO'E), M.A. Lieftinck, 2 October 1948. Paratypes USNM 227492, 3 specimens, 30.6-39.5 mm SL, collected with holotype; WAM P27279-001, 2 specimens, 52.9-58.5 mm SL, collected with holotype; AMZ 116.457, 17 specimens, 21.9- 52.7 mm SL, collected with holotype. Figure 1 Melanotaenia misoolensis, male, paratype, 58.5 mm SL. Diagnosis A species of Melanotaenia with the following combination of characters: dorsal rays IV to VM,12 to 14; anal rays 1,20 to 25; pectoral rays 13 to 16; horizontal scale rows 10; vertical scale rows 33 or 36; preopercle-suborbital scales 1 1 to 15; predorsal scales 13 to 17; colour in preservative generally pale brown on back and yellowish-tan below; a dark stripe running along middle of sides from rear edge of eye to caudal fin base, maximum width of stripe about IV 2 scales. Description Dorsal rays V-I,12 (IV to VI-1,12 to 14); anal rays 1,23 (1,20 to 25); pectoral rays 13 (13 to 16), horizontal scale rows 10; vertical scale rows 34 (33 to 36); predorsal scales 16 (13 to 17); preopercle-suborbital scales 13 (11 to 15); gill rakers on first arch 2 + 16 (2 or 3 + 14 to 16). 107 A New Freshwater Rainbowfish (Melanotaeniidae) from Indonesia Greatest body depth 34.5 (32.6 to 35.0); maximum body width 14.4 (13.6 to 14.9); head length 27.7 (27.0 to 29.8); snout length 8.9 (8.4 to 9.0); eye diameter 9.3 (9.2 to 11.0), interorbital width 10.4 (10.2 to 11.2); caudal peduncle depth 11.5 (10.9 to 12.2); caudal peduncle length 14.5 (14.1 to 17.6); pectoral fin length 20.3 (20.2 to 22.4); pelvic fin length 16.2 (16.7 to 19.7); caudal fin length 23.9 (21.0 to 27.0); predorsal distance 47.9 (48.1 to 54.1); preanal distance 50.9 (45.7 to 51.0); prepelvic distance 38.3 (34.6 to 38.7). Body ovate, laterally compressed, the snout somewhat pointed. Predorsal profile straight, the interorbital and adjacent nape flattened. Ventral, prepelvic profile rounded, the breast strongly compressed at ventral midline. Jaws oblique, approximately equal; premaxilla with an abrupt bend between the anterior horizontal portion and lateral part; rear edge of maxilla about level with front of eye; lips thin; both jaws with dense covering of teeth arranged in irregular rows; teeth conical with slightly curved tips; teeth on anterior and lateral portions of premaxilla invading lips and distinctly visible when mouth is closed; exposed teeth also visible at front of lower jaw; vomer with narrow band of villi- form teeth in 1 or 2 rows; palatines with similar teeth arranged in a single row. Scales relatively large, arranged in regular horizontal rows; scales with smooth to scalloped margins; predorsal scales extending to rear of interorbital; preopercle- suborbital scales in two rows. First dorsal fin originates about level with or slightly behind anal fin origin; first spine 2-3 times thickness of other spines of first dorsal fin; third spine the longest, its tip reaching base of about third soft ray of second dorsal fin when depressed. Last 2 or 3 soft rays of second dorsal fin the longest in males, anterior rays the longest in females; depressed tip of second dorsal fin extending nearly to caudal fin base in adult males and about half to two-thirds length of caudal peduncle of females. Anal spine slightly shorter than first dorsal spine which is about half head length; longest rays of anal fin in posterior part of fin in males and anterior portion in females; dorsal and anal fins with rectangular outline, pointed posteriorly with elongated rays in males; pectoral fins pointed; pelvic fin tips when depressed extending to about base of second or thud soft anal ray; caudal fin moderately forked. Colour in alcohol: light brown on upper back, yellowish-tan on ventral half; a brown stripe, slightly more than one scale wide at its broadest point, extending along middle of side from rear edge of eye to base of caudal fin; fins primarily tan with some dusky brown pigmentation; a browmish spot frequently present at base of upper pectoral rays. Live coloration is unknown. Comparisons Melanotaenia misoolensis is most closely allied to M. catherinae (Beaufort 1910), which is endemic to Waigeo, a large island lying approximately 160 km north of Misool. Both species are similar in colour; however, the mid-lateral stripe ol M, catherinae is significantly wider, having a maximum width of about 108 Gerald R. Allen three scales compared with IV 2 scales for M. misoolensis. Moreover, the mid- lateral stripe of M. misoolensis is nearly covered entirely by the pectoral fin, whereas it is broadly exposed (at least one scale row) above the pectoral fin of M. catherinae. In addition, the latter species lacks the dusky spot on the fin membrane behind the last dorsal spine and has a dusky soft dorsal fin which is often blackish in adult males. Melanotaenia misoolensis, in contrast, has a dusky spot behind the last dorsal spine and the soft dorsal fin is pale, possibly yellowish in life. The only meristic difference noted is related to counts for the soft anal rays. Melanotaenia misoolensis usually has 22 to 25 rays (one specimen with 20 rays) compared with 19 to 21 rays for M. catherinae. Comparative material included 35 specimens (ZMA 103.145, paralectotypes), 37-72 mm SL, of M. catherinae. Remarks The species is named misoolensis with reference to the type locality. Acknowledgements I thank Dr Han Nijssen, Curator of Fishes at ZMA for generously providing laboratory facilities and access to previously unstudied New Guinea fishes during a visit to Amsterdam in April 1981. Connie Allen prepared the typescript. References Allen, G.R. (1980). A generic classification of the rainbowfishes (Melanotaeniidae). Rec. West Aust. Mus. 8 (3): 449-490. Allen, G.R. and Cross, N.J. Rainbowfishes of the World. New Jersey (U.S.A.): T.F.H. Publica- tions. (In press.) Beaufort, L.F. de (1910). Weitere Bestatigiing einer zoogeographischen Prophezeiung. ZooL Anz., 36: 249-252. Received 5 February 1982 Accepted 17 May 1982 Published 13 December 1982 109 Rec. West Aust Mus. 1982 , 10 ( 2 ): 111-126 Walled Rock Shelters and a Cached Spear in the Pilbara Region, Western Australia P. Bindon* and M. Lofgrenf Abstract In late 1977, our attention was drawn to walled rock shelters in the Pilbara Region of Western Australia where a spear was discovered behind an intact wall section. Although the age of the spear is not firmly established, we suggest that it is of post- contact origin. This paper considers walled rock shelters and their relationship to various other Aborigin^ stone structures reported throughout Australia. We also describe the spear, and compare it to other pieces in the ethnographic collection of the Western Australian Museum. Finally, we suggest a possible use for walled rock shelters. Background In November 1977, Goldsworthy Mining Ltd personnel advised the Department of Aboriginal Sites, Western Australian Museum, that a spear and other pieces of wood had been discovered behind a man-made wall in a rock shelter on their mining lease. Subsequent investigation of the site led to the recovery of the spear and prompted search for further examples of stone walling. Physical Setting The vast accumulation of iron rich rocks is perhaps the best known feature of the Pilbara Region (Figure 1). At several locations these rocks are currently mined for export (Trendall 1974). One temporary mining reserve (Goldsworthy’s Area ‘C’, also known as Packsaddle) straddles a low range (c. 800 m elevation above sea level) where, following uplift, erosion of cracks and joint planes has resulted in the formation of fissures, sink holes, tunnels, and small caverns. In section these latter features are usually plano-convex or flattened ovals, with the long axis aligned horizontally. Most of the gullies and steeper sided gorges contain several of these features. * Department of Archaeology, Western Australian Museum, Francis Street, Perth, Western Australia 6000. f Department of Anthropology, Western Australian Museum, Francis Street, Perth, Western Australia 6000. Ill Walled Rock Shelters and a Cached Spear in Western Australia Tropical cyclones dominate the Pilbara climate with an average of two per year being recorded. They occur between December and March and result in about half the annual rainfall (average c. 250 mm). Summer thunderstorms and infrequent winter rains make up the remainder. Marble Bar, 225 km north of the study area and the closest settlement with adequate records, averages just less than five falls of rain greater than 20 mm per year. With daily high annual mean temperatures of 35.7°C, the humidity is correspondingly very low (Beard 1975). This semi-arid climate played a considerable part in the preservation of the spear discussed in this paper. Within the research area, the vegetation can be characterized as tree steppe, dominated by the Eucalyptus brevifolia-Triodia wiseana association (Beard 1980). The lower slopes and the gullies contain riverine woodland flora, including Eucalyptus camaldulensis, Acacia spp., Grevillea spp., Hakea spp., etc., while valley plain habitats carry mulga formations (Acacia aneura low woodland). Previous Research A number of studies have focused upon the rock art of the Pilbara (Dix 1977; McCarthy 1962; McCaskill 1977; Virili 1977; Worms 1954; Wright 1968, 1977) and Palmer (1975, 1977fl!, b, c) has related aspects of rock art to ethnographic site data. In light of the discussion to follow, it is worth noting that spears are depicted in the rock art of the region (e.g. Wright 1968: 43, 62). Studies into other aspects of the prehistory of the Pilbara Region include reports of archaeological sites in the Chichester Ranges (Dortch 1972), the Tom Price area (Bednarik 1977), and the Millstream area (summarized and reported in Clarke et al. 1978). These studies lack occupation dates and none mention the wooden artefacts of the region. Brown (1980) has analyzed stone artefact assemblages along a transect from coast to inland through the area, but once again without the benefit of dating. In addition, numerous unpublished surveys (Department of Aboriginal Sites, Western Australian Museum) further document the extent of prehistoric Aboriginal occupation in specific localities throughout the area. Maynard (1980) published the first radiometric date of 20,740 ± 345 BP (SUA 1041) for an inland site in the Pilbara (P0187, a rock shelter near Newman - Figure 1). Tindale’s (1974) map of tribal boundaries in Aboriginal Australia indicates that the site is within the territory of the Pandjima people. Population figures for this tribe are not available, but numbers around 500 are probably realistic given the semi-arid nature of this area (Radcliffe-Brown 1930: 688). Early ethnographic accounts (e.g. Withnell 1901) make no mention of walled rock shelters or spears of the type mentioned in this paper. Clement and Schmelz (1903) published ethnographical notes which include the first report of a spear of the type we describe in detail. 112 PORT HEDLAND P. Bindon and M. Lofgren 113 Figure 1 The Pilbara Region, Western Australia. Walled Rock Shelters and a Cached Spear in Western Australia The Site Located 105 km south-east of Wittenoom and 85 km north-west of Mt Newman, the rock shelter (P4349 — Figure 1) lies in a low ridge trending roughly east/west (22^59'S, 118^49'E). On this ridge, and on the adjacent plain, there are a number of other Aboriginal sites including artefact scatters, stone arrangements, and several other rock shelters, some containing dry-stone walling. The shelter described in this report faces south into a shallow but steep-sided gully (Figure 2). Other fissures, cracks and small shelters in the same strata also open into this gully. Except for stone walls and pieces of wood, none of them contain cultural material and none is large enough to provide shelter for humans. B Figure 2 Sketch of floor plan and cross-section of walled rock shelter (P4349) in which the spear (A) was found. 114 P. Bindon and M. Lofgren Inside the rock shelter, all the passages at the rear and sides have been walled (Figure 3). These walls consist of flattish slabs (to 30 cm^) of local rock, laid one slab wide, one on top of another, with random jointing patterns. Except where now broken down, the walls fill the whole space between floor and ceiling, and extend across the passages. It appears that these walls were not intended to completely seal the area enclosed behind them. In three cases, a naturally hollowed section of tree-trunk about 50 cm long and 10 cm internal diameter has been placed horizontally by the builder so as to penetrate the wall. The other- wise complete walls have one or two slabs omitted leaving an opening about 10 or 15 cm in diameter. Figure 3 Stone walling techniques inside the rock shelter. Behind the wall the floor is 40 cm higher than in the open part of the shelter. This enclosed floor is covered with fist-sized rocks fallen from the walls and ceiling of the shelter. It is littered with twigs and spinifex [Triodia sp.) leaf blades. The spear described later in this paper was lying on this surface when discovered. Some of this debris has entered the shelter through a small sink hole now choked with rubble. The walled section is too small to enter, and can only be examined from a prone position. In a review of Aboriginal stone structures, Mountford (1940) proposed two major categories: cairn-like structures and piles of pebbles (pebble mounds). Withnell (1901: 5) mentions piles of stone with ceremonial functions. Neither of 115 Walled Rock Shelters and a Cached Spear in Western Australia these two authors discuss walling. Kimber (1981) discusses a number of stone features including walling, but not within caves. However, stone-wall fish-traps are well known in Western Australia (e.g. Love 1936: 138), and Worsnop (1897: 108) describes walls across small gullies in the Kimberley. Stone circles, made by clearing a rocky area of stones which are then used to make a wall, have been found in the Pilbara, in other areas of Western Australia and in other places in Australia (e.g. McCaskill 1977: 184). Love (1936: 176) mentions small enclosures of stone in which dingo pups were confined. These walls, and other clearings found on rocky hillsides may have been used as hunting hides. They are in the open, not within rock shelters. Commonly in the Kimberley and more rarely elsewhere, secondary burial chambers were constructed by wedging a few stone slabs across a suitable small crevice. Sacred material was sometimes stored in walled niches and these places marked in one way or another to warn the uninitiated not to endanger themselves by approaching too closely. The complete lack of human bone fragments or appropriate cultural material together with the absence of any of the usual markers, rules out the use of the walled shelters described here as either burial chambers or repositories. The Spear The spear (Figure 4) found in the walled niche, is now registered in the W.A. Museum collection as A23064. It is of composite form, 2.73 m in total length, with a single barb cai-ved from the solid on the point section. This point (32 cm in length) appears to be made from Acacia sp., while the shaft (2.41 m in length) is a less dense wood of unknown genus. The spear point is affixed to the shaft with gum and sinew, and the shaft has been treated with fat and ochre at some time in the past. A crack in the main shaft has been repaired in the Museum, but no attempt has been made to straighten the curvature which has occurred due to warping. The spear weighs c. 300 g, and its point of balance is located 0.44 of its total length from the tip of the spear head. Scorch marks appear on the shaft, suggesting the usual practice of heating the wood during initial construction for the purpose of straightening the shaft and to assist bark removal (e.g. Hayden 1979: 75). There is also a small scorch mark at the barb suggesting that the barb’s angle of deviation from the main body of the spear head (c. 25*^) may have been adjusted by heating. The length of the barb is 42 mm and the vertical distance from the point of the barb to the adjacent spear shaft is 15 mm. Recently the head of a spear of this type, detached from its shaft, was found in a Pilbara rock shelter and donated to this Museum (WAM A23494, Figure 5). Although the tip and barb area is mostly lost, enough remains to indicate that again the barb was carved from the solid near a knot as in the case of the recently discovered specimen. This spear head has a series of about 30 oblique cuts running 116 P. Bindon and M. Lofgren 1 | 117 Figure 4 A23064 — the spear discovered in the rock shelter. Length - 2.73 m. ///j '/n ' ■/ // /{ Walled Rock Shelters and a Cached Spear in Western Australia 118 Figure 3 A23494 — head of a similar spear. Length = 24.8 P. Bindon and M. Lofgren along the barb side for 16 cm. Near the barb another group of 14 similarly shaped cuts join the longer series to produce a group of ‘V’ shapes. Both series of cuts have been produced with a thin metal blade. On the other side of this artefact is a further series of about 35 somewhat indistinct cuts three millimetres long and less than one millimetre wide. The ends and edges of these have been partly smoothed during the manufacturing process or perhaps at a later stage. Similar cuts appear on the shaft of another spear of this type in the WAM collection (A343, Figure 6) but not on other points which we have examined. We conclude that at least some of these cuts were produced during thickness reduction or paring when very short shavings were being taken from the shaft in places where long shavings might split off a wooden splinter of greater thickness than was required. We have observed this practice in contemporary Western Desert communities. A very neat binding of tail sinew from Macropus sp. extends for 72 mm over the resin-cemented joint between the head and shaft of the spear found in the cave. A similar spear head (WAM A23455, Figure 7) which has been sawn from its shaft is less carefully spliced and exhibits part of the deep ‘V’ shaped notches used to interlock the head and shaft of such spears (Clement and Schmelz 1903: 5). This example has a rag strip binding the cemented join, but all other similar spears in the Western Australian Museum collection have sinew bound splices. The splicing method can be clearly seen on a point removed from the shaft (Figure 5). The ‘V’ form of the grooves used to produce a strong joint between head and shaft is quite evident. A small patch of spinifex resin can still be seen deep in the cleft. This resin has been used to cement the point to the shaft in the first step of hafting. About ten oblique cuts on each of the horns of the splice were provided apparently to offer a more secure attachment for the resin cement. This method is at variance with that described by Clement and Schmelz (1903: 5) in which the splice is first bound together with sinew and then covered with resin. On A2304, the spear found behind the wall, there are 35 shallow cuts on one side of the spear head, extending from the point of the barb back towards the butt. These average 1 mm wide and 5.6 mm in length. Comparative examples have cuts in a similar position, but the patterns vary (cf. Figures 4-7). These cuts are not the same as those used for thickness reduction. They serve no obvious function, but perhaps parallel the practice of identification marks and good luck tokens better known from the Western Desert (Douglas 1977: 20). This spear shaft (A23064) appears to have been carefully finished by scraping most of the surface with either stone or glass scrapers. Many of the branches along the shaft have been broken off, while others have been cut with a sharp tool. The shape of the notches on the spear head indicates the use of a metal imple- ment. In addition, some thinning of the shaft has been undertaken by slicing long thin slivers of wood rather than by shaving and scraping, the actions performed by stone tools. Although metal tools could have arrived in this part of the State 130 years ago, or even earlier if Dutch East India Company shipwrecks provided 119 Walled Rock Shelters and a Cached Spear in Western Australia 120 Figure 6 A343 - a similar spear from the same area now in the WAM collection. Length = 3.5 m. P. Bindon and M. Lofgren 121 Walled Rock Shelters and a Cached Spear in Western Australia artefact metal, we consider that the use of metal tools on this spear suggests it was made between 60 and 100 years ago. A barb (WAM A23456) broken from a spear (similar to A23064) is illustrated in Figure 8. This barb, found adjacent to an adze flake with gum adhering (Bindon: in prep.) on the floor of a rock shelter, about 3 km from the walled shelter, along with the discovery of point A23494 confirms the use of barbed spears until relatively recently in this area. Figure 8 A23456 — the barb discovered in an adjacent rock shelter. Length = 48 mm. Taking this and all the environmental factors into account, we consider that even in this arid climate and partially protected location, the preservation of wood for more than a century is unlikely. Comparison with other spears in the literature (Clement and Schmelz 1903; Davidson, 1934, 1936; Gould 1970) and in the collection of the Western Aus- tralian Museum suggest that while well known, this type of spear is not the most common form collected from the Pilbara Region. Comparative measurements on a similar spear in the WAM collection (A343 - reference location ‘North-West’, 122 P. Bindon and M. Lofgren c. 1902, Figure 6) show that the walled-in spear is shorter and lighter than others of its type, and is more carefully finished. The functional nature of the recently discovered spear as determined by the criteria discussed by Palter (1977) is of interest. The length of the spear lies close to the mean of his hand-thrown sample (2.73 m vs. 2.66 m). However, the mass of this spear (c. 300 g) lies much closer to the mean mass of spear thrower projectiles (246.3 g) than to hand-thrown spears (740.0 g). Similarly, the point of balance (0.44) is in the region of overlap between hand-held and spear thrower projectiles. The lack of a functional indentation on the butt of the spear argues for its being a hand-thrown projectile, but it appears to be at the morphological extreme of the range of such weapons. In his study of Pilbara rock art, Wright reports spears similar to that which were described (1968: 43, 62 and Figures 1, 44, 764). None of these are being thrown or thrust by hand although future discoveries may reveal such an example. Discussion Ethnographic accounts of Aboriginal life in the Pilbara contain no references to walled rock shelters. In part, this reflects a lack of field observation; few early writers ventured far from the immediate environs of the white settlements. Perhaps also at the turn of the century when these writers were active, some of the traditional exploitative activities of the Aborigines were being abandoned. C. and A. Clarke (in prep.) have obtained a number of contemporary Aboriginal accounts concerning the utilization of rock shelters, including walling, as described to them by descendants of the original inhabitants of the region. We conclude that fissures and niches were walled to encourage habitation by small game and perhaps to aid in making their capture more certain. In our inter- pretation the walls can be regarded both as a hunting device and as a strategy to increase the resource potential of the area. Most of the small shelters and openings were too small for human habitation, but were easily modified with small walls to provide a large number of safe habitats for various animal species. Mammals like rats (Rattus sp.) and possums {Trichosurus sp.) as well as reptiles like pythons {Aspledites sp.) and goannas ( Varanus sp.) are the most likely occupants of such environments based upon considerations of both size of apertures left in the walls and the known species for the area. The provision of passages either by the insertion into the wall of a suitable length of hollow tree-trunk, or by simply omitting to wall a niche completely suggests that access was deliberately provided. If entry to these places was being denied, then the walls would have been complete. The walls, with restricted entry passages provided an expanded safe habitat into which the invad- ing species population was allowed to grow quickly to the carrying capacity that the niche afforded (Odum 1969: 182 ff). As a hunting strategy, the practise we 123 Walled Rock Shelters and a Cached Spear in Western Australia describe is similar to that of placing hollow logs in streams to exploit later what- ever occupant takes residency (Roth 1901). After providing an environment, the hunter leaves the area for some time allowing the animal to move in. Sometime later, the wall is broken down and any resident animals taken for food before the wall is rebuilt, providing opportunities for the remaining population to again expand into the area. There are seven stone arrangements and numerous large surface campsites in close proximity to the low ridge which holds most of the walling discovered in this area. The concentration of stone arrangements suggests that at some stage in the prehistory of this area, large numbers of people gathered to participate in ceremonies. Such increase in density of the human population would place con- siderable pressure on food resources, and we suggest that this may account for the large number of walls scattered through the area. The walled shelters both increased the number of animals available and made their location more certain. There is no suggestion that the type of resource management described here would have the same importance that Bunya nuts had in the Queensland forests (Petrie 1904: 11), or Bogong moths in the Australian alps (Flood 1980). The scale of involvement is clearly at a lower level, and may suggest a correspondingly small population involved in any activities in this area. Some of the walls found following the initial discovery had more than half their length broken down, especially the more obvious large walls. This destruc- tion allowed sufficient space to afford human access into the chamber behind. Whether this was done by Europeans or by Aborigines is not known; however, it seems most likely that they were pulled down by inquisitive investigators. At this stage we are unable to account for the regional localization of the hunt- ing technique which we have proposed. The discovery of the typical Pilbara spear behind one of these walls in a rock shelter, indicates that the walls were constructed by Aboi’iginal people. The spear was placed so that the removal of only two or three of the wall slabs revealed its position; whoever built the wall was certainly aware of the presence of the spear. Aboriginal artefacts have always been considered as interesting souvenirs by white settlers and this specimen would have been removed by a European if it had been seen. We suggest that the spear was cached behind the wall by the Aboriginal builder of that structure. There seems to be no further connection between these two examples of Aboriginal technology. This spear is an excellent example of one of the Pilbara spear types and is well provenanced, providing an excellent type specimen. Aboriginal man used numerous contrivances and techniques in his hunt for game. Nets, brush fences, pit-traps, and various ambush techniques were used throughout Australia. If our assessment is correct, the technique proposed in this paper expands the range of hunting methods and underlines yet again the diverse resourcefulness of Aboriginal exploitation of the environment. 124 P. Bindon and M. Lofgren References Beard, J.S. (1975). Vegetation of the Pilbara area. Vegetation Survey of Western Australia. Perth: Univ. of W.A. Press. Beard, J.S. (1980). A new phytogeographic map of Western Australia. Western Australian Herbarium Research Notes 3: 37-58. Bednarik, R.G. (1977). A survey of prehistoric sites in the Tom Price region, north Western Australia. .(4rc/ifleo/. Phys, Anthrop. Oceania 12: 51-76. Brown, S. (1980). Ashes to Ashes. Unpub. B.A. Hons. Thesis. Perth: Univ. of W.A. Clarke, J., Dix, W.C., Dortch, C.E. and Palmer, K. (1978). Aboriginal sites on Millstream Station, Pilbara, Western Australia. Rec. West. Aust Mus. 6: 221-237. Clement, E. and Schmelz, J.D.E. (1903). Ethnographical notes on the Western Australian Aborigines. Int. Archiv. fur Ethnogr. 16: 1-29. Davidson, D.S. (1934). Australian spear-traits and their derivations. /our. of the Polynesian Soc. 43: 143-162. Davidson, D.S. (1936). The spear-thrower in Australia. Proc. Amer. Philosophical Soc. 76: 445-483. Dix, W.C. (1977). Facial representations in Pilbara rock engravings. In; Form in Indigenous Art: Schematisation in the Art of Aboriginal Australia and Prehistoric Europe, pp. 277-285. (Ed. P.J. Ucko.) Canberra; Australian Institute of Aboriginal Studies. Dortch, C.E. (1972). An archaeological site in the Chichester Range, Western Australia; preliminary account. /.i?. Soc. West. Aust, 55: 65-72. Douglas, W.H. (1977). Illustrated Topical Dictionary of the Western Desert Language. Canberra: Australian Institute of Aboriginal Studies, . Flood, J.M. (1980). The Moth Hunters: Aboriginal Prehistory of the Australian Alps. Canberra: Australian Institute of Aboriginal Studies. Gould, R.A. (1970). Spears and spear-throwers of the Western Desert Aborigines of Australia. Amer. Mus. Novitates 2403: 1-42. Hayden, B. Palaeolithic Reflections: Lithic Technology and Ethnographic Excavations Among Australian Aborigines. Canberra: Australian Institute of Aboriginal Studies. Kimber, R.G. (1981). Some thoughts on stone arrangements. The Artefact 6: 10-18. Love, J.R.B. (1936). Stone-Age Bushmen Today. London: Blackie &: Son. Maynard, L. (1980). A Pleistocene date from an occupation deposit in the Pilbara Region, Western Australia, Archaeology 10: 3-8. McCarthy, F.D. (1962). The rock engravings at Port Hedland. Pap. Kroeber Anth. Soc. 26; 1-72. McCaskill, D.L. (1977). Schematisation in rock art of the Upper Gascoyne district, Western Australia. In; Form in Indigenous Art: Schematisation in the Art of Aboriginal Australia and Prehistoric Europe, pp. 184-190. (Ed. P.J. Ucko.) Canberra: Australian Institute of Abor- iginal Studies. Mountford, C.P. (1940). Aboriginal stone structures. Trans. Roy. Soc. S.A. 64: 279-287. Odum, E.P. (1969). Fundamentals of Ecology. Philadelphia: W.B. Saunders & Co. Palmer, K. (1975). Petroglyphs and associated Aboriginal sites in the north west of Western Australia, Phys. Anthrop. Oceania 10: 152-160. Palmer, K. (1977a). Myth, ritual and rock art. Archaeol. Phys. Anthrop. Oceania 12: 38-50, Palmer, K. (1977&), Aboriginal sites and the Fortescue River, north west of Western Australia. Archaeol. Phys, Anthrop. Oceania 12: 226-233. Palmer, K. (1977c). Stone arrangements and mythology. Mankind. 1 1 : 33-38. Palter, J.L. (1977). Design and construction of Australian spear-thrower projectiles and hand- thrown Archaeol. Phys. Anthrop. Oceania 12: 161-172. Petrie, C.G. (1904). Tom Petrie’s Reminiscences of Early Queensland. Brisbane: Watson. 125 Walled Rock Shelters and a Cached Spear in Western Australia Radcliffe-Brown, A.R. (1930). Former numbers and distribution of the Australian Aborigines. Official Yearbook of the Commonwealth of Australia 23: 687-696. Roth, W.E. (1901). Food: its search, capture, and preparation. North Queensland Ethno- graphy Bulletin No, 3. Brisbane: Govt. Printer. Tindale, N.B. (1974), Aboriginal Tribes of Australia. Berkeley: Univ. of California Press. TrendaU, A.F. (1974). Hamersley Basin. In: Geology of Western Australia Memoir II; Geological Survey of Western Australia, pp, 119-143. Perth: Govt. Printer. Virili, F.L. (1977).. Aboriginal sites and rock art of the Dampier Archipelago, Western Aus- traha. In. Form in Indigenous Art: Schematisation in the Art of Aboriginal Australia and Prehistoric Europe, pp. 439-451. (Ed. PJ. Ucko) Canberra: Australian Institute of Aborigin- al Studies. Withnell, J.G. (1901). The Customs and Traditions of the Aboriginal Natives of North Western Australia. Roeboume: H.B. Geyer. Worms, E.A. (1954). Prehistoric petroglyphs of the Upper Yule River, North-Western Australia. Anthropos 49: 1067-1088, Worsnop, T. (1897). The Prehistoric Arts, Manufactures, Works, Weapons, etc., of the Abor- igines of Australia. Adelaide: Govt. Printer. Wright, B.J. (1968). Rock Art of the Pilbara Region, North-West Australia. Canberra: Australian Institute of Aboriginal Studies. Wright, B.J. (1977). Schematisation in the rock engravings of north-western Australia: form and identification in a living context. In: Form in Indigenous Art: Schematisation in the Art of Aboriginal Australia and Prehistoric Europe, pp. 110 116. (Ed. P.J. Ucko) Canberra: Australian Institute of Aboriginal Studies. Received 5 January 1982 Accepted 12 July 1982 Published 13 December 1982 126 Rec. West Aust Mus. 1982 , 10 ( 2 ): 127-131 A New Species oi Acanthistius (Pisces: Serranidae) from Eastern Australia J.B. Hutchins* and R.H. Kuiterf Abstract A new species of serranid fish Acanthistius paxtoni, is described from New South Wales. It is closely related to A. cinctus (Gunther), also from eastern Australia, but is separable on the basis of colour pattern and scalation. Morphological and colora- tion discrepancies also distinguish it from the Easter Island species^, fuscus Regan, an apparent close relative previously assumed to be synonymous with A. cinctus. Introduction Australian members of the Southern Hemisphere serranid genus Acanthistius inhabit inshore and offshore warm temperate reefs in depths ranging from the intertidal zone to at least 64 m. They are distinguished from other Australian serranids by possessing 13 dorsal fin spines and 99 or more vertical scale rows above the lateral line. In a recent paper describing a new species of Acanthistius from Western Australia (Hutchins 1981), four Australian species were recognized: A. cinctus (Gunther, 1859) and ocellatus (Gunther, 1859) from eastern Aus- tralia, and A. serratus (Cuvier, 1828) and the new A. pardalotus from VVestern Australia. The four species are morphologically similar and best separated by their diagnostic colour patterns. The most distinctively marked is A. cinctus with its well defined body bars. The other three species possess spotted and/or blotched colour patterns. The present paper describes a new barred species oi Acanthistius from New South Wales and compares it with the closely related A. cinctus. It is also contrasted with A. fuscus Regan, 1913 from Easter Island because of a previous report (Randall 1976: 336) that A. cinctus and the Easter Island species may be synonymous. Measurements were made as in Hutchins (1981). Both type specimens are housed at the Australian Museum, Sydney, hereafter abbreviated AM. Other abbreviations used are BMNH — British Museum (Natural History), WAM — Western Australian Museum and SL — standard length. * Department of Ichthyology, Western Australian Museum, Francis Street, Perth, Western Australia 6000. t National Museum of Victoria, 285-321 Russell Street, Melbourne, Victoria 3000. 127 A New Acanthistius from Eastern Australia Systematics Acanthistius paxtoni sp. nov. Figure 1 ; Table 1 Holotype AM 1.21555-001, 204 mm SL, collected by fish trap at Seal Rocks, New South Wales (32°26'S, 152°32'E) at 64 m, R.H. Kuiter, 25 May 1980. Paratype AM 1.16997-001, 258 mm SL, Watsons Bay, Sydney Harbour, New South Wales (SS^^Sl'S, 151°17'E), no other data. Figure 1 Acanthistius paxtoni sp. nov., holotype, AM 1.21555-001, 204 mm SL. Diagnosis This species is placed in the genus Acanthistius on the basis of the 13 dorsal fin spines and the large number of vertical scale rows above the lateral line (99 or more). Within Acanthistius^ A. paxtoni is distinguished from other Australian members on the basis of its distinctive colour pattern of six poorly defined dark cross bars on the body and the numerous pale wavy lines on the head and body (brownish-orange in life), which may break up into spots. It is easily separated from A. cinctusy the only other species which possesses body cross bars, by the absence of wavy lines on the body of the latter species (see Hutchins 1981, Figure la). Also, the scales on the operculum and upper half of the preoperculum are 128 J.B. Hutchins and R.H. Kuiter cycloid in A. paxtoni and ctenoid in A. cinctus. A, paxtoni differs from the Easter Island species, A. fuscus, in colour pattern {A. fuscus possesses no body bars or pale wavy lines), in its longer caudal peduncle (caudal peduncle depth 1.4-1. 7 in its length, 1.0 for yl. fuscus)^ as well as the condition of its opercular and preopercular scales {A. fuscus possesses ctenoid scales). Description Measurements of the holotype and paratype are presented in Table 1. The following counts and proportions of the paratype are in parentheses when differ- ing from those of the holotype. Dorsal rays XIII, 15; anal rays III, 8; pectoral rays 18 (19); lateral line pores to caudal base 51 (53), vertical scale rows from upper origin of gill opening to base of caudal fin 114 (111); scales in diagonal row from upper origin of gill opening to base of first dorsal spine 26 (28); scales in diagonal row from origin of first anal spine to lateral line 50 (57); gill rakers (including rudiments) on lower half of first gill arch 14 (16). Greatest body depth 2.6, head length 2.3 (2.4), snout to origin of dorsal fin 2.6, lower lip to origin of anal fin 1.3, postorbital length of head 3.9 (4.0), length of spinous dorsal base 3.0, length of soft dorsal base 4.5 (4.6), all in Table 1 Measurements in mm of the type specimens of Acanthistius paxtoni. Holotype AM 1.21555-001 Paratype AM 1.16997-001 Standard length 204 258 Head length 87 109 Snout length 22 27 Eye diameter 17 19 Interorbital width 11 14 Postorbital length of head 52 65 Greatest depth of body 78 98 Least depth of caudal peduncle 25 30 Length of caudal peduncle 35 50 Snout to origin of dorsal fin 78 100 Lower lip to origin of anal fin 155 199 Length of spinous dorsal base 69 87 Length of soft dorsal base 45 56 Length of pectoral fin 51 61 Length of pelvic fin 40 48 Length of longest dorsal spine 27 33 Length of longest dorsal ray 35 37 Length of longest anal spine 30 35 Length of longest anal ray 40 50 Length of caudal fin 45 52 129 A New Acanthistius from Eastern Australia Standard length. Snout 4.0, eye 5.1 (5.7), least width of bony interorbital 7.9 (7.8), least depth of caudal peduncle 3.5 (3.6), length of caudal peduncle 2.5 (2.2), length of pectoral fin 1.7 (1.8), length of pelvic fin 2.2 (2.3), length of longest dorsal spine (fifth) 3.2 (3.3), length of longest dorsal ray (third) 2.5 (2.9), length of longest anal spine (second) 2.9 (3.1), length of longest anal ray (fourth) 2.2, length of caudal fin 1.9 (2.1), all in head length. Interorbital space slightly convex; maxilla reaching level below posterior half of eye; opercle with three spines, middle spine much closer to lower than upper one; opercular flap pointed; preopercular margin rounded, upper limb coarsely serrate (lowermost serration somewhat larger than rest and directed downwards), three to four strong recurved spines on lower limb (both specimens have one bifid spine), increasing in size anteriorly; scales on body and dorsal surface of head mostly ctenoid, those on sides of head and ventral surface of body cycloid; predorsal scales extend to posterior nostrils; outer row of small conical acute teeth in both jaws, and an inner band of villiform teeth separated at the symphysis (some symphysial teeth in upper jaw more cardiform); a V-shaped band of villi- form teeth on vomer and a band of similar teeth on each palatine. Colour of holotype in alcohol: ground colour greyish-brown with six poorly defined dark cross bars on body; numerous indistinct pale wavy lines on upper half of body, extending somewhat obliquely towards dorsal fin base; two indistinct broad dark bars radiate from posterior half of eye, upper reaching almost to uppermost opercular spine, lower to above angle of preoperculum (an extension of upper bar continues from front edge of eye to snout tip); spinous dorsal, pelvics and inner surface of pectorals dark brown, other fins greyish-brown; with the exception of the spinous dorsal, all fins are bordered distally by a narrow pale line. The colour of the paratype is similar to the holotype with the following exceptions: pale wavy lines on body more distinct and more numerous (about 10-11 on mid-body), those on upper two-thirds of body extending obliquely to dorsal fin base, those on lower third breaking up to spots; pale lines radiate irregularly from posterior half of eye; all fins dark brown. Colour in life (based on a colour transparency of the freshly caught holotype, see Figure 1): head and body pale greyish-brown with many brownish-orange wavy lines, those on upper half of body extending obliquely to dorsal fin base and breaking into spots posteriorly, those on lower half also contracting to spots; lines on head radiate irregularly from posterior half of eye; body with six poorly defined dark cross bars, first from base of first dorsal spine to opercular flap, last across base of caudal fin; two indistinct broad dark bars radiate from posterior half of eye, upper one almost to uppermost opercular spine, lower to above angle of preoperculum; snout and dorsal surface of head dark brown with indications of brownish-orange spots; upper portion of maxillary groove with a dusky to brownish-orange streak; throat and breast pinkish-grey; all fins dusky, the dorsal and pectorals with irregular faint brownish -orange markings; with the exception of the spinous dorsal, all fins possess narrow pale distal borders. 130 J.B. Hutchins and R.H. Kuiter Remarks Acanthistius paxtoni is so far known only from two localities in New South Wales, Seal Rocks to the north of Newcastle, and Watsons Bay in Sydney Harbour. Its absence from intertidal rock pools, a habitat utilized by the juvenile stages of all other Australian Acanthistius species, suggests that this serranid is restricted to deeper reefs. Further collecting in this habitat will probably yield additional specimens from other areas along the New South Wales coastline. Acanthistius cinctus, known from New South Wales, northern New Zealand, Lord Howe, Norfolk and the Kermandec Islands is closely related to A. paxtoni. Both are morphologically similar to A. fuscus from Easter Island, although the latter species lacks the body cross bars characteristic of the former two. Unfor- tunately the present shortage of specimens of both A. paxtoni and A. fuscus precludes any detailed analysis of this relationship. This species is named after J.R. Paxton, Head of the Department of Ichthy- ology at the Australian Museum, Sydney, in honour of his contributions to Australian ichthyology. Additional Material Examined In addition to the material listed in Hutchins (1981), the following specimens were studied. Acanthistius cinctus: AM 1.18497-012, 4 specimens, 92-218 mm SL, Emily Bay, Norfolk Island, 16 September 1975; AM 1.10699, 320 mm SL, Lord Howe Island. Acanthistius fuscus: BMNH 1913.12.7.1, holotype, 180 mm SL, Easter Island, April 1911. Acknowledgements We wish to thank J.R. Paxton (AM) and P.J. Whitehead (BMNH) for kindly providing specimens housed at their respective institutions. We are also grateful to D.F. Hoese (AM) for bringing to our attention the second specimen (paratype) of Acanthistius paxtoni, and to G.R. Allen (WAM) for his comments on the manuscript. J.E. Randall of the Bishop Museum, Hawaii, kindly made available a colour transparency of fuscus from Easter Island. References Cuvier, G. and Valenciennes, A. (1828). Histoire naturelle des poissons. 2. Paris: Levxault. Gunther, A. (1859). Catalogue of the acanthopterygian fishes in the collection of the British Museum. 1. London: British Museum (Nat. Hist.). Hutchins, J.B. (1981). Description of a new species of serranid fish from Western Australia, with a key to the Australian species of Acanthistius. Rec. West. Aust. Mus. 8: 491-499. Randall, J.E, (1976). Ichthyological expedition to Easter Island. Natl Geogr. Res. Rep. 1968: 333-347. Regan, C.T. (1913). A collection of fishes made by Professor Francisco Fuentes at Easter Island, /^roc. Zool. Soc, London: 368-374. Received 1 June 1982 Accepted 22 July 1982 Published 13 December 1982 131 Rec. West. Aust. Mus. 1982 , 10 ( 2 ): 133-165 Distribution, Status and Variation of the Silver Gull Lams novaehollandiae Stephens, with Notes on the Lams cirrocephalus species-group R.E. Johnstone* Abstract Data on distribution, seasonal dispersal, colour of unfeathered parts and plumage stages are given for the Silver Gull (Larus novaehollandiae). Geographic variation within the species is analysed. Three subspecies are recognized, L. n. novaehollandiae Stephens of Australia (including Tasmania), L, n. forsteri (Mathews) of New Caledonia, and L. n. scopulinus Forster of New Zealand. Hartlaub’s Gull {Larus hartlaubii Bruch) of south-western Africa is treated as a full species. The Larus cirrocephalus species-group comprises four species of grey-headed and white-headed gulls from the Southern Hemisphere, L. cirrocephalus Vieillot, L. hartlaubii, L. novaehollandiae and L. bulleri Hutton. Introduction In his monograph on the world’s gulls, Dwight (1925) recognized five subspecies within the Silver Gull; Larus n. novaehollandiae from the coasts, islands and lakes of southern Australia north to Bernier Island in the west and the Five Islands in the east; L. n. gunni Mathews of Tasmania; L. n. forsteri of New Caledonia and coastal northern Australia from Port Darwin east and south to the Capricorn group; L, n. scopulinus of New Zealand; and L. n. hartlaubii of south-western Africa. Although Dwight was hampered by a shortage of specimens and data on soft parts etc. his treatment of the Silver Gull has remained virtually unchanged to the present day. Peters (1934) followed Dwight and recognized all five of his sub- species. Condon (1975) recognized only two instead of three subspecies for the Australian region: L. n, novaehollandiae for Tasmania and Australia except northern Queensland, and L, n. forsteri for the south-west Pacific from New Caledonia west to Torres Strait and south along the eastern coast of Queensland to Mackay. The main purpose of this paper is to examine geographic variation in Australia and to see if it can form the basis for the recognition of subspecies. To put the Australian variation in perspective it has been necessary to examine overseas populations of the Silver Gull and closely related species. * Department of Ornithology and Herpetology, Western Australian Museum, Francis Street, Perth, Western Australia 6000, 133 Distribution, Status and Variation of the Silver Gull Two terms used in the descriptions may need explanation. ‘Mirrors’ are the white areas towards the end of the primaries occurring on both webs of the first, second and in some cases the third primary (see Figure 1). ‘Tongues’ are the elongate areas of white or grey extending for variable distances from the bases of the primaries, usually on the outer web, sometimes on both. The primary wing patterns of gulls provide one of the most important characters used to classify them. For the purpose of this paper the primaries are numbered from the outer- most inwards. Movements are based on Garrick et al, (1957) and an analysis of recovery records in the Corella and its predecessor [Australian Bird Bander). OUTER WEB Figure 1 Primary feather showing mirror and tongue. Materials and Methods Two-hundred and three Silver Gulls [Larus novaehollandiae) held in the Western Australian Museum, South Australian Museum, National Museum of Victoria, Australian Museum, Queensland Museum, Australian National Wildlife Collection, Tasmanian Museum, Queen Victoria Museum (Tasmania), Dominion Museum (New Zealand), and American Museum of Natural History were examined in addition to 30 specimens of Hartlaub’s Gull [Larus hartlaubii) and 50 specimens of the Grey-headed Gull [Larus cirrocephalus) from the British Museum, Durban Museum, Transvaal Museum, and East London Museum. Measurements of specimens were taken as follow's: length of chord of flattened wing; length of tail (along a central rectrix); length of tarsus; length of entire culmen; and bill depth at the gonys (measured vertically from point of inflection on the lower mandible to upper edge of premaxilla). The white outer bar (mirror) on the primaries was measured along the shaft on each web or vane (see Figure 134 R.E. Johnstone 1). The length of white at the base of the first three primaries (tongue) was measured along the shaft on each vane (see Figure 1), and the length of black on the fourth primary was measured from the tip to the white tongue on the outer web. Data on unfeathered parts were taken from labels unless otherwise stated. Variation in the Silver Gull The Australian, New Caledonian and New Zealand populations of the Silver Gull are treated under the following geographic regions: (1) Tasmania, (2) New South Wales, (3) Victoria, (4) South Australia, (5) Archipelago of the Recherche (West- ern Australia), (6) Albany to Perth (Western Australia), (7) Fisherman Islands to Houtman Abrolhos (Western Australia), (8) Shark Bay to Exmouth Gulf (Western Australia), (9) Barrow Island to Broome (Western Australia), (10) Northern Territory, (11) Queensland, (12) New Caledonia, and (13) New Zealand. This facilitates comparison of the more distinctive populations, and data on soft parts and movements are more easily discussed. (1) Tasmania The Silver Gull is common on all coastal waters around Tasmania. It penetrates well up the major rivers and also occurs on the lakes in the Central Highlands. The breeding distribution is mapped in Figure 2. Tasmanian birds differ from all other Australian populations in having the greatest amount of white on the primaries (on both tongues and mirrors), especially on the third primary (see Tables 1, 2 and Figure 7). Females (and in some measurements males) are slightly larger than birds from Victoria, New South Wales and South Australia (see Table 4). Primaries in Adults The first or outermost primary is usually tipped black and is black over most of the lower basal portion of the feather. In some birds a white tip is followed by a black bar. There is a long white mirror the full width of the feather near the tip (see Table 1) and a white tongue at the base of the feather on each web. The shaft is in most cases black through the black portions of the feather and white through the white. The second primary is similar to the first, usually having a black tip and a large, white, subterminal mirror. The tongue is more extensive on both webs, being broader and longer (often one-third to half the length of the feather). Tiie shaft is hlack in the small black portion near the tip and the remainder, in most cases, white. The third primary is usually tipped white, followed first by a black bar, then by an irregular-shaped white mirror often continuous (especially on the outer web) 135 Distribution, Status and Variation of the Silver Gull Figure 2 Map of Tasmania, King I. and Furneaux Group, showing location of breeding sites (circles) and specimens studied (dots). 136 R.E. Johnstone with the broad, long white tongue (see Table 3). The shaft is white except through the black tip. The fourth primary has a white tip, a black bar and black inner margin grading into grey and white on the outer web. There is less terminal black than in main- land birds; length of black averages 25 mm in males and 24 mm in females, compared to 33 mm and 32 mm in Victorian birds (see Table 3). The fifth primary is similar to the fourth but the grey portion is decidedly greater. The remaining primaries are grey, slightly darker than the secondaries, with a blackish margin near the tip on the inner web (most noticeable on the sixth); the shafts are also grey. Compared toother Australian populations Tasmanian birds are easily recognized by their large wing mirrors, with the mirror on the third primary often confluent with the white tongue. One adult male from Trumpeter Bay, Tasmania, has continuous white (i.e. tongue joins mirror) on first, second and third primaries, resembling closely the wing pattern of Larus bulleri of New Zealand. Non-adult Plumages Downy chicks are mottled buffy-grey to blackish-brown above and whitish- grey below. Young up to six months of age have small mirrors on the first and second primaries, and extensive light brown markings on the wing coverts and secondaries. They also have biackish-brown subterminal bars on all but the outer tail feathers. Immatures eight to ten months old still retain some brownish mark- ings on the wings, the mirrors on the first three primaries are smaller than in adults (in some specimens there is no mirror on the third); the remaining primaries and the secondaries have broad blackish-brown subterminal bars. Unfeathered Parts In young up to six months of age, the bill is described on labels as black, the iris blackish-brown and the legs grey. In immatures the bill is red-brown (darker terminally) and the iris dark to white. Adults have the bill described on labels as red, vermilion-red or orange-brown; the iris white; the orbital ring red; the legs and feet cadmium red, red or orange-brown. Movements Two banded Tasmanian birds have been recovered in South Australia 950 km to the north-west. Tasmanian birds have been collected in Victoria and New South Wales. (2) New South Wales The Silver Gull is common in coastal New South Wales and is found on many islets and lakes along or near the coast. The breeding distribution in New South Wales is mapped in Figure 3. 137 Distribution, Status and Variation of the Silver Gull Figure 3 Map of New South Wales and Victoria showing location of breeding sites (circles) and specimens studied (dots). In size, birds from this region match fairly well with Tasmanian specimens (see Table 4), but on average are slightly shorter in the bill in both sexes. In colora- tion, birds from New South Wales differ from Tasmanian birds in having reduced white on the first four primaries, in both the mirrors and the tongues. Primaries in Adults The first primary has little or no white tongue, and the mirror is often mar- gined with black on the inner web. The second primary also has a narrow reduced white tongue (compared to Tasmanian birds) occupying a third to a half the length of the feather, and the mirror is often margined with black on the inner web. The third primary has a longer and broader white tongue (over half the 138 R.E. Johnstone length of the feather) than the first and second, but this is still shorter than in Tasmanian birds. No New South Wales specimen has the tongue meeting the mirror on the third primary as in many Tasmanian birds. The mirror on the third primary is in most cases oval-shaped or a spot, margined on both webs with black. It is absent in three out of eight adult males (all three from the Sydney area) and in one adult female from Byron Bay. It is also reduced in several other females, being absent on the outer web in six out of eleven specimens and on the inner web in three out of eleven. The fourth primary has a shorter white tongue than in Tasmanian birds. It is of interest that the specimens with reduced white on the primaries are all from north of Sydney, indicating a decline in the number of mirrors and the extent of the white tongues from south to north (see Discussion). Two specimens in the Australian Museum (030270 from the Grant collection labelled Sydney Harbour, and 042134 from near Sydney) have wing patterns of Tasmanian birds. The second specimen is possibly a migrant from Tasmania, but the first (like much Grant material) is probably mislabelled. Unfeathered Parts These are similar to Tasmanian birds, except for some specimens wdth a dark reddish-black tip to the bill. Movements Birds banded in New South Wales have been recorded in Queensland, Victoria and the Northern Territory. In New South Wales banding has shown that there is an average dispersal of young gulls in their first summer for about 400 km north- wards from the breeding colonies, with some travelling up to 800 km northwards. A few birds move south for varying distances up to 370 km but the main trend is northward. After the first winter there is much less movement with no recoveries more than 400 km from the place of birth, and the majority of recoveries and observations are within 80 km. The movement in New South Wales is nearly entirely coastal, in contrast to Victoria and South Australia where many birds move inland. Some New South Wales birds do occasionally move inland as indicated by a bird from Five Islands which was recovered onMt Ebenezer Station (160 km south-west of Alice Springs) in the Northern Territory. (3) Victoria The Silver Gull is common along the Victorian coastline and is more widely distributed inland than in any other State. The breeding distribution is mapped in Figure 3. Many adult specimens from Victoria, New South Wales and South Australia have varying amounts of grey on the head and nape. In size and coloration Victorian birds match most specimens from New South Wales, but have on average, a little more white on the third primary (see Table 3). 139 Distribution, Status and Variation of the Silver Gull Primaries in Adults Eight out of the ten adults studied have large mirrors on the first three primaries. One adult male and one female show no trace of a mirror on the third primary. Three out of six females studied have the mirror on the second primary margined with black. The white tongue at the base of the first primary occupies up to one-third the length of the feather (on both webs); just over one-third the length on the second primary; and up to two-thirds the length on the third. Unfeathered Parts These are similar to New South Wales and Tasmanian birds, but no adults were noted as having dark tips to the bill. Most were recorded as scarlet. Movements As in New South Wales there is a general dispersal of young gulls of one to two years old in all directions to over 1200 km from the birthplace. There is then a retraction of birds back towards the birthplace and most recoveries of birds over two years old are within 80 km of their birthplace. A substantial proportion of young gulls do, however, remain permanently at their place of birth. Banded Victorian birds have been recovered in Tasmania, South Australia and New South Wales. There are very few recoveries of banded Victorian birds form Tasmania, so it appears that there is little movement across Bass Strait. Unlike New South Wales where the main movement is northwards along the coast, most Victorian populations breed inland necessitating southward, eastward or westward move- ments to reach the coast hence the more varied dispersal. (4) South Australia In South Australia the Silver Gull is common along all coasts and on offshore islands. Apart from coastal areas, it breeds in the interior as far inland as Lake Eyre. As in Victoria this gull is part of the farming scene, birds following the plough and frequenting piggeries, slaughter yards and rubbish tips. The breeding distribution in South Australia is mapped in Figure 4. In size and coloration birds from South Australia are similar to Victorian birds but have slightly less white on the third primary. Primaries in Adults Six out of eight adult males and three out of eleven females have reduced mirrors on the third primary (compared to Victorian specimens), the outer web being black. One adult female has no mirror at all. The white tongue at the base of the first three primaries also matches well with Victorian birds; on the first the tongue is short (usually less than one-third the length of the feather, in most cases on both webs); on the second the tongue is over one-third the length of the 140 R.E. Johnstone Figure 4 Map of South Australia showing location of breeding sites (circles) and specimens studied (dots). feather on both webs; and on the third primary it is over half the length of the feather. As mentioned, South Australian birds are similar to Victorian birds, but there is a sharp contrast between them and the easternmost breeding population in West- ern Australia (Archipelago of the Recherche). The South Australian specimens are larger than the latter and have longer bills and much more white on the third primary (see Tables 3 and 4). Unfeathered Parts In young birds the bill is described as black or dark yellow with a black tip; the legs and feet dark yellow, yellow-brown or brownish. In adults the bill is red, 141 Distribution, Status and Variation of the Silver Gull deep dull red, dark red, orange-red, red with dark tip, red tipped with black, or orange with a black tip. The orbital ring is red the legs and feet are red, deep dull red, orange-red or dull orange; and the iris is white. Movements Birds banded in South Australia have been recovered in Victoria, New South Wales and Tasmania. In April 1979 two specimens were collected from a flock of 25 at Lera Waterhole (Gregory Salt Lake), Western Australia, which are in measurements and coloration most like South Australian birds and probably came from there. The back and wings are darker grey than in Western Australian birds and the shape and size of the mirrors match South Australian specimens. Small flocks of Silver Gulls were again recorded on Gregory Salt Lake in June 1980. A specimen collected at Newman, Western Australia, in November 1981 also matches best in coloration with South Australian birds. As the Western Australian populations are fairly sedentary and almost strictly coastal, many inland records from that State (Pilbara, Nullarbor Plain, Eastern Goldfields and Wheat Belt) are probably visitors from South Australia. (5) Archipelago of the Recherche The Archipelago of the Recherche contains the easternmost breeding popula- tion in Western Australia. The break between this and the westernmost breeding colony in South Australia is over 1000 km. The breeding distribution in Western Australia is mapped in Figure 5. Birds from the Archipelago of the Recherche are the smallest in Western Australia, especially in wdng, bill and weight measurements (see Table 4). They have smaller primary mirrors (particularly the third) than South Australian specimens or those from the Albany to Perth region. All five specimens collected in December 1979 (the other two are old specimens) had a rosy tinge on the breast and flanks. This faded soon after death, but it has not been observed in any other western population or referred to in any eastern birds. A rosy tinge to the plumage is, however, recorded for the New Zealand race of the Silver Gull Lams novaehollandiae scopulinus and for the closely related Black-billed Gull Larus bulleri, also of New Zealand. Primaries in Adults The white mirrors on the first two primaries are only slightly smaller than in populations 6 and 7 (birds from Albany to Houtman Abrolhos); however, the third primary has no mirror or a greatly reduced mirror. The two adult male specimens from the area have no white on the outer web of the third primary and only a narrow strip of white enclosed by black on the inner web. Of five adult females only one has a white mirror on the outer web of the third primary (four birds having the outer web all black), and only two out of five have a white mirror on the inner web of the third primary (three having the inner web black). Six out 142 .a • R.E. Johnstone Figure 5 Map of Western Australia showing location of breeding sites (circles) and specimens studied (dots). 143 Distribution, Status and Variation of the Silver Gull of seven birds have no mirror on the outer web of the third primary, and three out of seven no white mirror on the inner web. An adult male collected near Cape Arid on the mainland has no trace of a mirror on the third primary matching most island females. The basal white tongues match specimens from Albany to Houtman Abrolhos, being long and broad and not reduced as in the north-western populations, which also have small or only two mirrors. Unfeathered Parts The bill is maroon, orbital ring red or bright red, legs and feet maroon, iris white and mouth bright red. Movements Based on the male collected near Cape Arid it appears there is some movement of birds between the Archipelago of the Recherche and the mainland. Silver Gulls commonly seen at Esperance have not been studied, (6) Albany to Perth The Silver Gull is common on south-western coasts. It is far more sedentary in Western Australia than in any other State, specially New South Wales, Victoria and South Australia. Occurrences inland are also much rarer than in other States. In the Perth to Fremantle area (where this species is in plague numbers) they ascend the Swan River only as far as Guildford (27 km) and are absent from many near-coastal lakes south and north of the Swan River. Although gulls have been seen on several large inland lakes such as Lake Grace, Lake Dumbleyung, the Wagin lakes, and at Northam and Rawlinna, no specimens are available from these areas and, as at the Gregory Salt Lake, the birds could be visitors from south- eastern Australia. Tlie only permanent inland breeding colony is one of about 40 pairs that nest on islets in Lake Muir in the south-west of the State. Specimens from the Albany to Perth region are larger than Recherche birds (see Table 4). Compared to South Australian birds, specimens from this area have slightly less white on the mirrors, and south-western females have shorter bills. Primaries in Adults Five out of nine males have no mirror on the outer web of the third primary, and in three others the white is narrow; five have all black inner webs, and in two others the white is very narrow. Two out of six females have no mirror on the third primary, and in the other four the white is narrow. Nearly all the specimens with only two or greatly reduced mirrors are from the northern part of this region, the Perth area. A small proportion of breeding specimens from Carnac 1. (near Perth) have only two mirrors, and some individuals have just a faint white spot on the third primary. Along the west coast it is within 144 R.E. Johnstone this and the following population that the change from three to two mirrors occurs. The white tongue on the first primary is short and narrow; on the second longer and wider (up to half the length of the feather on both webs); and on the third long and broad (to almost two-thirds the length of the feather on both webs) (see Figure 7). Non-adult Plumages Birds five weeks old have the entire underparts white; the face and crown grey, with a blackish-brown spot in front of the eye; head grey; cheeks whitish; mantle, back and wing coverts light grey to greyish-white, each feather with a subterminal blackish-brown band (darkest on wings) and fringed pale reddish-brown. The primaries are black with small mirrors on the first and second. The secondaries have black subterminal bars. The rump is light bluish-grey, and the tail white with blackish-brown to buff subterminal band on all but the outer two rectrices. At five weeks they still have traces of brown on wings and tail. The tail band is lost first, then the speckled coverts and finally the black bar on the secondaries. Adult plumage is attained in 10 months. Unfeathered Parts Adult breeding specimens from Carnac I. have the bill red, dark maroon with tip tinged black, orange -brown with blackish-brown tip, brick-red with reddish- brown tip or maroon to brick red with tip tinged brown. The iris is white. The orbital ring is reddish-orange, orange or orange-yellow. The mouth is orange or reddish-orange. The legs are brick-red, dark reddish-brown, reddish-brown, reddish- orange or orange-yellow. Nicholls (1964) noticed that in her captive colony the adult birds showed a waxing and waning in the coloration of the bill, legs, feet, mouth and eyelids, with peaks in late autumn and mid-spring. This is consistent with the breeding pattern on Carnac I. which Wooller and Dunlop (1979) have shown to be trimodal with peaks in autumn (March to May), winter (June) and spring (September to November). In immature birds the iris is dark brown; orbital ring greyish-black; legs grey; basal half of bill greyish-brown, the rest blackish-brown; and the mouth pink. Movements The largest populations of Silver Gulls in Western Australia occur near Perth and Albany, but there appears to be less movement along the south coast than along the lower west coast. Despite large numbers banded in Western Australia, none have been recovered in any other State. The longest recorded distance travelled by a banded bird is of one ringed near Perth and recovered at Busselton, 195 km to the south. Young birds banded on Carnac 1. have been recovered breeding on Seal I., Lancelin I. and Rottnest I. There is little evidence supporting 145 Distribution, Status and Variation of the Silver Gull the view of Serventy et al. (1971), that there is a mainly southern movement in this State. In fact the above-mentioned birds recovered on Lancelin I. show the opposite trend. Furthermore I have seen no birds from south of the Houtman Abrolhos that could be ascribed to populations to the north. Banding records combined with geographic variation indicate that there is local movement up to about 200 km along the coast. As mentioned earlier, movements inland seem to be on a still smaller scale. (7) Fisherman Islands to Houtman Abrolhos In size and coloration this population is generally similar to population 6 (see Tables 1, 2, 3 and 4). There is a slight increase in the amount of white on the first two mirrors, but a slight decrease in the amount of white on the third mirror, it being more deeply enclosed by black. Primaries in Adults In one of three males the mirror on the third primary is not present on the outer web, and in the other two males the white mirror is narrow (enclosed by black) on both webs. One of four females has no mirror on the third primary, and in the other three the white is very narrow on both webs. Unfeathered Parts These are as in population 6. (8) Shark Bay to Exmouth Gulf Primaries in Adults Specimens from this region (Dorre I., Lake MacLeod and Mangrove Bay) show a marked decrease of about 20 mm in the size of the mirror on the first primary on both webs. In females there is no overlap in the size of the mirror with population 7, and in males very little overlap (see Table 1 and Figure 7). On the second primary the mirror is similarly 20 mm shorter than in population 7. The third primary has no white mirror, and no specimens from Lake MacLeod north- wards in Western Australia have any trace of a white mirror on this feather. The white tongues on the primaries, particularly on the second and third, are also reduced. On the first primary there is a little white at the base or no white at all; on the second it is on both webs but narrow and under one-third the length of the feather; and on the third it is a little longer and wider but less than half the length of the feather (see Figure 7). The length of black on the fourth primary is greater than in population 7, but it compares well with more northern birds (see Table 3). Unfeathered Parts These are similar to populations 6 and 7. 146 R.E. Johnstone (9) Barrow Island to Broome The Silver Gull is generally scarce along the Pilbara and Kimberley coasts but is locally common in these regions about coastal towns and ports and at breeding colonies of Brown Boobies and Lesser Frigate-birds (Bedout I. and Lacepede Is). There are few records from northern coasts fringed with mangroves, the gulls preferring the sandy beaches. Specimens from this region match well with population 8, but females have a slightly longer bill than elsewhere in the State (see Table 4). Primaries in Adults There is a slight decrease in the size of the mirrors on the first two primaries. The inner webs of both are usually fringed with black. As in population 8 there is no mirror on the third primary and the white tongues are short. Unfeathered Parts The bill is red with a blackish-brown tip, iris white, orbital ring grey-brown, legs deep red and mouth red. Movements In both the Pilbara and Kimberley this species is almost purely coastal and rarely wanders more than a few kilometers inland. Storr (1980) gives the status of the Silver Gull in the Kimberley as very common at certain ports (Broome, Cockatoo I., Koolan I.); uncommon elsewhere. Breeding probably occurs on the Lacepede Is but at present the only known breeding locality in the Kimberley is Jones I., but this island is possibly not currently used, for gulls are scarce in this area and there are no other breeding records since 1901. The main Kimberley population is centred around the Lacepede Is. There is thus a huge break in the breeding range from south-west Kimberley to the north coast of the Northern Territory. (10) Northern Territory According to Storr (1977) the Silver Gull occurs along northern coasts and islets including Melville I., Goulburn Is and Groote Eylandt. It is locally common, e.g. on Groote Eylandt, but is generally uncommon. It favours blue-water seas off rocky and sandy coasts and islands. Breeding is recorded on Haul Round 1. and in Melville Bay. Only five specimens are available from this area, and on measurements and coloration they appear not to be visitors from elsewhere and are probably resident birds. They match better with specimens from north Queensland than Kimberley. Primaries in Adults The mirrors on the first two primaries are slightly larger than in Kimberley birds, and the two males have a small white spot or small mirror on the inner web 147 Distribution, Status and Variation of the Silver Gull of the third primary. On one specimen the spot on the third primary is only present on one wing. The white tongue on the first primary is short and narrow and only on the extreme base of the feather. On the second primary the tongue is over one-third the length of the feather and on the third over half the length of the feather. On both the second and third the tongues are narrow, especially on the inner web. Unfeathered Parts The bill is recorded as dark scarlet, dull scarlet, dull dark red, or blood-red. The iris is white or ash grey. The orbital ring scarlet-orange. The legs scarlet-orange, pale scarlet, orange -red or blood-red. The tip of the bill is not recorded on any specimens as being darker than the base as in most Western Australian populations. Movements The Northern Territory population appears to be sedentary with some local movement away from the coast to outer islands especially in the breeding season. This area could also expect migrants from south-eastern Australia (particularly South Australia); which could be distinguished by their three large wing mirrors. (11) Queensland The Silver Gull is fairly common along the Queensland coast as far north as Cairns. Gulls are occasionally reported well inland, particularly after cyclones. Storr (1973) gives the range oi Larits novaehollandiae forsteri in Queensland as northern and mid-eastern coasts and islands, south to Mackay and Lady Elliot 1.; and the nominate subspecies as south-east coast, estuaries, and islands, north certainly to Bribie I., and in the western and southern interiors. The breeding distribution in Queensland is mapped in Figure 6. There are two large breaks in the breeding range. (1) between Cook Island New South Wales and the Bunker Group (Queensland) a distance of about 470 km, and (2) between Holbourne I. and the Low Is, a distance of 450 km. Six of the 17 specimens available from Queensland are probably migrants from New South Wales, having three large mirrors or in the case ol immatures larger mirrors than Queensland birds of the same age. In all, only ten adults were studied that probably came from the Queensland breeding population: two from Litel I. (Torres Strait), three from Purangi 1. (Cape York), one from Tukna Creek (Cape York), two from the Endeavour River, one from Mt Inkerman and one from near Sarina. Primaries in Adidts The Queensland birds match well with Northern Territory specimens. They have less white on the primaries than most New South Wales birds but more than Kimberley birds. One Queensland male has a small white spot on the outer web of the third primary and only two have small white spots on the inner web. The 148 R.E. Johnstone basal white tongues are slightly narrower and shorter than in Northern Territory specimens. An immature female from Rennell 1. (Torres Strait) has a small white spot on the outer web of the first primary and on both webs of the second primary. Figure 6 Map of Queensland showing location of breeding sites (circles) and specimens studied (dots). 149 Distribution, Status and Variation of the Silver Gull Unfeathered Parts Queensland specimens have the bill described as purplish-red or vermilion red, the iris pale grey or chrome yellow and the legs scarlet or vermilion red. Movements Little is known about the movements of Queensland birds. The similarity in colour pattern and size between Northern Territory and Queensland birds suggests that the Northern Territory has been colonized from Queensland. (12) New Caledonia An adult male from New Caledonia is the only specimen examined from within the range of forsteri {sensu stricto). Usually Queensland and coastal Northern Territory birds are placed in L. n. forsteri (type locality New Caledonia). However the single specimen examined from New Caledonia differs from all adult Aus- tralian birds in having only one wing mirror. It also has a greyish crescent on the nape and broad white tips on the second, third, fourth, fifth and sixth primaries. The mirror on the first primary matches well with north Queensland specimens in size and shape. The second primary is a new feather and shows no trace of a white mirror, nor does the third primary which is also new. Judging from this specimen and Mathews’ (1912) description of forsteri the New Caledonian birds have much less white on the primaries, and in this respect differ markedly from Queensland birds. Further collecting within this region is necessary for clarifying the status of L. n. forsteri. (13) New Zealand The Red-billed Gull [Larus novaehollandiae scopulinus) occurs on North and South Islands, and on Three Kings and Chatham Islands, Stewart, Great, Little Barrier, Kapiti, Snares, Auckland and Campbell Islands. It is a straggler to the Kermadec Islands. It is very common along the coasts, especially about harbours, but is also found on inland lakes, breeding inland at Rotorua. Mills (1969) maps the breeding distribution in New Zealand. New Zealand birds are smaller than those from all Australian populations (see Table 4), especially in wing and bill measurements. Most of them have only two mirrors, which are smaller than in Australian birds (see Tables 1, 2, 3 and Figure 7), and the mirrors run out on the inner web at right angles to the feather shaft (the angle is more acute in Australian birds). The basal white tongues on the first three primaries, especially the third, are longer and broader than in most Australian populations. In this respect they differ markedly from Hartlaub’s Gull of Africa, which also has only two mirrors but greatly reduced tongues. Several New Zealand specimens show a faint greyish tinge to the nape, and Buller (1888) mentions that some birds in nuptial plumage have the breast and sides suffused with a delicate roseate tint. The white mantle 150 R.E, Johnstone Figure 7 Typical wing patterns of adult Silver Gulls from (left to right) New Zealand, Tasmania, Victoria, Archipelago of the Recherche, Albany to Perth and Shark Bay to Exmouth Gulf. 151 Distribution, Status and Variation of the Silver Gull merges into the pale grey back. The back and wings are pale grey or blue-grey as in Australian birds, and in most specimens are lighter than in Hartlaub's Gull. Primaries in Adults The first primary has a short to moderately long white tongue, usually on both webs but longest on the inner web (up to just over one-third the length of the feather). The white mirror runs out on the inner web at right angles from the shaft (see Figure 7). The rest of the feather is black except for a white tip in specimens showing no wear. The second primary has a broad white tongue, longest on the inner web and up to half the length of the feather. The outer edge of the tongue is occasionally margined with black. The mirror again runs out from the shaft at right angles on the inner web (on the outer web the angle is more acute). The tip is white in unworn plumage. The third primary has a long broad white tongue up to two-thirds the length of the feather. The tongue is edged on the inner web with greyish-white or greyish- black. The rest of the feather in 12 out of 14 specimens is black, i.e. lacking a mirror but tipped white in unworn plumage. Two adults, one from Lower Hutt (No. 13384) and the other a male from Stewart I. (No. 15227), have a fairly large white spot or small mirror on the inner web of the third primary. Although the nominate form has been recorded in New Zealand, the above specimens are certainly not Australian, for they have the square-cut mirrors on the first two primaries characteristic of scopulinus. Oliver (1930: 267) depicts a bird from Kapiti I. with three primary mirrors. The fourth primary has a white tongue on the outer web. The inner web is grey and white margined with black the rest of the feather is black except for a white tip in unworn plumage. Unfeathered Parts The bill is dark red, lighter on the ridge and towards the tip, iris silvery white, orbital ring red, and legs and feet red. Movements No New Zealand birds have yet been found in Australia. Dwight’s (1925) records of the Australian form as casual in New Zealand could be based on the odd New Zealand bird with three primary mirrors. At any rate they are not accepted by the Checklist Committee of the Ornithological Society of New Zealand (1970). Variation in Hartlaub’s Gull Hartlaub’s Gull (L. hartlaubii) breeds on islands off south-western Africa from Walvis Bay south at least to Cape Agulhas (Clancey 1964). It favours the cold 152 R.E. Johnstone coastal waters of western South Africa but ranges occasionally as far east as Natal. Hartlaub’s Gull has long been treated as a race of the Australian Silver Gull. African birds are smaller and darker than all Australian populations and match best in size with New Zealand birds. They average slightly longer in the wing than New Zealand specimens and have longer and narrower bills. The upper and under wing and the back are darker grey than in Australia and New Zealand. Primaries in Adults Most L. hartlaubii have only two white mirrors in the wing (two females have a small white spot on the inner web of the third primary), and the mirrors average smaller than in New Zealand specimens (see Tables 1, 2, 3 and 5). The basal white tongues on the first three primaries are much shorter and narrower than in all populations of novaehollandiae. In 15 of 18 adult hartlaubii there is no white tongue at the base of the first primary, and in the two that have a short tongue it is greyish-white rather than white. The mirror on the first primary is in most specimens margined with black on the inner web. On the second primary the tongue is longer (up to half the length of the feather) on the outer web. On the inner web the tongue is absent or greatly reduced and is often only present as a greyish-white line along the shaft. The mirror is broadly margined with black on the inner web in 12 out of 18 adults. The third primary has a long white tongue on the outer web, usually about half, but up to two-thirds the length of the feather. Where present on the inner web the tongue is narrow (1-4 mm wide) and merges with the grey (in other words scarcely crossing the shaft on to the inner web). In 16 of 18 adults there is no mirror on the third primary. The exceptions are two females (one from Port Nolloth, the other from Berg River); which have a small white spot on the inner web of the third primary. The fourth primary has a long white tongue on the outer web. On the inner web the tongue is reduced to a thin greyish-white line running along the shaft and widening near the end. The black on the inner web is in most specimens more extensive than in Australian and New Zealand birds; i.e. there is less grey or white. In this character some L, hartlaubii match with Grey-headed Gull (L. cirrocephal- us)\ however the amount of black is variable, some specimens having a fairly broad greyish area on the inner web. In several other characters L. hartlaubii is more like L. cirrocephalus than L. novaehollandiae. The white mantle in adult hartlaubii and cirrocephalus is well defined and contrasts sharply with the dark grey back (in novaehollandiae the white mantle grades into the light grey back). Breeding hartlaubii assume a grey bar on the nape that commonly extends to the side of neck and less frequent across the throat. This collar is usually most distinct on the hind neck. It is here 153 Distribution, Status and Variation of the Silver Gull that the dark hood of cirrocephalus meets the white of the neck and it is here that the margin of the hood is darkest. Unfeathered Parts The bill is cherry red (3), reddish-black (3), reddish-brown (2), flesh (1). The iris is brown (10), greyish-brown (4), yellowish-brown (2). The orbital ring is red (2), dark maroon (1), cherry red (1). The legs are red (6), cherry red (2), brownish (2), reddish-brown (1), tan brown (1), flesh (1), black (1). Variation in the Grey-headed Gull The Grey-headed Gull [Larus cirrocephalus) is divided into two subspecies, separated by the Atlantic Ocean. The nominate subspecies occurs in South America north to Ecuador and north-eastern Brazil, and L. c. poiocephalus in Africa (north to Gambia and Ethiopia) and Madagascar. On both continents it occurs on fresh waters and coasts. The African subspecies L. c. poiocephalus is smaller than the nominate form and has smaller mirrors (see Tables 5 and 6). The mantle is purer white and well demarcated from the back, rather than being tinged with grey as in many South American specimens. Grey-headed Gulls differ from Hartlaub’s Gull in their larger size, more exten- sive grey on the head (even in immature and non-breeding plumages), smaller mirrors on the first two primaries and less white or greyish-white on the fourth primary. The iris is yellow in most cirrocephalus and brown in most hartlaubii. The Grey-headed Gull favours fresh waters, hartlaubii cold coastal waters; though each may be found in the preferred habitat of the other species (Clancey 1964). Primaries in Adults The colour pattern of the primaries is much the same as in L. hartlaubii. The first primary is black with no trace of a tongue and with a squarish white sub- terminal mirror, margined with black on the inner web. The second primary is mostly black with a narrow white tongue on the outer web and a small white mirror margined with black on the inner web. The third primary is black, without a mirror and with a moderately long tongue (usually only on the outer web; sometimes a trace of white on the inner web). The fourth primary has a long white tongue on the outer web, the rest of the feather being mostly black but becoming blackish-grey at the base on the inner web. The length of the sub- terminal black bar on the outer web averages 70 mm in the nominate form and 71 mm in poiocephalus. Unfeathered Parts The bill is red (3), dark red (2), light red (1), dull crimson (1), dark coral (1), dull red (1) and dark brown (1). The iris is yellow (3), maple yellow (2), light yellow (1), pale (2), white (1), brown (2). The orbital ring is light red (1). The legs 154 R.E. Johnstone are red (4), light red (1), coral (1), brown (2), dull vermilion (1), dull red (1), dark brown (1). Movements Generally the breeding ranges of Hartlaub’s Gull and the Grey-headed Gull do not overlap. In southern Africa cirrocephalus breeds regularly and in good numbers in the southern Transvaal, and occasionally elsewhere in the interior if conditions are suitable. They disperse widely from their breeding grounds, cover- ing all of South Africa and commonly reaching southern .Mozambique and southern Angola in the north. Hybridization Each year small numbers of Grey-headed Gulls enter the range of L. hartlaubii and occasionally form mixed pairs or pair themselves among hartlaubii. Hybridization between Larus cirrocephalus and Larus hartlaubii has been record- ed from time to time (Sinclair 1977), but obvious hybrids are rare in collections. It would be interesting to note the iris colour of a mixed pair, as some breeding cirrocephalus have the iris described as brown (rather than yellow) and odd hartlaubii have it yellowish-brown (rather than brown). In many gulls unfeathered part coloration and calls play an important part in species recognition and pair bond formation, and it can be seen from the unfeathered part data that some cirrocephalus and some hartlaubii could match. Judging from descriptions, the breeding displays of L. cirrocephalus j L. hart- laubii and L. novaehollandiae, are almost identical. 155 Table 1 Measurements (mm) of Larus novaehollandiae showing the length of the white mirror on the first primary, with means and sample size in parentheses. Distribution, Status and Variation of the Silver Gull ^ w u 2 u . 0 ’■2 C ® — 4^ WD 0 5 o J5 w £ u ^ ^ S fa. w *2 nJ £ 1-^ 00 Tf O CO O) CM CO 40 if) 40 40 ^ lO If) if) if) -w »• CM IT) 00 I-H CO if) 2 o 40 CO r>- r>. lO 40 40 if) if) 40 if) if) CO lO 05 1— 1 yb 40 CM CM CO CM lO IT) CO CM CO CM CO .Q V •* •M U 3 «2 O 3 s, 3 O • H G .3 O •A 4^ 3 ® 0 ^ o r>. f>- CM if) 40 if) if) if) CO 00 if) if) if) if) CO if) if) •5 f £ C fa. qj .j- faJ £ 40 <—1 00 O) 00 iO if) 40 40 40 O CO 00 if) 40 40 CM CO ^ if) o o 40 CO ■'t CM ^ CM CO if) UD Tj* T-H 40 40 ^ o rt* „ ^ o . 1—1 40 f'. T— H to' CM 40 T— ' S-S, CM if) S-S^ 00 if) S-5- *o o *0 o *o o *0 o *o o *o o V) j: 4.) 2 •M C.M ■£ fa. 2 ’S fj .o -*-• 3 o c/3 *2 "IS fa. V5 3 < O o ho qj qj 4J a ^ fa. ■t-t CA 3 < G fa. 4J fa. OJ CL. 0 >« G g +■> •M fd E ’fa. a trt rtS H 3 0 c/3 U CJ fa. 4J < cc: CA o; OJ < M fa. CM CO Tt' if) 40 156 Western Australia 9 (N5) 40-61 (53) 32-53 (45) Tongue joins Tongue joins mirror Mirror Continuous Continuous inner web Length of white white outer Outer web Length of white white inner enclosed First primary mirror outer web web black mirror inner web web bv black R.E. Johnstone •-( CO o xO O o o lO XT) CO CO CM CO CO CO ' — ' r>- ^ <£) 00 CM KD CM a-i t£> xo xO Tf CO CO CO CO ^ 1-H 00 t-H ^ r'. O CM rH CO ^ T— 1 1—1 CO CO CO CO CM O lO CO O CO 1— ( 00 o 'S' s s «5 CO CO XO CO ' '«— «• i—t O 00 ^ o CO 00 r>- KD xO rf XO xo xO xO xo y5 00 o o o ^ xf3 CO CM CO CO CO CO CM CO XO ^0 TjH xO r-- CO 2 2 2 2 2 2 2^ 2^ 2^ 2 2 ' >—>• '■ — ' - — ' ' — ' *o o *0 o *o o *o o x> o *o *o o o V3 T3 C/3 1-2 2 o c« cr 0 c ci o u >-• -9 « < 3 0 U-i 3 O ■M cn 3 4-> -c 3 Xh 4^ (A 3 •M *C X4 1) -c 2 *3 0 T3 3 c C ^ >* Pi -C < Vi < 3 3 *0 11 d 19 1 £ h 11 CQ w 3 0 S s-< 11 o V E 3 X4 11 Xh 11 X Vi 3 u 11 N -C 2 V) Xh d 6 ■M Vi fc 0 Vi •M X4 u u ■ 2 0 H X V w Sm 11 o 3 11 u Ui K ^ c/D CjJ CQ CQ 2 a 2 2 00 o r-H CM CO 157 31-62 (46) 34-57 (45) Table 2 Measurements (mm) of Larus novaehollandiae showing the length of the white mirror on the second primary, with means and sample size in parentheses. Distribution, Status and Variation of the Silver Gull c 4 , ^ Si ^ w O u O b. u .a b d w ^ 5 c = ^ 'S ® .2, o c I) O -ZJ « g I g I -8 H £ tj ^ ^ £ ^ 5 ■€ g Z ^ w ^ u s w jg 0^3 Si ^ -a ^ S ^ fj 0 H o 3 c >« U 9|> lO urj vT) S 1 — ( t-H m xTi CM o CO ^ r'- ^ ITJ CO CO 52- ^ 00 CO oT © CO CO u u -H 00 © r- o CO © CO ® 0 © lO •3- 51- UD S to 0 C k. O 00 UD © -H o r— ( t-H © r-- iO lO yp yp m © © H '5 Oi © CM CO ^ r-^ © Id r-. CM CO CO 4^ CO CO CO ^ CM ^ ^ © © 00 r-i r-* T— ( CM © 2^ 2 2 2 2 - — ' - — - *0 o *o o *o o *o o ’O o (y O a z u H D 158 Albany to Perth d (N8) 29-55 (46) 1 26-49 (41) Western Australia 9 (N5) 33-54 (45) 30-58 (41) Tongue joins Tongue joins mirror Mirror mirror Mirror Continuous outer web Outer Length of Continuous inner web Inner white outer enclosed web white mirror white inner enclosed web R.E. Johnstone u 3 JSi 3 St CO « -s •= ^ ® o u • — u B o csr CO ■4' CO CO a> rh CM CM CO CO o o CM CO cf> s CO 00 CO CM <£> Tj< CO CO o CO CM (9 E •M M s < o 3 o ■£ 4-> W) 3 < C CO V B H c T3 C 3 O T3 o; 5 o c u \Ci Tfi CO Csl 1-H CM CO CO CM CM CM 1-H O O 00 Gt O CM CM O) CO CO (jp ^ 5 CM CO 5 5 CO 5 5 1-H r>- iO CO o 4« 4* VO CO O O CO CO CM 1-H CM CO CM CO CM o r-H __ CO iO CO xO xn c<^ Z Z 5,S, z z Z Z z z z ' — ' ' — ' *0 o •o o *o o •o o *0 o •o *o o 159 Table 3 Measurements (mm) of Larus novaehollandiae showing the length of the white mirror on the third primary, and the length of the black subterminal bar on the fourth primary, with means and sample size in parentheses. Distribution, Status and Variation of the Silver Gull I - Si (A V JS J XI ^1 ■s o- e X £ ^ U 0 a ri i 3 X u -O X ^ ^ ^ is ^ 2 «« 0 fc- O — b u X .a a ^ ^ o s 3 e c * o C t C -2 X o •- ox W H E U ^ 5 s h. c Sc P ’•*-» X s I f .S £ J 1 1 a ^ X Si G >• U X 3 V g 3 s h e g) o *3 C fc C .3 X 0 -S 0 X 4J H E U ^ ^ 2. 1 I a> o t t O r-. CO CM O CO 't* CO CM CM I ®? lO If) CM CM CM CO 00 o X CO CM . — . 00 O CO CO CO CM m 1-H l-H J-H TjH o ^ rt’ f" Tf If) cn lo CM CO cO cO CM CO CM CM CM 1— ( 00 O O d> o o o o o o o CM CM 0^ x" oo' —1 X oo 1—1 oT X Z Z Z Z - — ' ' — ' ' — ' ■ — ' *o o •O CM- *0 CM- *0 o *0 CM- x> o* o be D rJ ^ < y c .2^ s t 'S U ^ ^ a: 5 160 Tongue joins Tongue joins mirror Mirror mirror Mirror Length of Length of Continuous outer web Outer Length of Continuous inner web Inner black sub- white mirror white outer enclosed web white mirror white inner enclosed web terminal bar on R.E. Johnstone ^ 04 CO tJ- o CO 04 04 04 d) d> Tf r-* r~( 04 o o CO 04 00 CO CO CO cO 00 ic r- lO o o o 04 6 o ZZ ZZ E r-- CO z. *o O* *o o *0 CK *0 o- *o 0+ c JD 3 V) < V) o 3 -o C % B 3 < 4-> >> o s: 3 < Vi V B 3 < B 4 -) s C 02 0 0 u 5 o Q c k. o V o o 0 a> s: X v • in -iS ”0 oj 'U IM u 0 •£ *c3 ^ b ^ c nq j- " s I 5 “K ■^0 3 ■a c ■fc V H TD 2 '2 0 T3 S 2 ^ is*' < 4> 2 B 3 c T3 o 13 S C v g 3 E V o w ca 1 c ^ o U > 0 0 2 C a> Wl c 13 U o N •C X v> ^ S « t pa IM V ? £ «j tn u 2 2= V5 U3 ^ 1 5 o Z & u Z u Z to r- OO O o :5 CM CO 162 R.E. Johnstone Table 5 Measurements (mm) showing the extent of the white mirrors of Larus hartlaubii and Larus cirrocephalus. First primary Length of white mirror outer web Mirror outer web enclosed by black Length of white mirror inner web Mirror inner web enclosed by black Larus hartlaubii 6 {N9) 25-50 (40) 26-41 (34) 5 9 (N9) 31-45 (43) 21-38 (29) 7 Larus c. cirrocephalus d (N3) 36-44 (40) 24-38 (33) 3 9 {N4) 23-49 (40) 11-37 (27) 4 Larus c. poiocephalus 6 (N13) 20-40 (32) 17-29 (23) 12 9 (N9) 22-39 (30) 16-26 (21) 9 Second primary Larus hartlaubii d (N9) 21-39 (27) 19-33 (26) 6 9 (N9) 13-34 (26) 11-33 (22) 8 Larus c. cirrocephalus d (N3) 23-39 (31) 9-34 (23) 2 9 (N3) 31-37 (35) 22-30 (26) Larus c. poiocephalus d (N12) 16-32 (24) 2 7-28 (20) 10 9 (N9) 9-27 (20) 1 12-22 (17) 4 Table 6 Measurements (mm) of Larus hartlaubii, Larus cirrocephalus cirrocephalus and Larus cirrocephalus poiocephalus with means and sample size in parentheses. Wing Culmen length Culmen depth Tail Tarsus Larus hartlaubii 6 (N9) 274-291 (283) 45.0-50.5 (47.8) 8.2-9.3 (8.8) 104-114 (108) 39-44 (42) 9 (N9) 265-293 (276) 42.0-47.0 (44.5) 7.5-9.4 (8.0) 102-112 (106) 38-43 (40) Larus c. cirrocephalus d (N4) 313-338 (325) 49.5-56.0 (54.1) 9.9-11.1 (10.5) 116-134 (125) 50-55 (51) 9 (N8) 300-328 (312) 48.0-53.5 (49.9) 8.9-9.4 (9.2) 110-132 (118) 44-50 (48) Larus c. poiocephalus d (N18) 283-330 (313) 46.5-57.5 (52.1) 7.9-10.7 (9.5) 102-124 (117) 44-51 (47) 9 (N13) 290-317 (302) 45.0-51.0 (47.4) 8.0-9.9 (8.8) 111-120 (114) 40-48 (44) Discussion Most of the variation within Australian Silver Gulls is clinal, e.g. increasing white on the primaries from Queensland to Tasmania and from Western Australia to Tasmania. As expected the dine down the east coast is not smooth due to large breaks in the breeding range. The situation is also complicated by long-distance movements of many south-eastern birds and the lack of specimens from breeding localities. In contrast, the Western Australian populations are far more sedentary 163 Distribution, Status and Variation of the Silver Gull and show a more gradual increase in mirror size and number from north to south (two mirrors in the north, three in the south). The dine along the south coast would also be relatively smooth except for the peculiar population in the Archi- pelago of the Recherche; these birds are small, have only two mirrors or a greatly reduced third mirror, and in life have a rosy tinge to the under parts. Using a combination of measurements and coloration the region of origin of most suspected vagrants and migrants can be detected. In Australia, Tasmanian birds appear to be the most distinctive and the name giinni is available for them. However, the unnamed populations from the Archipelago of the Recherche and from northern Western Australia (Carnarvon to Kimberley) have diverged at least as far. Nevertheless in view of the predominantly clinal variation within Australia it seems pointless to add names to those already available. The Lotus cirrocephalus species group is made up of four species: L. cirro- cephalus, L. hartlaubii, L. novaehollandiae and L. bulleri, and is entirely restrict- ed to the Southern Hemisphere. The evolution of this group could be explained by the following steps: (1) an early form of Southern American cirrocephalus invaded South Africa and in isolation evolved into L. hartlaubii; (2) an early form of hartlaubii spread east to Australia and New Zealand, evolving into novae- hollandiae in the first region and bulleri in the second; (3) Australian novae- hollandiae subsequently re-invaded New Zealand and evolved into scopulinus; (4) recently Southern American cirrocephalus re-invaded Africa and evolved into poiocephalus. Conclusion The following nomenclature for iheLarw^- cirrocephalus species group is proposed: Larus cirrocephalus cirrocephalus Vieillot of South America, Larus cirrocephalus poiocephalus Swainson of Africa, Larus hartlaubii Bruch of south-west Africa, Larus novaehollandiae novaehollandiae Stephens of Australia and Tasmania, Larus novaehollandiae forsteri (Mathews) of New Caledonia, Larus novaehollandiae scopulinus Forster of New Zealand, and Larus bulleri Hutton of New Zealand. References Buller, W.L. (1888). A History of the Birds of New Zealand, 2nd ed; (Privately published, London). Garrick, R., Wheeler, W.R. and Murray, M.D. (1957). Seasonal dispersal and mortality in the Silver Gull Larus novaehollandiae Stephens, and Crested Tern Sterna hergii Lichtenstein. CSIRO Wildl. Res. 2: 116T44. Glancey, P.A. (1964). The Birds of Natal and Zululand. (Oliver and Boyd: Edinburgh and London). 164 R.E. Johnstone Condon, H.T. (1975). Checklist of the Birds of Australia, Part 1, Non-passerines. (Melbourne: RAOU). Dwight, J. (1925). The Gulls (Laridae) of the World. Bull. Amer. Mus. Nat. Hist 52; 63-401. Ornithological Society of New Zealand (The Checklist Committee, F.C. Kinsky, convener), (1970). Annotated Checklist of the Birds of New Zealand including the Birds of the Ross Dependency. (A.H. and A.W. Reed: Wellington). Mills, J.A. (1969). The distribution of breeding Red-billed Gull colonies in New Zealand in relation to areas of plankton enrichment. Notornis 16: 180-186. Nicholls, C.A. (1964). Double-broodedness in the Silver Gull Larus novaehollandiae. West. Aust Nat 9: 73-77. Oliver, W.R.B. (1930). New Zealand Birds. (Fine Arts [NZ] Ltd: Wellington). Peters, J.L, (1934). Checklist of Birds of the World. Vol. II. (Harvard Univ, Press: Cambridge, Mass.). Serventy, D.L., Serventy, V. and Warham, J. (1971). The Handbook of Australian Seabirds. (A.H. and A.W. Reed: Sydney). Sinclair, J.C. (1977). Interbreeding of Grey-headed and Hartlaub’s Gulls. Bokmakierie 29: 70-71. Storr, G.M, (1973). List of Queensland Birds. West. Aust Mus. Spec. Pubis No. 5. Storr, G.M. (1977). Birds of the Northern Territory. West. Aust Mus. Spec. Pubis No. 7. Storr, G.M. (1980). Birds of the Kimberley Division, Western Australia. West Aust Mus. Spec. Pubis No. 11. Tinbergen, N. and Broekhuysen, G.J. (1954). On the threat and courtship behaviour of Hart- laub’s Gull, Ostrich 25: 50-61. Wheeler, W.R. and Watson, I. (1963). The Silver Gull Larus novaehollandiae Stephens.£’mw 63: 99-173. Wooller, R.D. and Dunlop, J.N. (1979). Multiple laying by the Silver Gull Larus novae- hollandiae Stephens on Carnac Island, Western Australia. Aust. Wild. Res. 6: 325-335. Addendum On 13 October 1982 a female Silver Gull was collected from a flock of 200 at Wyndham, east Kimberley, Western Australia. It has three primary mirrors and the bill was red, slightly darker at the tip. In wing pattern and bill coloration it matches birds from the Northern Territory rather than west Kimberley. Several immature gulls still begging for food were also seen at Wyndham; so it appears that the species breeds in this area. Received 15 April 1982 Accepted 13 July 1982 Published 13 December 1982 165 Rec. West. Aust. Mus. 1982, 10 (2): 167-197 Taxonomy and Evolution of Living Species of Breynia (Echinoidea: Spatangoida) from Australia KJ. McNamara* Abstract Three living species of Breynia are described: B, australasiae (Leach, 1815) from eastern Austriia; B. desorii Gray, 1851 from western and northern Australia; and B. neanika sp. nov. which is described on the basis of specimens from north-eastern Australia and from the Arafura Sea. Speciation is shown to relate to variation in rates of morphological development, B, desorii undergoing most morphologiceil change during its ontogeny, B. neanika the least. The post-larval ontogeny of Breynia is described in detail for the first time. A revised key for the living species of Breynia is presented. Introduction The spatangoid echinoid Breynia, although being one of the commonest of the larger heart urchins around the coast of Australia, has been little studied. Indeed, surprisingly few detailed taxonomic studies have been carried out on any of the living irregular echinoids of Australia apart from recent papers by McNamara and Philip (1980) and McNamara (1982a). These authors established that in living Australian schizasterid echinoids it is important to examine large collections as there is often a high degree of intraspecific variation. This is the case with eastern and Western Australian forms of Breynia in which a certain amount of confusion over the taxonomic status has occurred during the last hundred years. Gray (1851, 1855) originally distinguished the Western Australian B. desorii, from the eastern species, B. australasiae (Leach, 1815). However, many later workers (Agassiz 1872-74; Studer 1880; H.L. Clark 1914, 1917, 1925, 1938, 1946; Alexander 1914; Mortensen 1918) synonymized B. desorii with austral- asiae without having examined adequate material. Some workers (Mortensen 1951; James 1966; A.M. Clark and Rowe 1971) have considered B. desorii to be specifically distinct from B, australasiae, although for different reasons. The aim of this paper is to describe the western form of Breynia in detail, on the basis of a large collection from Norbill Bay, Rosemary Island, Western Aus- tralia, in order to re-evaluate its taxonomic status. The range of variation between * Department of Palaeontology, Western Australian Museum, Francis Street, Perth, Western Australia 6000. 167 Taxonomy of Breynia adults, both within and between populations, and the ontogeny of Breynia are described, and the relationship between ontogeny and phylogeny within the genus is assessed. Furthermore, examination of the eastern Australian form, largely on the basis of material from Lord Howe Island, and a collection recently made from the Arafura Sea by the CSIRO vessel Soela, reveals the presence of two species, both distinct from the western species; one of these is described as new. Specimens used in this study are housed in the Western Australian Museum (WAM), Australian Museum (AM), Queensland Museum (QM) and British Museum (Natural History) (BM). In species descriptions ‘percentage of test length’ is abbreviated to % TL. All measurements made are self-evident, except for width of internal fascicle; this is measured across the apical system. Key to Living Species oi Breynia 1 Test with flattened aboral surface; short petals 2 Test with vaulted aboral surface; long petals 3 2 Internal fasciole narrow; periproct oval B. neanika Internal fasciole narrow; periproct circular B. vredenburgi 3 Many primary tubercles 4 Few primary tubercles B. australasiae 4 Internal fasciole long; plastron long B. desorii Internal fasciole short; plastron short B. elegans Systematics Order Spatangoida Claus, 1876 Family Loveniidae Lambert, 1905 Genus Breynia Desor, 1847 Type Species Spatangus australasiae Leach, 1815: 68. Breynia australasiae (Leach, 1815) Figures 1, 2 Spatangus australasiae Leach, 1815: 68, PL 82. Breynia australasiae — Mortensen 1951: 132-139, PL 10, figs 1-5, PL 12, figs 8, 9, 11, 13, PL 14, figs 3-5, PL 49, figs 1-5, 20, 25-30, 32; with full synonymy. Breynia desorii Gray, 1851; — Mortensen 1951, PL 12, fig. 7, PL 14, figs 6, 7. 168 K.J. McNamara Diagnosis (emended herein) Aborally test bears up to 15 primary spines and tubercles in each of the anterior interambulacra, 2 and 3, and up to 25 in 1 and 4. Peripetalous fascicle only close to ambitus anteriorly; broad and relatively short internal fascicle. Peristome broad and sunken; periproct large. Generally between 5 and 6 subanal pore pairs. Remarks This species has been described by a number of authors (see Mortensen [1951] for detailed synonymy) and in detail by Mortensen (1951). Further general description would be superfluous. However, a number of further observations can be made, amplifying, in particular, characters hitherto not considered to be of great taxonomic significance, but in this study shown to be important in charac- terizing the species. Compared with other species of Breynia the peristome is wide (Figure 5) occupying up to 16% TL. In large adults the labrum may project slightly anterior- ly giving a slight lunate shape to the peristome, in contrast to the semicircular shape in juveniles and young adults. Mortensen (1951) characterized B. australasiae as having a labrum which reaches only to the second ambulacral plate, although rarely just approaching the third. Examination of almost 100 specimens has revealed that the labrum reaches the second ambulacral plate in only 40% of the specimens; in the remaining 60% it reaches the third. The periproct is large for the genus, its long axis reaching almost 15% TL. It is enclosed by plates 4-8 of interambulacrum 5. Although Mortensen (1951) observed no more than 6 pore pairs either side of the mid-line within the subanal fasciole, specimens are known which have 7 ; others have as few as 4. Generally there are 5 or 6. The internal fasciole is relatively broad (Figures 1, 7) occupying generally 20- 24% TL, but up to almost 30% in some large individuals. It is relatively short (Figure 6), between 36 and 44% TL. As a result of the peripetalous fasciole being set well in from the ambitus anterolaterally, the width across the anterior petals is only about 56-60% TL (Figure 9). The smaller interambulacral area within the peri- petalous fasciole results in fewer primary tubercles than in species in which the peripetalous fasciole approaches closer to the ambitus (Figure 7). In anterior interambulacra 2 and 3 there are no more than 15 primary tubercles, while in the lateral interambulacra 1 and 4 there are up to 25. Rate of production of primary tubercles through ontogeny is slow compared with B. desorii (Figure 8). Examination of an ontogenetic suite shows similar morphological development to B. desorii (described below), although rate of development is reduced. The significance of this is discussed below in the section dealing with the phylo- genetic relationships between the Australian species. Opening of genital pores is thought to correspond to onset of maturity. This generally occurs between test lengths of 38 and 44 mm. 169 Taxonomy oi Breynia Figure 1 Breynia australsasiae (Leach, 1815); AM G12604, from Lord Howe Island; (A) aboral view; (B) lateral view; both x 1. 170 K.J. McNamara Figure 2 Breynia australasiae (Leach, 1815); AM G12604, from Lord Howe Island; (A) adoral view; (B) posterior view; both x 1. 171 Taxonomy of Breynia Distribution H.L. Clark (1946) and Mortensen (1951), following Tenison Woods (1878), recorded B. australasiae along the entire Queensland coast from ‘Cape York to Port Jackson’. It does not appear, however, to be common from the Queensland coast on the basis of materid held in Museum collections. Single specimens are known from East Tongue Reef, north of Cairns; Lindeman Island; Swain Reef and Turtle Head Island, Cape York (all Australian Museum specimens); north Wistari Reef; Bowen; Yepoon (all Queensland Museum specimens); and Hervey Bay (Endean 1961). Four specimens from Torres Strait are in the British Museum (Natural History) collection. B. aiistralasiae is particularly common around Lord Howe Island in the Tasman Sea, where it burrows in sand in shallow water depths. Although Tenison Woods (1878) recorded the species from Port Jackson there are no specimens from N.S.W. in the collections of the Australian Museum or any other collections examined. There is no I'ecord of B. australasiae from either the Victorian or South Australian coasts (Figure 14). Although Mortensen (1951: 137) considered B. australasiae to be restricted to the eastern coast of Australia and B. desorii to the west, three of the specimens he figured from eastern Australia, one from Bowen (Mortensen 1951, PI. 12, fig. 7) and two from ‘Queensland’ (Mortensen 1951, PL 14, figs 6, 7) he called B. desorii, contrary to his opinion on their distribution expressed in the text. These specimens appear to be B. australasiae. Breynia desorii Gray, 1851 Figures 3,4, 10 Breynia desorii Gray, 1851: 131; — Gray 1855: 46; — Mortensen 1951: 129-131, 139-141, PI. 11, figs 3-5, PI. 12, figs 1, 12, non 7, PI. 14, figs 1, 2, non 6, 7; - James 1966: 79; - A.M. Clark and Rowe 1971 : 146-149, 165, fig. 81b. Breynia australasiae — Agassiz 1872-74: 578 (pars.); — Studer 1880; 881; Alexander 1914. 112; - H.L. Clark 1914: 169;- H.L. Clark 1917: 250 (pars.); - Mortensen 1918: 20, PI. 5, figs 2, 18, 19, 22; - H.L. Clark 1925: 228; - H.L. Clark 1938: 438^39; - H.L. Clark 1946: 381. ?Breynia australasiae var. aroensis Currie, 1924: 63-66, PI. 4. Holotype British Museum (Natural History) specimen 39.6.10.36, a dry test lacking spines; probably female. Purchased from a Mr Turner prior to 1839 when it was registered. Miss A.M. Clark of the British Museum informs me (litt. comm. 21.10.80) that the specimen is the only one in the collections registered as coming from ‘Swan River, Western Australia’, the type locality, even though Gray (1851: 131) recorded several specimens being in existence at the time of his description of the species. It must therefore be regarded as the holotype as it is the sole remain- ing specimen from the type series. It had earlier been considered by A.M. Clark and Rowe (1971) that all of Gray’s original series was lost. 172 K.J. McNamara Figure 3 Breynia desorii Gray, 1851; BM 39.6.10.36, holotype, from ‘Swan River, Western Australia’; (A) aboral view; (B) lateral view; both x 1. 173 Taxonomy of Breynia Figure 4 Breynia desorii Gray, 1851; BM 39.6.10.36, holotype, from ‘Swan River, Western Australia'; (A) adoral view; (B) posterior view; both x 1. 174 K.J. McNamara Figure 5 Plot of test length against peristome width for B. australasiae (triangles), B. desorii (circles) and 5. neanika (stars). 175 Taxonomy of Breynia Figure 6 120 * 100 * 80 - U) c 'andra State Forest (62239); North Tarin Rock Reserve (44437); Lake Grace Reserve (43825-6); Kukerin (6102); 29 km SE Newdegate (47607-8); Dongolocking Reserve (49620-3, 49728-9); ‘8 km SW Collie’ (19581); Lake Chinocup (43466-70); 27 km E Pingrup (39861-3,45299); Lake Magenta Reserve (39950); 32 km W Ravensthorpe (44870) and 26 km W (44865); Jerramungup (14493); 10 km SE Ongerup (42622); Culham Inlet (78181); 4 km W Hopetoun (56064); East Mt Barren (78228); near Woolbernup (41147); Dempster Inlet (38998-9); lower Gairdner River (52090); ‘Cape Leeuwin’ (299, 12784, 29984). Eucla Division CW.A.) Near McDermid Rock (65262, 65269, 65287, 65291, 74240, 74293); between McDermid Rock and Lake Cronin (65278); North Ironcap (71172): near Lake Cronin (65097, 65099, 65100, 65160, 68007, 68028, 68032, 68037, 68052, 74104, 74218); 8 km WNW Forrestania (71180); Frazer Range (54770-1); 30 km NNE Balladonia Hotel (46606); 20 km E Jyndabinbin Rockhole (62360) and 27 km E (62359); 38-40 km ESE Norseman (57920, 57951); 42 km ENE Clear Streak Well (59756) and 8 km N (59587-8, 59861-2); Charlina Rock (57967) and 18 km NNE (59871); 10 km SE Mt Newmont (59757-8); Frank Hann National Park (78342); Peak Charles (56884); Salmon Gums (30790) and 20 km E (22653) and 60 km E (62361, 62364-5); Esperance (19587) and 32 km E (21994) Cape LeGrand National Park (41945, 67753). Northern Territory Docker River (45201, 45206); 16 km E Lasseters Cave (34197-8); Armstrong Creek (34182); Ayers Rock (46634-5). South Australia 250 km N Cook (34536) and 160 km N (31858-9); Everard (24511). 207 Revision of the Bearded Dragons with Notes on Amphibolurus Pogona minor minima (Loveridge, 1933) Amphibolurus barbatus minimus Loveridge, 1933, Proc. New Engl. zool. Club 13: 69. West Wallabi I., W.A. Diagnosis A small, relatively slender, long-limbed Pogona, similar in all respects to P. minor minor of the opposite mainland except for its longer appendages and more numerous subdigital lamellae. Description Snout-vent length (mm); 39-115 (N 82, mean 90.9). Length of appendages (% SVL): foreleg 42-54 (N 81, mean 47.0), hindleg 60-77 (N 81, mean 69.2), tail 175-246 (N 74, mean 203). Upper labials 12-17 (N 81, mean 14.7). Lamellae under fourth toe 20-31 (N 80, mean 25.3). Femoral pores 2-6 (N 68, mean 4.1) on each side. Pre-anal pores 2-5 (N 68, mean 2.6) on each side. Scalation and coloration as in southern P. m. minor. Figure 3 A Pogona minor minima from North I., Houtman Abrolhos, W.A., photographed by P. Griffin. 208 G.M. Storr Distribution The larger and more northerly islands (North, East Wallabi and West Wallabi) of the Houtman Abrolhos, off the midwest coast of Western Australia (see Figure !)• Material South-West Division (W.A.) North L (47815-8); East Wallabi I. (19547-62, 21846, 30183-9, 30201,46554,47826-7, 78249); West Wallabi I. (19501-46, 29496-7, 57996). Pogotia minor mitchelli (Badham, 1976) Amphibolurus mitchelli Badham, 1976, Aust. J. Zool. 24: 435. Derby, W.A. Diagnosis A moderately large Pogona, distinguishable from P. m. minor by its brighter coloration (more brownish, less greyish), greater size, wider head, stronger contrast between scales in front of and behind transverse series of occipital spines, and lack of longitudinal series of spines on side of nape. Description Snout-vent length (mm): 37-163 (N 63, mean 109.6). Length of appendages (% SVL): foreleg 36 46 (N 62, mean 39.6), hindleg 49-64 (N 60, mean 54.6), tail 143-205 (N 58, mean 170). Upper labials 14-19 (N 53, mean 15.9). Lamellae under fourth toe 17-25 (N 54, mean 20.8). Femoral pores 4-6 (N 49, mean 4.3) on each side. Pre-anal pores 2-4 (N 49, mean 2.5) on each side. Transverse series of occipital spines meeting longitudinal series of supra-auricular spines at less than a right angle, and only curving forwards near midline; spines usually stronger and more contiguous than in other subspecies of P. minor. Scales behind occipital spines smaller than in other subspecies, with a greater step down from plane of occipital scales to plane of nuchal scales. Longitudinal series of nuchal spines of other subspecies reduced in mitchelli (as in northernmost P. minor minor) to a small cluster of spines continuous with transverse series of nuchal spines. Cluster of spines on scapular fold larger than in other subspecies. Lateral scales at rear of throat (continuous with ‘beard’) larger and more spinose than in other subspecies. Dorsal ground colour yellowish-brown, orange -brown or greyish-brown. Dorsal and ventral pattern in juveniles not so marked as in other subspecies and seldom persisting in adults. Distribution Arid and semi-arid north-western Western Australia, i.e. western Kimberley and northern Pilbara south to the Port Hedland and Marble Bar districts (see Figure 1). 209 Revision of the Bearded Dragons with Notes on Amphibolurus Figure 4 APogona minor mitchelli, photographed by P. Griffin. Remarks Badham (1976) treated mitchelli as a full species. However it hybridizes with P. m. minor in southern Pilbara, but the narrowness of the hybrid zone (less than 200 km wide) permits the recognition of two mainland subspecies of P. minor. Material Kimberley Division (W.A.) Drysdale River National Park (15°16'S, 126°43'E) (50696); Lombadina (46416) and 6 km S (60911); Martins Well {16^34'S, 122°5rE) (58527. 28532); Pender Bay (58524-5); Coulomb Point (40246-7, 58526); Point Torment (60916); Inglis Gap (27727); Derby (15182-4, 15823, 26828-30, 31037, 33624, 49986) and 24 km SSE (32166, 32189); 10 km WNW Mt North (70527) and 3 kra SSW (70534); 5 km NNW Mt Percy (70655-6); 24 km E Deep Creek (68983); Broome (14109, 14135a-2, 58857-8); Mt Anderson (32196-8); Edgar Ranges (53999, 54011, 54070); Injudinah Creek (27723); LaGrange (3445, 13067, 46489); Cape Bossut (40536). North-West Division (W^A.) DeGrey (2116, 73038); Great Northern Highway, 63 km ENE of DeGrey River crossing (46068); 25 km ESE Port Hedland (46498); Mundabullangana (19556) and 23 km SE (19377); 70 km N Marble Bar (40860); Mt Edgar (13066, 45758-60). 210 G.M. Storr Pogona microlepidota (Glauert, 1952) Amphibolurus barbatus microlepidotus Glauert, 1952, West. Aust. Nat. 3: 168. Drysdale River Mission [Pago] , W.A. Diagnosis A large Pogona with relatively small and narrow head and wide and strongly depressed body. Further distinguishable from P. minor by its 3-5 rows of large spines (rather than one row of small spines) along dorsoventral angle of body. Description Snout-vent length (mm); 93-180 (N 13, mean 148.7). Length of appendages (% SVL); foreleg 38-45 (N 13, mean 42.5), hindleg 59-71 (N 13, mean 65.7), tail 181-212 (N 13, mean 198). Upper labials 15-19 (N 13, mean 17.0). Lamellae under fourth toe 20-23 (N 13, mean 21.3). Femoral pores 3-5 (N 13, mean 3.8) on each side. Pre-anal pores 2-3 (N 13, mean 2.1) on each side. Figure 5 A Pogona microlepidota from Crystal Head, W.A., photographed by R.E. Johnstone. 211 Revision of the Bearded Dragons with Notes on Amphibolurus Transverse series of occipital spines well separated from longitudinal series of high supra- auricular spines; a single central occipital spine and 1-3 outer spines moderately large, remainder small. At back of nape a small circular cluster of spines close to midline; further out (in same position as in P. minor mitchelli) a larger, higher, oblique cluster of spines narrowly separated from transverse series of nuchal spines (posterior sector of series of high spines curving around rear of head behind ear). ‘Beard’ a little better developed than inP. minor and extending further on to ventral surface of throat. Small cluster of spines on scapular fold; parallel to and behind fold a short series of spines. Dorsal ground colour dull yellowish-brown, becoming more greyish-brown towards middle of back. Little evidence of dorsal pattern apart from a tendency for yellowish brown and greyish-brown to be disposed in vague, narrow, alter- nating transverse bands of each colour. Side of rear of head (especially streak through temple) and side of neck blackish-grey. Tail greyish-brown, narrowly banded with pale dull yellowish-brown. Lower surfaces brownish-white occasion- ally spotted with grey. Distribution Far north of Western Australia, i.e. subhumid north-west Kimberley from Napier Broome Bay south-west to the Prince Regent River (see Figure 1). Material Kimberley Division (W.A.) Pago (951-2, syntypes); Kalumburu (27728-9, 34075, 74943); near Crystal Head (14®30'S, 125®47T) (43028, 56232); Mitchell Plateau (14°53'S, 125049T) (44258); Bigge I. (57108); Prince Regent River National Park (46847, 46962, 47246). Pogona nullarbor (Badham, 1976) Amphibolurus nullarbor Badham, 1976, Aust. J. Zool. 24: 440. 16 km NW Naretha, W.A. Diagnosis A moderately large, short-snouted, short-tailed Pogona with wide, strongly depressed body. Further distinguishable from P. minor by 3-7 rows of large spines (rather than a single row of small spines) along dorsoventral angle of body, pale narrow transverse dorsal bands, and smooth mucronate (rather than keeled) ventrals. Description Snout-vent length (mm): 51-141 (N 16, mean 121.7). Length of appendages (% SVL): foreleg 34-43 (N 16, mean 38.7), hindleg 51-62 (N 16, mean 56.6), tail 122-160 (N 15, mean 138). Upper labials 14-20 (N 16, mean 16.7). Lamellae 212 G.M. Storr under fourth toe 17-23 (N 16, mean 19.3). Femoral pores 3-5 (N 15, mean 3.9) on each side. Pre-anal pores 2-3 (N 15, mean 2.3) on each side. Transverse series of occipital spines curving forwards towards midline, its alignment forming acute angle with that of short series of low supra-auricular spines; occipital spines small and not contiguous. On side of nape a longitudinal series of 3-6 spines (as in southern and western P. minor but weaker), separated from series of high spines curving around rear of head behind ear. ‘Beard’ weak but extending (as a band of narrow, elongate, strongly mucronate scales) right across throat. Cluster of spines on scapular fold small. Dorsal ground colour reddish-brown, orange -brown or greyish-brown. Six or 7 narrow creamy-white cross-bands on neck and back; pale bands on tail wider and darker. Throat greyish or whitish, marked with 3 or 4 hollow chevrons, the smaller inside the larger. Ventral surfaces as in P. minor, i.e. greyish in juveniles with ocelli (longitudinally elongate, whitish spots laterally edged with greyish brown), whitish in adults and marked with short longitudinal greyish-brown streaks (sides of obsolete ocelli). Figure 6 A Pogona nullarbor from Madura, W.A., photographed by R.E. Johnstone. 213 Revision of the Bearded Dragons with Notes on Amphibolurus Distribution The Nullarbor Plain of arid south-eastern Western Australia and western South Australia (see Figure 1). Material Eastern Division (W^A.) Kanandah (37789); Naretha (19592, 48172) and 16 km NW (29667, holotype). Eucla Division {IV. A.) 112 km NNE Rawiinna (31964-5); 16 km W Seemore Downs (39055); 16 km S Loongana (29486); Forrest (16888, 16896) and 24 km S (41632); WHson Bluff (28127); 40 km SW Eucla (66449); 33 km NE Madura (28901); Cocklebiddy (24655). South Australia 75 km S Cook (31620). References Badham, J.A. (1976). The Amphibolurus harbatus species-group (Lacertilia: Agamidae), Aust. J. ZooL 24: 423-443. Houston, T.F. (1978), Dragon Lizards and Goannas of South Australia. (S. Aust. Mus.: Adelaide). Storr, G.M. (1965). The Amphibolurus maculatus species-group (Lacertilia, Agamidae) in Western Australia./. Proc. R. Soc. West. Aust. 48; 45-54. Storr, G.M. (1966). The Amphibolurus reticulatus species-group (Lacertilia, Agamidae) in Western Australia./. Proc. R. Soc. West. Aust. 49: 17-25. Storr, G.M. (1977). The Amphibolurus adelaidensis species-group (Lacertilia, Agamidae) in Western Australia. Rec. West. Aust. Mus. 5: 73-81. Witten, G.J. (1982). Phyletic groups within the family Agamidae (Reptilia: Lacertilia) in Australia. In: Evolution of the Flora and Fauna of Arid Australia (Eds W.R. Barker and P.J.M. Greenslade): 225-228. (Peacock Publications). Received 15 April 1982 Accepted 9 September 1982 Published 13 December 1982 214 GUIDE TO AUTHORS Subject Matter Reviews and papers reporting results of research in all branches of natural science and human studies will be considered for publication. However, emphasis is placed on studies pertaining to Western Australia. Material must be original and not have been published elsewhere. Presentation Authors are advised to follow the layout and style in the most recent issue of the Rec. West. Aust. Mus. including headings, tables, illustrations and references. The title should be concise, informative and contain key words necessary' for retrieval by modem searching techniques. Names of new taxa must not be included. An abridged title (not exceeding 50 letter spaces) should be included for use as a running head. An abstract must be given, summarizing the scope of the work and principal findings. It should normally not exceed 2% of the paper and should be suitable for reprinting in reference periodicals. Contrary to Recommendation 23 of the International Code of Zoological Nomen- clature it may include names of new taxa. Footnotes are to be avoided, except in papers dealing with historical subjects. The International System of units should be used. Numbers should be spelled out from one to nine in descriptive text; figures used for 10 or more. For associated groups, figures should be used consistently, e.g. 5 to 10, not five to 10. Spelling should follow the Concise Oxford Dictionary. Systematic papers must conform with the International Codes of Botanical and Zoological Nomenclature and, as far as possible, with their recommendations. Synonymies should be given in the short form (taxon, author, date, page) and the full reference cited at the end of the paper. Manuscripts The original and two copies of manuscripts should be submitted to the Publications Officer, Western Australian Museum, Francis Street, Perth, Western Australia 6000. They must be in double-spaced typescript on A4 sheets. All margins should be at least 30 mm wide. Tables plus headings and legends to illustrations should be typed on separate pages. The desired positions for insertion of tables and illustrations in the text should be indicated in pencil. Tables should be numbered consecutively, have headings which make them understandable without reference to the text, and be referred to in the text. Drawings must be suitable for direct photographic reproduction. Photographs must be sub- mitted as high quality black and white prints. The original and two copies (the latter at desired reproduction size) of all illustrations are required with figure numbers lightly pencilled on the back. Lettering on them must be of appropriate size for reduction, if this will be necessary. Scale must be indicated, preferably on the illustration. All illustrations, whether line drawings or photographs, should be numbered in sequence and referred to as ‘Figure/s’ in the text. Each must have a brief, fully self-explanatory caption. In papers dealing with historical subjects references may be cited as footnotes. In all other papers references must be cited in the text by author and date and all must be listed alpha- betically at the end of the paper. The names of journals are abbreviated according to World List of Scientific Periodicals. The use of ‘unpublished data’ or ‘personal communication’ is dis- couraged. Processing Papers are reviewed by at least two referees and acceptance or rejection is then decided by an editorial committee. The senior author is sent two sets of galley proofs (one to be retained) and one set of page proofs which must be returned promptly to the Publications Officer after correction. The senior author will receive fifty free offprints of the paper. Additional offprints can be purchased and should be ordered in advance through the Publications Officer. Contents Allen, Gerald R. and Boeseman, M. A Collection of Freshwater Fishes from Western New Guinea with Descriptions of Two New Species (Gobiidae and Eleptridae) Allen, Gerald R. A New Species of Freshwater Rainbowfish (Melanotaeniidae) from Misool Island, Indonesia Bindon, P. and Lofgren, M. Walled Rock Shelters and a Cached Spear in the Pilbara Region, Western Australia Hutchins, J.B- and Kuitcr, R.H. A New Species oi Acanthistius (Pisces: Serranidae) from Eastern Australia Johnstone, R.E. Distribution, Status and Variation of the Silver Gull Laras novae^ hollandiae Stephens, with Notes on the Larus cirrocephalus species-group McNamara, KJ. Taxonomy and Evolution of Living Species of Breynia (Echin- oidea: Spatangoida) from Australia Storr, G.M, Revision of the Bearded Dragons (Lacertilia: Agamidae) of Wes- tern Australia with Notes on the Dismemberment of the Genus Amphiholurus ■■ ■■ I A A- • m-'-r S AA ■ RECORDS OF THE WESTERN AUSTRALIAN Volume 10, Part 3, 1983 Records of the Western Australian Museum Editorial Committee Chairman P.F. Berry Natural Science Human Studies G.R. Allen KJ. McNamara I.M. Crawford GJ. Henderson M.E. Lofgren Publications Officer A.N. Browne The Records of the Western Australian Museum (Rec, West. Aust. Mus.) is published irregularly with up to four parts appearing each year. A series of supplements, usually devoted to recording the basic data and results of specific Museum studies, is also produced. The journal and supple- ments are available for sale or exchange, the current price per part being $5.00 plus postage. All but a few back issues are available and orders and enquiries should be addressed to the Book- shop, Western Australian Museum, Francis Street, Perth, W.A. 6000, Australia, Cover Male of the bee, Ctenocolletes rufescens sp. nov., endemic to southern Western Australia. Drawn by Jill Ruse. © Western Australian Museum, 1983 ISSN 0312 3162 Published by the Western Australian Museum, Francis Street, Perth, Western Australia 6000. Printed in Western Australia by Advance Press Pty Ltd. Rec. West. Aust. Mus. 1983 , 10 ( 3 ): 215-234 Herpetofauna of the Geraldton Region, Western Australia G.M. Storr* *, T.M.S. Hanlonf and J.N. DunlopJ Abstract The study area is located on the west coast of Western Australia approximately between latitudes 28® and 30®S; it extends for 30-50 km inland to about Ajana, Eradu, Strawberry and Eneabba, and for 50-70 km offshore to the Houtman Abrolhos. The climate ranges from semi-arid in the north to subhumid in the south, most of the rain falling in the cooler half of the year. Much of the land out- side reserves has been cleared for farming. Brief notes are given on the local distribution, relative abundance and habitat preferences of the 96 species of frogs, turtles, lizards and snakes recorded from the region. Some aspects of regional zoo- geography are discussed. Introduction The present paper is one of several dealing with the herpetofauna of the west coast of Western Australia. To the immediate north of the present region L.A. Smith (pers. comm.) has studied the herpetofauna of the Kalbarri National Park, and Storr and Harold (1980) that of the Zuytdorp area; to the immediate south Dell and Chapman (1977) listed the amphibians and reptiles of the Cockleshell Gully Reser\'e; and to the east Burbidge et al. (1978) described the fauna of the Wandana Nature Resei've near Yuna. Published information on the herpetofauna of the present region hitherto covered only the Houtman Abrolhos (Alexander 1922, Green 1972, O’Loughlin 1965, 1966 and 1969, and Storr 1960 and 1965) and the islands fui'ther south (Ford 1963). However, much information on the mainland was available in the registers of the Western Australian Museum, for the farming country between Northampton and Dongara has yielded many specimens since the turn of the century.* More recently naturalists have explored the sandplains, wetlands and * Department of Ornithology and Herpetology, Western Australian Museum, Francis Street, Perth, Western Australia 6000. t 38 Genesta Crescent, Dalkeith, Western Australia 6009. $ 12 Dame Pattie Drive, Burrendah, Western Australia 6154. * A collection (WAM R26304-26355) said to have been made by F.W. Pearson at Greenough between 1851 and 1881 is now believed to have come from the Eastern Goldfields. 215 Herpetofauna of the Geraldton Region G.M. Storr, T.M.S. Hanlon and J.N. Dunlop caves of the region (P'igure 1). To these data we have added those gathered by T.M.S. Hanlon in December 1980 and J.N. Dunlop in April 1982; the expenses of these surveys were met by generous grants from Mr and Mrs W.H. Butler to the Western Australian Museum. The Environment Mean annual rainfall ranges from 32 cm in the north-cast to 65 cm in the far south, 73-81% of it falling from May to September. Along the mainland coast, and still more in the Houtman Abrolhos, the climate is mild. In the interior summers are much hotter (mean daily maximum temperature c. 35^C in January and February), and frosts are occasionally reported in winter. The south and far north of the region consist essentially of a low sandy plateau rising gradually to 200-250 m in the east and terminating in the west in coastal dunes or low cliffs of aeolian limestone. The central and lower northern parts of the region are dissected by the Bowes, Chapman, Greenough and Irwin Rivers, and the sandy plateau has been reduced to remnants on the mesas between Geraldton and Northampton and to a much larger area between the Chapman and Greenough Rivers east of the ‘Greenough Flats’. The valleys of these rivers, especially their alluvial plains, have long been cleared for farming. In this central sector, unconsolidated white sand dunes stand between the red-soil plains and the sea. The sandy plateaux carry a rich assemblage of shrubs with scattered low trees of Banksia, Eucalyptus^ Xylomelum, Nuytsia etc. In the dissected areas of the north-east, thickets of jam [Acacia acuminata) and scattered York gums [Eucalypt- us loxophleba) occupy or used to occupy the undulating country below the residual sandplains; here the soils are generally heavy and often stony. Water- courses and lagoons are fringed with woodlands of Eucalyptus camaldulensis (replaced by E. rudis in the far south) and Casuarina obesa and thickets of Melaleuca. The seaward slopes of coastal dunes are sparsely covered with Spinifex longifolius and low shrubs; with increasing shelter from the sea the dunes support a higher and more varied heathland and thickets oi Acacia (especially A. rostel- lifera) suid Melaleuca. The numerous islands of the Houtman Abrolhos are low-lying and consist largely of flats of marine limestone and piles of debris thrown up by storms from the fringing coral reefs. Only the largest of the northern islands (North, East Wallabi and West Wallabi) contain extensive areas of sandy beaches and dunes. On most of the islands the vegetation is sparse and the Oora depauperate, but parts of East and West Wallabi are clothed in moderately rich scrub. There is no fresh surface water on any of the islands. The surrounding seas are considerably warmer than those along the mainland coast. For further information on the physiography, soils and vegetation of the region see Beard (1976). In the following list specimens in the R series of the Western Australian Museum are cited without prefix. 217 Herpetofauna of the Geraldton Region Annotated List Leptodactylidae Heleioporus albopunctatus Gray, 1841 Throughout the mainland, wherever fresh surface water is available in winter and early spring (collected on the Hutt, Bowes, Chapman, Greenough and Irwin Rivers and at White Peak and Stockyard Gully). Begins to call in late April. Heleioporus eyrei (Gray, 1845) Near-coastal areas north to the lower Greenough (collected at Greenough, 12 km E of Dongara 12 km E of Irwin, 16 km E of Coolimba and the Three Springs^ ). Uncommon. Swamps that fill in late autumn or early winter. Heleioporus psammophilus Lee and Main, 1954 One record from the lower Irwin River: three specimens (30516, 30527, 32993) collected at 2 km E of Irwin. Limnodynastes dorsalis (Gray, 1841) Throughout the mainland (collected on the Hutt River at 12 km ESE of Gregory, the Bowes River at 12 km W of Northampton, the Chapman River, the Irwin River at Mountain Bridge, and at the Three Springs, and heard at 2 km N of Northampton and at Lake Arrowsmith). Myobatrachus goiddii (Gray, 1841) One record from the eastern interior: a specimen collected under a stone in sandplain country at Eradu (Harrison 1927). Neobatrachus pelobatoides (Werner, 1914) Probably occurring throughout the mainland, but there is only one record, a specimen (8593) collected at White Peak. There are several records from immediately north and immediately south of our region. Neobatrachus sp. Two frogs (12479-80) collected at Greenough were registered in June 1957 as Heleioporus centralis'. These specimens cannot be found for checking, but they were more likely to belong to Neobatrachus sutor Main (only then being described) than to N. centralis (Parker). N. sutor has been collected just north of our region at Galena on the Murchison River. ^ The soak Three Springs is referred to in this paper as ‘the Three Springs’ to distinguish it from the town of Three Springs to the east of our area. 218 G.M. Storr, T.M.S. Hanlon and J.N. Dunlop Pseudophryne guentheri Boulenger, 1 882 Throughout the mainland (collected on the Hutt River north of Northampton, the Chapman River, the Greenough River at Ellendale and Newmarracarra, and the Irwin River, and at Stockyard Gully and the Three Springs). Moderately common. Requiring fresh surface water in late autumn and winter. Ranidella pseudinsignifera (Main, 1957) Throughout the mainland (collected at East Chapman, Bookara, Pell Bridge on the lower Irwin River, and Stockyard Gully, and heard at the Three Springs). Moderately common. Requiring fresh surface water in winter. Hylidae Litoria moorei (Copland, 1957) Greater part of mainland, north to the Hutt River (collected on the Hutt River at 12 km ESE of Gregory, on the Bowes River at 12 km W of Northampton, at an Tm-named cave near Eneabba’, and at Stockyard Gully and the Three Springs, and observed at Ellendale Pool on the Greenough River). Common. Requiring permanent freshwater river-pools or springs. Cheloniidae Chelonia mydas (Linnaeus, 1758) Common summer visitor to North L, Houtman Abrolhos, where it is said to nest (Storr 1960). Also recorded by Alexander (1922) from West Wallabi 1. Eretmochelys imbricata (Linnaeus, 1766) A specimen was received from Geraldton in 1902. Gekkonidae Crenadactylus ocellatus ocellatus (Gray, 1845) Regionally confined to the floutman Abrolhos. In the Wallabi Group un- common on the larger islands (East Wallabi and West Wallabi) and moderately common on the smellier islands (Seagidl and Tattler); in the Easter Group common on Rat, Hut and Helsinki Is; and in the Pelsaert Group common on Middle 1. Sheltering under limestone rocks and reef debris. Diplodactylus alboguttatus Werner, 1910 Two records: a specimen in the ARiseum of Comparative Zoology (Harvard) from Geraldton and one in the WAM from 4 km S of Geraldton. Common in the lateritic country immediately south of our region (Dell and Chapman 1977). 219 Herpetofauna of the Geraldton Region Diplodactylus granariensis Storr, 1979 One record: three specimens (27397-9) collected at 12 km E of the mouth of the Hutt River. Moderately common in the latcritic uplands immediately south of our region. Diplodactylus ornatus Gray, 1845 Coastal and near-coastal areas throughout the mainland. Uncommon. Has been collected in Acacia rostellifera thickets in white coastal dunes, in Melaleuca- Acacia scrub on near-coastal grey loamy sand, and in Xylomelum-Banksia scrub on inland sandplains. Diplodactylus poly ophthalmus Gunther, 1867 One record from far south: a specimen (78106) collected in sandplain close to a lateritic ridge 10 km S of Eneabba. Diplodactylus pulcher (Stcindachner, 1870) Two records of single specimens from far north-east (Ajana and Eradu). Diplodactylus spinigerus Gray, 1842 Common on all coasts, including East Wallabi and West Wallabi Is in the Houtman Abrolhos, in thickets of Acacia rostellifera, open Spinifex longifolius and Olearia axillaris, and scrubs of Melaleuca, Acacia and Casuarina. Also common on inland sandplains from Binnu south to Eneabba. Diplodactylus squarrosus Kluge, 1962 One record from far north-east: a specimen (24851) collected at Binnu. Diplodactylus strophurus (Dumchil and Bibron, 1836) One record from far north-east: a specimen (26006) collected at Binnu. Gehyra variegata (Dumeril and Bibron, 1836) Throughout the mainland and on some of the Houtman Abrolhos (East Wallabi and West Wallabi Is in the Wallabi Group, and Murray and Middle Is in the Pelsaert Group). Very common on the mainland in a wide variety of habitats; very common on Middle I. (O’Loughiin 1969); scarce on other islands. Heteronotia binoei (Gray, 1845) Northern half of mainland south to Greenough; also North and West Wallabi Is in the Houtman Abrolhos. Common on North I. and locally on the mainland but generally uncommon. A wide variety of habitats including coastal limestone. Nephrurus levis occidentalis Storr, 1963 Northern half of mainland south to Waggrakine and Wicherina. Rare. White sandplains with relatively open vegetation. 220 G.M. Storr, T.M.S. Hanlon and J.N. Dunlop Phyllodactylus marmoratus (Gray, 1845) On numerous islands in the Houtman Abrolhos: North I.; East Wallabi, West Wallabi, Tattler, Pelican and Seagull Is and islet between Seagull and East Wallabi Is in the Wallabi Group; Rat, Helsinki, Morley and Wooded Is in the Easter Group; and Gun I., three islets south of Gun L, and Murray, Shark, Basile and Pelsaert Is in the Pelsaert Group. Very common on Pelican, Tattler, Wooded, Gun and Murray Is and south end of Pelsaert L; scarce to moderately common on other islands. Sheltering under limestone and reef debris. Occurs on the mainland just south of our region (Cockleshell Gully). Phyllurus milii (Bory, 1825) Sparsely distributed on the mainland (collected on coast at Horrocks and Geraldton and near Greenough, and in the interior at a Stockyard Gully cave). Also the Houtman Abrolhos: East Wallabi, West Wallabi, Seagull and Pigeon Is in the Wallabi Group; and Gun I. in the Pelsaert Group. Very common on the islands; scarce on the mainland. Marine and aeolian limestone, including caves and sea- cliffs on the mainland. Pygopodidae Aclys concinna Kluge, 1974 One record from far south-west: a specimen (72983) found dead on road through scrubby heath on whitish sand over limestone 5 km E of Coolimba. Aprasia repens Fry, 1914 Southern and central interiors, north to Eradu. Scarce. Sandplains. Delma australis Kluge, 1974 Regionally known only from Rat I. in the Easter Group, Houtman Abrolhos, where it is common under slabs of limestone. In Western Australia mainly distribu- ted east of the Darling Range; on the west-coastal mainland only found in the country immediately south of Shark Bay, Delma fraseri Gray, 1831 Greater part of mainland north to Northampton. Uncommon. Delma gray a Smith, 1849 Southern part of mainland north to Beagle Point and inland to Eneabba; also West Wallabi I. in the Houtman Abrolhos; the specimen of 'Delma fraserP from East Wallabi I. (O’Loughlin 1966: 22) probably belonged to this species. Scarce. Coastal dunes and near-coastal sandplains. 221 Herpetofauna of the Geraldton Region Delma nasuta Kluge, 1974 One record from north-east: a specimen (47709) collected at Northampton. Delma tincta DeVis, 1888 Northern interior south to East Chapman and Eradu. Moderately common. Mainly on heavy red soils carrying jam [Acacia acuminata). Lialis burtonis Gray, 1835 Throughout the mainland; also East Wallabi and West Wallabi Is in the Hout- man Abrolhos. Moderately common. A wide variety of habitats but preferring coastal dunes and limestone. Pletholax gracilis gracilis Cope, 1864 One record from southern interior: a specimen (25071) collected at Eneabba. Pygopus lepidopodus (Lacepede, 1804) Throughout the mainland. Moderately common. A wide variety of habitats including near-coastal sandplains. Pygopus nigriceps nigriceps (Fischer, 1882) Northern interior south to the Irwin River. Known only from single specimens collected at Ajana, Binnu and Irwin. Probably restricted to heavier soils. Agamidae Amphibolurus adelaidensis adelaidensis (Gray, 1841) Throughout the mainland. Common. Sandplains with heath or mallee scrub. Amphibolurus inermis (DcVis, 1888) Apparently confined to two discrete areas: the far north-west around Balline, and the upper south-west between Dongara and Lake Arrowsmith. Uncommon. Lightly vegetated sandy loams. Amphibolurus maculatus maculatus (Gray, 1831) Mainly in the south (north to Cliff Head), where it is common on sandplains. Further north largely restricted to the white coastal dunes between Dongara and Geraldton and again at Gregory, where it is moderately common among Spinifex longifolius\ there is also a specimen from Ajana. A little east of our region, e.g. at 3 1 km E of Eneabba it is replaced by A. m. griseus Storr. Amphibolurus minor minimus Lovcridge, 1933 Endemic to the Houtman Abrolhos (North, East Wallabi and West Wallabi Is). Very common in all vegetated habitats but favouring sandy areas with clumps of 222 G.M. Storr, T.M.S. Hanlon and J.N. Dunlop Spinifex longifolius\ also among low halophytic shrubs on shell grit and in open mixed scrub on limestone. This subspecies differs from A. m. minor in its lesser size and relatively longer tail and hindlegs (Storr 1965; 8). Amphibolurus minor minor Sternfeld, 1919 Mainly southern, north nearly to Geraldton, extending inland in south, but north of Dongara confined to coastal dunes. Also far north-east (single specimens from Ajana and Binnu). Common in south, where it is found in a variety of habitats including Acacia rostellifera thickets in white coastal dunes, low coastal heaths, mallee scrubs on grey sandy loam, and the rich proteaceous heaths of interior sandplains. In the southern population the mouth is yellow; mouth colour has not been recorded in the north-east, but to the immediate north of our region (Murchison House) it has been noted as white. Amphibolurus reticulatus (Gray, 1845) Northern, south to Waggrakine and Eradu, mainly in the interior but following the Hutt, Bowes and Chapman Rivers downstream nearly to their mouths. Moder- ately common. Preferring the heavier soils. Lophognathus longirostris Boulenger, 1883 Regionally confined to the Greenough River. Common. Mainly the woodlands of Eucalyptus camaldulensis and scrubs of Melaleuca-Casuarina along the banks of the river. Moloch horridus Gray, 1841 Patchily in interior (24 km W of Binnu, Eradu, Mt Fanny and Eneabba). Locally common but generally scarce. Heath on sand or laterite. Scincidae Cryptoblepharus carnabyi Storr, 1976 Regionally confined to the Houtman Abrolhos: East Wallabi and West Wallabi Is and islet between East Wallabi and Seagull Is in the Wallabi Group; Rat, Helsinki and Morley Is in the Easter Group; and Gun L, an islet south of Gun I., and Shark I. in the Pelsacrt Group. (Specimens of Ablepharus boutonii' from Seagull, Pigeon, Hut and Pelsaert Is no doubt belong to C. carnabyi but are not available for checking.) Locally common in the Easter and Pelsacrt Groups but uncommon in the Wallabi Group. These skinks are wholly terrestrial, sheltering in limestone crevices and under reef debris. On the mainland found along the Murchison River just north of our region. Cryptoblepharus plagiocephalus (Cocteau, 1836) Northern interior south to the Greenough River, and southern interior north 223 Herpetofauna of the Geraldton Region to Lake Arramall (30 km SSE of Dongara). Uncommon. Mainly trees along water- courses and around lagoons, especially Eucalyptus rudis in south. Ctenotus fallens Storr, 1974 Southern half of mainland north to Arramall Farm (30 km SSE of Dongara) and inland to Eneabba; and the Houtman Abrolhos (North I.; East Wallabi, West Wallabi and Seagull Is in the Wallabi Group; Rat, Hut and Helsinki Is in the Easter Group; and Middle 1. in the Pelsaert Group). Common. Open or lightly wooded sandplains, coastal dunes and coastal limestone. Has been collected in the Kalbarri National Park to the immediate north of our region. Ctenotus impar Storr, 1969 Two records from southern interior: five specimens from Burma Road Reserve (28°55'S, 115°0UE) and two from 10 km S of Eneabba. Locally common. Sandplains. One specimen was taken from a scorpion burrow. Ctenotus lesueurii (Dumeril and Bibron, 1839) Two records: a specimen (41658) from coastal dunes near the lower Greenough River, and one (72985) from a yellowish sandplain 15 km E of Coolimba. Ctenotus mimetes Storr, 1969 One record from extreme north-east: a specimen (30321) collected at 3 km W of Ajana. Ctenotus pantherinus pantherinus (Peters, 1866) Patchily in interior (Binnu, 32 km N of Eneabba and 10 km S of Eneabba). Scarce. Heath on laterite and nearby sandplains. Ctenotus schomburgkii (Peters, 1863) Patchily in interior (Burma Road Reserve, 5 km SSE of Eneabba and 10 km S of Eneabba). Scarce. Sandplains. Egernia kingii (Gray, 1839) Mainland coast north to the Hutt River, and the Houtman Abrolhos (North L; East Wallabi, West Wallabi, Tattler, Seagull and Pigeon Is in the Wallabi Group; and Murray, Middle and Pelsaert Is and islet south of Gun 1. in the Pelsaert Group). Scarce on the mainland; moderately common to very common on the islands. Coastal dunes, cliffs and flats of broken limestone and reef debris, especially in vicinity of nesting terns. Egernia multiscutata bos Storr, 1960 Three records from far south; a specimen (26746) collected at Stockyard Gully, 224 G.M. Storr, T.M.S. Hanlon and J.N. Dunlop and observations by G.M. Storr on Eucalyptus todtiana sandplains 13 km SW and 31 km E of Eneabba. Egernia stokesii stokesii (Gray, 1845) Endemic to the Houtman Abrolhos: East Wallabi, West Wallabi, Tattler, Seagull and Pigeon Is in the Wallabi Group; and Murray and Middle Is in the Pelsaert Group; formerly occurring on Rat I. in the Easter Group, but disappearing between 1889 and 1913, following the introduction of domestic cats (Alexander 1922). Very common. Sheltering mainly under slabs of limestone; also in hollow stems of dead shrubs. Eremiascincus richardsonii (Gray, 1845) Northern interior south to White Peak. Rare. Favouring the heavier soils. Lerista christinae Storr, 1979 One record from southern interior: a specimen (70721) pit-trapped in rich proteaceous heathland 5 km SSE of Eneabba. Lerista distinguenda (Werner, 1910) On the mainland regionally confined to the valley of the Greenough River (collected at Eradu, Newmarracarra and Greenough); also Rat I. in the Easter Group, Houtman Abrolhos. Uncommon. Lerista elegans (Gray, 1845) Far south of mainland north to 16 km N of Eneabba, and East Wallabi and West Wallabi Is in the Houtman Abrolhos. Uncommon. Sandy country with low open vegetation. Lerista gerrardii (Gray, 1864) Northern interior south to the Greenough River (Newmarracarra). Mainly in leaf litter beneath acacias, especially jam [A. acuminata), on red loamy soils. Lerista greeri Storr, 1982 A specimen (188) was collected by W.B. Alexander in the Wallabi Group, Houtman Abrolhos, in November 1913. Otherwise this skink is only known from the Kimberley. Lerista lineopunctulata (Dumeril and Bibron, 1839) Greater part of mainland, inland to Northampton and Eradu; and West Wallabi I. in the Houtman Abrolhos. Moderately common in white coastal dunes; scarce on sandplains of interior. 225 Herpetofauna of the Geraldton Region Lerista macropisthopus (Werner, 1903) Extreme north-east (Ajana and Galena). Moderately common. Mainly in leaf litter beneath acacias growing on red loamy soils. Lerista planiventralis decora Storr, 1978 One record from southern interior: two specimens (73112-3) collected in white sand with heath and scattered eucalypts 16 km N of Eneabba. Lerista praepedita (Boulenger, 1877) Greater part of mainland, inland to Eradu and Eneabba; and the Houtman Abrolhos (North I.; East Wallabi and West Wallabi Is in the Wallabi Group; and Middle I. in the Pelsaert Group). Uncommon. Sandplains. Menetia greyii Gray, 1845 Sparsely distributed on the mainland (collected at Binnu, 6 km E of Dongara and 16 km E of Coolimba); and the Houtman Abrolhos (Eastern I. in the Wallabi Group, and Leo, Suomi, Rat and Morley Is in the Easter Group). Scarce on the mainland; common on the islands. Menetia surda Storr, 1976 One record from far north-east: a specimen (71048) collected under litter on a yellow sandplain 4 km N of Binnu. Morethia butleri (Storr, 1963) One record from the northern interior: a specimen (28003) collected at 12 km E of the mouth of the Hutt River. Morethia lineoocellata (Dumeril and Bibron, 1839) Northern and central coasts south to the vicinity of Greenough; and North, East Wallabi and West Wallabi Is in the Houtman Abrolhos. Common. Well- vegetated, white coastal dunes. Morethia obsenra Storr, 1973 Sparsely distributed in the interior (collected at Northampton, Burma Road Reserve, 10 km S of Eneabba and 15 km E of Coolimba); also Gun L, an islet south of Gun 1., and Pelsaert I. in the Pelsaert Group, Houtman Abrolhos, and possibly the Beagle Is (Ford 1963: 138). Rare on the mainland; common in the Pelsaert Group. Omolepida branchialis (Gunther, 1867) Common in far south-west, north to Beagle Point and inland to Stockyard Gully, in white coastal dunes and near-coastal sandplains. Scarce further north (collected at Galena, ‘presumably Geraldton’ and Newmarracarra). 226 G.M. Storr, T.M.S. Hanlon and J.N. Dunlop Tiliqua occipitalis (Peters, 1863) Coastal areas throughout the mainland, inland to Balline and Lake Arrow- smith. Moderately common. Mainly well-vegetated coastal dunes and near-coastal sandplains; also observed on a saltbush flat. Tiliqua rugosa rugosa (Gray, 1827) Throughout the mainland. Moderately common in farming country and near coast; scarce in uncleared parts of the interior. Varanidae Varanus caudolineatus Boulenger, 1885 Extreme north-east (two specimens from Ajana). Varanus gouldii (Gray, 1838) Sparsely distributed throughout the mainland. Preferring sandy country. Varanus tristis tristis (Schlegel, 1839) Two records: a specimen (1735) collected at Newmarracarra, and one observed by T.M.S. Hanlon in Acacia scrub 5 km N of Leeman. Typhlopidae Ramphotyphlops australis (Gray, 1845) Two records from southern interior: single specimens collected at Irwin House (5688) and Arrowsmith (21856). Ramphotyphlops hamatus Storr, 1981 Two records from northern half of mainland: single specimens from Geraldton (32368) and Newmarracarra (1733). Ramphotyphlops leptosoma Robb, 1972 Northern interior south to Newmarracarra. Moderately common. Apparently favouring the heavier soils. Ramphotyphlops waitii (Boulenger, 1895) Five of the six regional specimens come from the vicinity of Geraldton (includ- ing Wonthella and Greenough); the sixth is from the southern interior (2 km N of Eneabba Spring). [The specimen of 'Typhlina bituberculata' reported by Storr and Harold (1980) for the Zuytdorp area was in fact an R. waitii.] 227 Herpetofauna of the Geraldton Region Boidae Aspidites ramsayi (Macleay, 1882) Northern interior south to Newmarracarra. Scarce. One specimen was found in a rabbit warren in sandplain country at Eradu. Liasis *childreni’ Gr 2 iy, 1842 Three records: single specimens from 30 km N of Geraldton (76412), Moon- yoonooka (6142) and 5 km E of Coolimba (72984). Python spilotus imbricatus L.A. Smith, 1981 Coastal areas of -mainland north to Geraldton; and East Wallabi, West Wallabi and Seagull Is in the Wallabi Group, Houtman Abrolhos. Scarce on the mainland; common on the Wallabi Is, formerly occurring on North L (Storr 1960). Elapidae Demansia reticulata reticulata (Gray, 1842) Throughout the mainland. Moderately common. A specimen from Bookara had swallowed a frog (Pseudophryne guentheri). Denisonia gouldii {Gxdiy, 1841) One record from far southern interior: three specimens (28087-9) collected at Stockyard Gully Cave. Denisonia monachus Storr, 1964 Northern half of mainland south to the Greenough River (Newmarracarra). Uncommon. Notechis curtus (Schlegel, 1837) Far south of mainland north to the lower Arrowsmith River. Uncommon. Sandplains and coastal dunes. Pseudechis australis (Gray, 1842) Throughout the mainland. Moderately common in the interior; scarce or absent near the coast. Pseudonaja modesta (Gunther, 1872) Northern interior south to the Greenough River (Walkaway). Uncommon. Pseudonaja nuchalis (Gunther, 1858 Throughout the mainland. Very common in and around towns and settle- ments; moderately common in coastal dunes; apparently uncommon in uncleared parts of the interior. 228 G.M. Storr, T.M.S. Hanlon and J.N. Dunlop Vermicella bertholdi (Jan, 1858) Southern interior north to the Greenough River (Newmarracarra). Scarce. Vermicella bimaculata (Dumeril, Bibron and Dumeril, 1854) One record from far south: a specimen (72971) collected at 10 km N of Leeman. The specimens from ‘Greenough’ (cited by Storr 1967: 86) are now believed to have come from the Eastern Goldfields. Vermicella fasciolata fasciolata (Gunther, 1872) One record from extreme south-west: a specimen (62262) collected at 10 km N of Green Head. Vermicella littoralis Storr, 1968 Coastal dunes and limestone north to Horrocks; and East Wallabi and West Wallabi Is in the Houtman Abrolhos. Moderately common. Vermicella semi fasciata semifasciata (Gunther, 1863) Northern half of mainland south to Geraldton. Moderately common. Hydrophiidae Pelamis platura (Linnaeus, 1766) One record: a specimen (29609) found on the beach at Dongara after a storm in early October 1967. Discussion The 46 genera and 97 species and subspecies of amphibians and reptiles are distributed in 12 families as follows: Leptodactylidae Hylidae Cheloniidae Gekkonidae Pygopodidae Agamidae Scincidae Varanidae Typhlopidae Boidae Elapidae Hydrophiidae 6 genera, 9 species 1 genus, 1 species 2 genera, 2 species 7 genera, 14 species 6 genera, 1 1 species 3 genera, 8 species and subspecies 9 genera, 29 species 1 genus, 3 species 1 genus, 4 species 3 genera, 3 species 6 genera, 12 species 1 genus, 1 species 229 Herpetofauna of the Geraldton Region In view of the number of species regionally known from only one or two records, we expect several more reptiles to be added to the list. These addition- al species are most likely to be among those recorded from the area immediately north of our region, viz. Aprasia smithi, Amphibolurus parviceps, A. scutulatiis, Ctenotus severusy Lerista connivens, L. humphriesiy L. muelleriy L, 'nichollsi' and Varanus eremius. Species recorded from the immediate south are less likely to be found in our area, for their habitat (lateritic uplands with dry sclerophyll forest) ceases just before our region is reached; in this category are Egernia napoleonis and the mainland population of Crenadactylus o. ocellatus. Two species, Chelodina oblonga and Notechis scutatuSy have their northern limit on the Hill River, and Hemiergis peronii quadrilineata extends north to Jurien Bay. Nevertheless many south-west Australian frogs and lizards extend into the present region, eleven of them attaining their northern limit here, viz.: Heleioporiis eyrei (north to the lower Greenough) H. psammophilus (to the lower Irwin) Litoria moorei (to the lower Hutt) Aprasia repens (to Geraldton) Diplodactylus polyophthalmus (to Eneabba) Pletholax g. gracilis (to Eneabba) Ctenotus impar (to the Burma Road Reserve) Egernia kingii (to the lower Hutt on the mainland; to North I. in the Houtman Abrolhos) and Lerista christinae (to Eneabba) Lerista distinguenda (to Greenough on the mainland; to Rat 1. in the Houtman Abrolhos) and Python spilotus imbricatus (to Geraldton on the mainland; to North 1. in the Houtman Abrolhos) To these could be added Egernia multiscutata boSy which extends only to Eneabba on the mainland but reappears much further north on Bernier I. in Shark Bay. An even larger number of arid-zone reptiles have their southern limit (at least on the west coast or in its vicinity) within the present region: Diplodactylus squarrosus (extending south to Binnu) Diplodactylus strophurus (to Binnu) Nephrurus levis occidentalis (to the hinterland of Geraldton) Delma nasuta (to Northampton) Delma tincta (to East Chapman) Pygopus n. nigriceps (to the lower Irwin) Amphibolurus inennis (to Lake Arrowsmith) Amphibolurus reticulatus (nearly to Geraldton) Moloch horridus (to Eneabba) 230 G.M. Storr, T.M.S. Hanlon and J.N. Dunlop Ctenotus mimetes (to Ajana) Ctenotus schomburgkii (to Eneabba) Lerista gerrardii (to the Greenough) Lerista macropisthopus (to Ajana) Menetia surda (to Binnu) Morethia butleri (to the Hutt) Varanus caudolineatus (to Ajana) Ramphotyphlops hamatus (to the Greenough) Ramphotyphlops leptosoma (to the Greenough) Aspidites ramsayi (to the Greenough) Denisonia monachus (to the Greenough) and Pseudonaja modesta (to the Greenough) It follows from the foregoing discussion that a high proportion of the species inhabiting the northern half of the region do not extend to the southern half, and vice versa. In previous analyses of west-coast herpetofaunas we have remarked on the high rate of latitudinal replacement of species between regions (Storr and Hanlon 1980). This, however, is the first time we have encountered a high rate of latitudinal replacement within a region, but then this is the first region studied in which there is a marked latitudinal gradient in rainfall, the annual mean doubling from north-east to south-west. The only endemic taxa are the two lizards restricted to the Houtman Abrolhos: Amphibolurus minor minimus and Egernia stokesii stokesii. We return to the problem of broken distributions along the mid-west coast of Western Australia. The principal disjunctions are in: Phyllodactylus marmoratus — 460 km between Cockleshell Gully and False Entrance Well, Edel Land Aclys concinna ^ 370 km between Coolimba and Tamala Pletholax gracilis — 440 km between Eneabba and Useless Loop, Edel Land Egernia multiscutata — 600 km between Stockyard Gully and Bernier 1. Lerista elegans — 425 km between 15 km N of Eneabba and Useless Loop, and Morethia lineoocellata — 190 km between Badgingarra and Greenough In December 1980 T.M.S. Hanlon was only able to reduce the disjunction in two of the above species [Aclys concinna and Lerista elegans)^ and then only by c. 30 km. However the time spent in the field (three weeks) was very short in relation to the large area covered. As it is difficult to distinguish between complete absence and very low density, we feel that most of the above disjunc- tions are not yet proved. The exceptional species are Pletholax gracilis and Egernia multiscutata. The separability of the northern population of Pletholax gracilis as a distinct subspecies [edelensis) implies geographic isolation. As for Egernia multiscutata, we concede that the skinks themselves may have been overlooked by naturalists, but hardly their conspicuous burrows. 231 Herpetofauna of the Geraldton Region Table 1 Distribution of the reptiles of the Houtman Abrolhos by island groups. North Wallabi Easter Pelsaert Cheloniidae Chelonia my das X X Gekkonidae Crenadactylus ocellatus X X X Diplodactylus spinigerus X Gehyra variegata X X Heteronotia binoei X X Phyllodactylus marmoratus X X X X Phyllurus milii X X Pygopodidae Delma australis D. gray a X X Lialis burtonis X Agamidae Amphibolurus minor X X Scincidae Cryptoblepharus carnabyi X X X Ctenotus fallens X X X X Egernia kingii X X X E. stokesii X X X Leris ta distinguenda L. elegans X X L. greeri X L. lineopunctulata X L. praepedita X X X Menetia greyii X X Morethia lineoocellata M. obscura X X X Boidae Python spilotus X X Elapidae Vermicella littoralis X TOTAL 9 22 8 10 232 G.M. Storr, T.M.S. Hanlon and J.N. Dunlop Finally some comments on the Houtman Abrolhos. Geographically the islands are divided into four quarters, namely (from north to south) North I. and the Wallabi, Easter and Pelsacrt Groups. North I. and the western islands of the Wallabi, Easter and Pclsaerl Groups are continental remnants that were separated from the mainland by rising sea-level c. 1 1500 years ago (Main 1961). The eastern islands in each group are recent accumulations of coral fragments (O’Loughlin 1969, Green 1972). Twenty-five species of reptile occur in the Houtman Abrolhos; their distribu- tion by island groups is set out in Table 1. Not surprisingly the Wallabi Group is the richest; in East Wallabi and West Wallabi it contains much the largest and most varied islands in the archipelago. Generally the fauna of North L and the Pelsaert Group are attenuated versions of the Wallabi Group fauna. However, the fauna of the Easter Group is quite peculiar in its composition. Only eight species have been found in the group, but two of them, Delma australis and Lerista distin- guenda, are not known from any other group. Moreover, Delma australis is absent from the opposite mainland, and Lerista distinguenda is restricted to one small part of the Geraldton region. A third Easter Group species, Menetia greyii, is unknown from other Groups except for its occurrence on Eastern 1. in the Wallabi Group. Now Eastern 1. is one of the islands recently formed from reef debris; and Menetia greyii, the sole reptile on the island, must have arrived there from over the sea. Perhaps the peculiarities of the Easter Group are similarly due to trans- marine dispersal. Comprising seven families, 17 genera and 21 species of reptiles, the fauna of the Wallabi Islands is continental in its diversity. Nevertheless this fauna differs considerably from that of any comparable area on the mainland. The last 11500 years has seen not only the evolution of distinct subspecies but also the extinction of such widespread west-coast reptiles as Diplodactylus ornatuSy Amphibolurus adelaidensisIparvicepSy A. maculatus, Ctenotus lesueurii, Omolepida branchialis, Tiliqua occipitalis, T. rugosa and Pseudonaja af finis Inuchalis. Nor has the main- land fauna gone unchanged in this period, as indicated by the withdrawal of Crenadactylus ocellatus, Phyllodactylus marmoratus and Cryptoblepharus car- nabyi. References Alexander, W.B. (1922). The vertebrate fauna of Houtman’s Abrolhos (Abrolhos Islands), Western Australia./. Linn. Soc. (ZooL) 34: 457-486. Beard, J.S. (1976). Murchison. Explanatory notes to Sheet 6, 1:1 000 000 series. Vegetation Survey of Western Australia. (University of Western Australia Press: Nedlands). Burbidge, A.A., Fuller, PJ. and McCusker, A. (1978). The wildlife of the proposed Wandana Nature Reserve, near Yuna, Western Australia. Report No. 32, (Dept of Fisheries and Wild- life: Western Australia). Dell, J, and Chapman, A. (1977). Reptiles and frogs of Cockleshell Gully Reserve. Rec. West. Aust. Mus. Suppl. No. 4: 75-87. 233 Herpetofauna of the Geraldton Region Ford, J, (1963). The reptilian fauna of the islands between Dongara and Lancelin, Western Australia. West. Aust. Nat. 8: 135-142. Green, G.A. (1972). Fifth Abrolhos expedition, 1970. (Aquinas College: Manning). Harrison, L. (1927). Notes on some Western Australian frogs, with descriptions of new species. Rec. Aust. Mus. 15: 277-287. Main, A.R, (1961). The occurrence of Macropodidae on islands and its climatic and ecological implications./. Proc. R. Soc. West. Aust. 44: 84-89. O’Loughlin, P.M. (1965). Aquinas College expedition to Wallahi Islands of Hout man’s Abrol- hos. (Aquinas College: Manning). O’Loughlin, P.M. (1966). Aquinas College second expediton to Wallabi Islands of Houtman’s Abrolhos. (Aquinas College: Manning). O’Loughlin, P.M. (1969). Aquinas College third and fourth expeditions to the Pelsart Group of Houtman’s Abrolhos. (Aquinas College: Manning). Storr, G.M. (1960). The physiography, vegetation and vertebrate fauna of North Island, Hout- man Abrolhos./. Proc. R. Soc. West. Aust. 43: 59-62. Storr, G.M. (1965). The physiography, vegetation and vertebrate fauna of the Wallabi Group, Houtman Abrolhos./. Proc. R. Soc. West. Aust. 48: 1-14. Storr, G.M. (1967). The genus Vermicella (Serpentes, Elapidae) in Western Australia and Northern Territory. /. Proc. R. Soc. West, Aust. 50; 80-92. Storr, G.M. and Hanlon, T.M.S. (1980). Herpetofauna of the Exmouth region, Western Aus- tralia. Rec. West. Aust. Mus. 8: 423-439. Storr, G.M. and Harold, G. (1980). Herpetofauna of the Zuytdorp Coast and hinterland. West- ern Australia. Rec. West. Aust. Mus. 8: 359-375. Received 3 April 1981 Accepted 9 June 1982 Published 20 April 1983 234 Rec. West. Aust. Mus. 1983 , 10 ( 3 ): 235-242 Human Skeletal Remains from Cheetup, Western Australia Leonard Freedman* and Mancel Lofgrenf Abstract This paper describes Pleistocene human skeletal material from an excavation by M. Smith in 1979 at Cheetup, Western Australia. Recovered from a hearth now dated at over 12 000 BP, these remains include fire-damaged cranial, post-cranial, and dental fragments. Analysis of the material reveals that it represents a child, possibly female, 3-9 months of age at death. The source of the material and its condition are consistent with deliberate cremation. Background The material described in this paper comes from Cheetup (33°52'S, 122°27'E), a rockshelter approximately 55 km E of Esperance, Western Australia (Figure 1). Excavated by M. Smith (1982), this site is listed in the catalogue of the Depart- ment of Aboriginal Sites, Western Australian Museum as W0594. These fragments (WAM registration number A23441) were recovered during the first season of excavation in 1979 at a depth of 7-33 cm in trench F12, spit 4, in a feature subsequently identified as a hearth. Radiocarbon dating of charred wood from this feature is 12 845 ± 310 BP (GX-6604). Initial sorting by J. Balme of the bone fragments recovered, indicated the possibility of some being of human origin. The following paper confirms that preliminary identification, and describes the material in detail. Previous descriptions of Australian Aboriginal skeletal material from Western Australia include three teeth and a pelvic fragment from DeviTs Lair (Davies 1968, 1973; Freedman 1976; Allbrook 1976), a study of cranial morphometries (Margetts and Freedman 1977), the Cossack material (Freedman and Lofgren 1979a, b), a study of odontometrics (Freedman and Lofgren 1981), middle ear ossicles (Blumcr, Freedman and Lofgren 1982), and discrete non-metric cranial traits (Milne, Schmidt and Freedman [in press]). Because of the already known length of human occupation of Australia (> 40 000 BP, Jones 1979), as well as the * Department of Anatomy and Human Biology, University of Western Australia, Nedlands, Western Australia 6009. t Department of Anthropology, Western Austredian Museum, Francis Street, Perth, Western Australia 6000. 235 Human Skeletal Remains from Cheetup considerable variation in form demonstrated by described material, there is par- ticular importance in all discovered Pleistocene human skeletal material. Also, problems of regional variation in space and time require the close study of all available material so that existing models may be refined. This paper offers new Material The fire-damaged skeletal remains recovered from this excavation include more than 100 small bone fragments, of which the largest piece is only 25 mm x 16 mm. The identifiable fragments are primarily from the skull vault, but 5 pieces of bone are clearly from the post-cranial skeleton, and there are parts of 6 teeth also present. 236 Leonard Freedman and Mancel Lofgren Skull Bones More than 60 small fragments of the cranial vault are present (Figure 2). Only 4 fragments are more than 15 mm x 10 mm, and no attempt has been made to reconstruct the vault. In most instances only a single table of bone (usually the outer) is present, but a few fragments include both tables and many of the others have some of the inner cancellous bone still attached. When both tables of bone are present, the thickness of the fragments vary from 1 mm to 1.5 mm, but it is not possible to identify the regions of the vault from which these fragments came. Figure 2 Cranial fragments, Cheetup, W.A. Small Unidentified Fragments of Bone There are over 25 small slivers and thicker pieces of bone (Figure 3). Some of the thicker pieces could come from the cranial base or face. Only one piece of these is more than a fragment. That piece is about 25 mm x 16 mm and over 7 mm at its maximum thickness, and could be part of the mastoid region of a temporal bone. Some of the other fragments could be parts of post-cranial bones, particularly long bones. Post-cranial Bones Of the 7 post-cranial fragments (Figure 4), 5 are identifiable. Three are proximal or distal parts of limb long bones. Two of these are probably the right and left proximal parts of the shafts of humeri (Figure 4A and B), the cross- sectional measurements of these pieces being approximately 7.8 mm x 9.1 mm 237 Human Skeletal Remains from Cheetup Figure 3 Unidentified fragments, Cheetup, W.A. Approximate dimensions of the Cheetup teeth, and comparisons with some other deciduous teeth mentioned in the text (in mm). Tooth Dimension Cheetup Devil ’s Lair Mod. Aust. Ab. Means $ 9 d di^ height (crown) 7.2 length (m-d) 6.0 7.0* 7.20 7.35 breadth (b-1) — 5.25* 5.30 5.47 di^ length (m-d) — 6.0t 5.93 6.00 dc height (crown) 6.0 length (m-d) 5.7 6.16 6.31 breadth (b-I) 5.2 5.84 6.05 dm^ length (m-d) 7.0 7.28 7.55 breadth (b-1) 8.0 (1 & r) - 8.77 9.07 * Davies, P.L. (1968) t Freedman, L. (1976) t Margetts, B. and Brown, T. (1978) 238 Leonard Freedman and Mancel Lofgren cm Figure 4 Post-cranial fragments, Cheetup, W.A. (see text for identifications). and 7.5 mm x 9.1 mm. The third end fragment (Figure 4C) is damaged but might be the distal end of the shaft of a humerus. The fourth fragment (Figure 4D) is an epiphysis, probably from the distal end of a humerus and it measures about 10 mm X 7.5 mm. The fifth fragment (Figure 4E) is a small portion of the epiphysis of a long bone, but it is too small to be more specifically identified. Teeth There are 6 individual teeth or parts of teeth present in the remains found. The specimens arc all the unworn calcified crowns, or parts of crowns, of deciduous 239 Human Skeletal Remains from Cheetup teeth; no roots are present (Figure 5). The specimens include: (Figure 5A) the ^ial part of the crown of the right di^ ; (Figure 5B) the crown, probably of a dc; (Figure 5C and D) the whole left, and part of the right, crown of dm^ , (Figure 5E) part of the crown ol a right dmi, and (Figure 5F) another dm crown frag- ment, possibly the left dmi . The approximate measurements of the teeth are given in Table 1 . cm Figure 5 Deciduous teeth, Cheetup, W.A. (see text for identifications). Conclusions The small size and partially charred condition of the bone fragments described make reconstruction and more extensive analysis impractical. The majority of the fragments are from the cranial vault and probably represent virtually the whole of that region. The few possible cranial base and face fragments are too small for any certain assignments to be made. Of the post-cranial skeleton, only 5 identifiable fragments were found and these are the ends of long bones or their epiphyses. The coloration of these frag- ments suggests extensive charring much more so than does the condition of the skull Iragments, in which the colour and damage suggest less exposure to heat. These differences may account for the paucity of post-cranial remains. The 6 teeth and tooth fragments arc all ol deciduous teeth lacking roots. They show no obvious signs of attrition and it does not appear as if the roots had been formed. On the basis of the apparently unworn state and lack of roots of the teeth found, an approximate age of about 6 months would seem the most reason- able assessment, using the ossification data summarized by Gabriel (1965). In 240 Leonard Freedman and Mancel Lofgren view of individual variability and the fact that initially there would be little wear on the teeth, a range of 3-9 months of age is postulated. In view of these dimen- sions being smaller than even the female mean values of Margetts and Brown (1978), it is possible that the remains are those of a female child. The location of the remains within a discrete feature identified as a hearth, and condition of the fragments of bone suggest some tentative conclusions. As with the Lake Mungo cremation (Bowler et al. 1970), the pyre did not entirely consume all bone fragments, and differential charring was the result. Charring is consistently more intense on external surfaces of individual fragments. While other explanations are possible for the presence and state of the remains, it appears most likely that the child w'as cremated as a complete and fully-fleshed cadaver. Acknowledgements We thank Moya Smith for the opportunity to study the material, Jane Balme for initially recognizing its importance, and John McGeachie for assistance with tooth identification. References Allbrook, D.B. (1976). A human hip bone from Devil’s Lair, Western Australia. ArchaeoL Phys. Anthropology Oceania 11: 48-50. Blumer, W.F.C., Freedman, L. and Lofgren, M. (1982). Middle ear ossicles of Australian Aborigines. ArchaeoL Oceania 17: 127-131, Bowler, J.M., Jones, R,, Allen, H. and Thorne, A.G, (1970). Pleistocene human remains from Australia: a living site and human cremation from Lake Mungo, western New South Wales. World Archaeology 2: 39-60. Davies, P.L. (1968). An 8000 to 12,000 human tooth from Western Australia. ArchaeoL Phys. Anthropology Oceania 3: 33-40. Davies, P.L. (1973). A human tooth from Devil’s Lair, Appendix to Dorch, C.E. and Merrilees, D. Human occupation of Devil’s Lair, Western Australia during the Pleistocene. ArchaeoL Phys. Anthropology Oceania 8: 89-115. Freedman, L. (1976). A deciduous human incisor tooth from Devil’s Lair, Western Australia. ArchaeoL Phys. Anthropology Oceania 11: 45-46. Freedman, L. and Lofgren, M. (1979(2). Human skeletal remains from Cossack, Western Aus- tralia./. Hum. Evolut. 8: 283-299. Freedman, L. and Lofgren, M. (19796). The Cossack skull and a dihybrid origin of the Aus- tralian Aborigines. Nature 282: 298-300. Freedman, L. and Lofgren, M. (1981). Odontometrics of Western Australian Aborigines. ArchaeoL Oceania 16: 87-93. Gabriel, A.C. (1965). Anatomy of Teeth and Jaws. University of Sydney: Sydney. Jones, R. (1979). The fifth continent: problems concerning the human colonization of Aus- tralia, ^nn. Rev. AnthropoL 8: 445^66. Margetts, B. and Brown, T. (1978). Crown diameters of the deciduous teeth in Australian Aboriginals. Am. J. Phys. Anthrop. 48: 493-502. 241 Human Skeletal Remains from Cheetup Margetts, B.M. and Freedman, L. (1977). Morphometries of Western Australian Aboriginal skulls. Rec. West, Aust. Mus. 6: 63-105. Milne, N., Schmidt, L.H. and Freedman, L. (In press). Discrete trait variation in Western Aus- tralian Aboriginal skulls. ], Hum. EvoL Smith, M. (1982). Late Pleistocene Zamia exploitation in southern Western Australia. Archaeol. Oceania. 17: 117-121. Received 7 July 1982 Accepted 22 September 1982 242 Published 20 April 1983 Rec. West. Aust. Mus. 1983 , 10 ( 3 ): 243-261 A Review of the Gudgeon Genus Hypseleotris (Pisces: Eleotridae) of Western Australia, with Descriptions of Three New Species Douglass F. Hoese* and Gerald R. Allenf Abstract The five Western Australian members of the eleotrid genus Hypseleotris Gill are reviewed. These small fishes are inhabitants of brackish water, freshwater streams, swamps and ponds, primarily in northern Australia and New Guinea. The species treated in the present paper include H. aurea (Shipway) and H. compressa (Krefft). In addition, the following three species from the IGmberley district are described as new: H. ejuncida, H. kimberleyensis, and//, regalis. The first two species are closely allied, differing from each other primarily on the basis of coloration, body depth, and number of vertebrae. Hypseleotris regalis is most similar to H. aurea, but differs with regard to squamation and number of vertebrae. A key to Western Australian Hypseleotris and diagnostic illustrations are provided. Introduction Prior to 1970 relatively few comprehensive studies had been carried out on the taxonomy of Australian freshwater fishes. Some areas, such as the north-west, had not been sampled extensively primarily because of the inaccessibility of much of this region. Recent collections from the Kimberley district of far northern Western Australia by the Western Australian Museum has resulted in the discovery ol a number of undescribed fishes, including a new genus and six new species of eleotrids. Three of the new species belong to the genus Hypseleotris, which occurs in fresh and brackish waters of the Indo-west Pacific. They are small free-swimming fishes which feed mainly on tiny invertebrates. Breeding males are often colourful and make attractive aquarium fishes. Currently few species are recognized, but virtually no work has been done on species occurring outside Australia. Although not recorded from the New World, it is very likely thdLi Hemieleotris from Central America will prove identical to Hypseleotris. Similarly, Batanga from Africa appears close to Hypseleotris. * Department of Ichthyology, The Australian Museum, 6-8 College Street, Sydney, New South Wales 2000. t Department of Ichthyology, Western Australian Museum, Francis Street, Perth, Western Australia 6000. 243 A Review of the Gudgeon Genus Hypseleotris Although several species have been described from Australia, generally only four were previously recognized as valid: H. compressa (Krefft),//. galii (Ogilby), H. klunzingeri (Ogilby), and H. simplex (Castelnau). Recently Hoese, Larson and Llewellyn (1980) recorded two additional undescribed species from south-eastern Australia. Although we are unable to locate the type specimen of//, simplex, it is regarded here as a junior synonym of H, compressa, which is widely distributed in northern and eastern Australia. The xemdimmg Hy pseleotris treated herein, includ- ing H. aurea and three new species are apparently restricted to north-western Western Australia. Our numerous collections from streams in the Northern Territory and the Gulf of Carpentaria drainage of northern Queensland contain only H. compressa. Other Hypseleotris species are known in Australia from coastal drainages extending from the Atherton Tablelands of northern Queensland to New South Wales and Victoria, and from inland drainages, including the Murray-Darling system, of southern Queensland, New South Wales, Victoria and South Australia. Methods Methods for counts and measurements mainly follow those of Hubbs and Lagler (1958). The longitudinal scale count was taken from the upper pectoral base obliquely to the midline and then to the end of the hypural. The transverse scale count was taken from the origin of the second dorsal fin downward and backward to the anal base (TRDB). The postdorsal scale count is taken from the end of the second dorsal fin to the caudal base middorsally, and includes only scales on the middle of the top of the caudal peduncle. Gill raker counts include all rudiments. Counts for vertebrae, fin rays, scales and gill rakers are presented in Tables 1-6. In addition, the body^ depth and pelvic fin length of the five species are compared in Tables 7 and 8 respectively. Specimens studied, including types of the new taxa, are deposited at the following institutions: Australian Museum, Sydney (AM); British Museum (Natural History), London (BMNH); Museum National d’Histoire Naturelle, Paris (MNHN); and Western Australian Museum, Perth (WAM). Systematics Hypseleotris Gill Hypseleotris Gill, 1863: 270 (Eleotris cyprinoides Valenciennes, by original designation). Carassiops Ogilby, 1897: 732 {Eleotris compressus Krefft, by original designation). Austrogobio Ogilby, 1898: 784 {Carassiops galii Ogilby, by original designation). Caulichthys Ogilby, 1898: 784 {Asterropteryx guentheri Bleeker, by original designation). Shipioayia Whitley, 1954a: 155 {Eleotris aurea Shipway, by monotypy and subsequent designa- tion by Whitley, 1954b: 30). 244 Douglass F. Hoese and Gerald R. Allen Description Head and body distinctly compressed. Mouth small, very oblique, posterior margin under or anterior to middle of eye. Gill opening moderately broad, extending forward to below posterior end of preoperculum. Pectoral base narrow, with rays developed ventrally; a free fold of skin extending to upper attachment of opercular membrane above uppermost ray. Body scales large and ctenoid. Teeth small in several I'ows in both jaws. Tongue tip truncate. First dorsal VI- VIII (rarely V or IX). Second dorsal 1,8-12 (rarely 13). Anal 1,10-13 (rarely 14). Pectoral rays 14-17 (rarely 13). Longitudinal scale count 24-43. Transverse scale count (TRDB) 7-15. Segmented caudal rays 15. Vertebrae 25-33. Head pores present or absent, sometimes with up to 5 preopercular pores and sometimes two pores above each eye. Cheek sensory papillae normally in longitudinal rows, with few vertical rows (except in H. klunzingeri). Body scales ctenoid, often cycloid on belly and nape. Head scales present or absent. Remarks Until recently Australian species of Hypseleotris were placed in the genus Carassiops Ogilby. Examination of type material of the type species of Hypseleo- tris (H. cyprinoides at MNHN) and Carassiops (//. compressa at BMNH and AM) revealed that these two forms represent a closely related, apparently allopatric, species pair. H. compressa differs from H, cyprinoides primarily in lacking inter- orbital scales before the middle of the eye. H. compressa is known from Australia and possibly New Guinea while H. cyprinoides is found in fresh and brackish water throughout the western tropical Pacific outside Australia. Whitely (1954a) separated Shipwayia from Carassiops {=Hypseleotris) only on the basis of the higher scale counts in Shipwayia', 45 to 50 rows versus 27 to 35. Our counts indicate scale counts of 33 to 43 for H, aurea compared with 24 to 34 in other Australian species. Because H. aurea is closely related to H. regalis, sp. nov., an otherwise typical member of the genus which has low scale counts, there appears to be no justification for separating Shipwayia and Hypseleotris. The Western Australian species can be conveniently grouped into three categories on the basis of colour similarities. The first contains only H. compressa; the second H. aurea and the allopatric H. regalis, sp. nov.; and the third H. kimber- leyensis, sp. nov. and the allopatric //. ejuncida, sp. nov. On the basis of fin coloration, it appears likely that the two complexes which are restricted to Western Australia were derived from the same stock as H. compressa. They do not show any close similarity to species from south-eastern Australia. //. compressa is the only known Australian species to occur in salt as well as fresh water, and is the only species ranging throughout tropical and subtropical Australia. All species from Western Australia are basically similar in sensory papillae patterns (Figure 1), a feature which is often diagnostic in eleotrids. 245 A Review of the Gudgeon Genus Hypseleotris Figure 1 Diagrammatic sketch of sensory papillae on head of Hypseleotris regalis. The draw- ing is a composite, based on several specimens. Papillae numbers are approximate and the pattern shown may be incomplete as the papillae are often difficult to detect. Key to Western Australian Species of Hypseleotris la Preoperculum with 2-4 pores. Predorsal scales reach forward to above middle of eyes. Scales from under first dorsal fin to upper attachment of opercular membrane ctenoid; nape scales often ctenoid. Second dorsal fin-rays modally 1,9 (Murchison River northward to Kimberleys) H. compressa (Krefft) lb No preopercular pores. Predorsal scales reach for- ward to or behind middle of eye. Scales from under first dorsal fin to upper attachment of opercular membrane cycloid; nape scales if present cycloid. Second dorsal fin-rays usually 1,10 2 2a Longitudinal scale count 34-43. Postdorsal scale count 10-13. Body relatively deep, body depth at pelvic fin origin 20-23% of SL (Murchison River) H. aurea (Shipway) 2b Longitudinal scale count 24-32. Postdorsal scale count 7-9. Body more slender, depth at pelvic fin origin L4-20% of SL 3 3a Longitudinal scale count 24-26, Body moderately slender, body depth at pelvic fin origin 16-20% of SL. Predorsal scaled completely to behind eyes. Gill rakers on outer face of second arch modally 10. Body without vertical bars, but with 15-20 246 Douglass F. Hoese and Gerald R. Allen chevron marks on side. Pelvic fin origin under posterior opercular margin (West Kimberley) H. regalis sp. nov. 3b Longitduinal scale count 28-32. Body very slender, depth at pelvic fin origin 14-18% of SL. Predorsal naked or partly scaled, with naked patches. Gill rakers on outer face of second arch 8-9. Body with faint vertical bars anteriorly. Pelvic fin origin under or behind pectoral fin insertion 4 4a Predorsal completely naked. Operculum naked. Second dorsal and caudal fins clear to dusky with- out distinct spots or wavy bands; caudal sometimes with 2 faint vertical bands. Dark bar on pectoral fin base developed dorsally only, usually only above uppermost pectoral fin ray. Body depth at anal origin 13-17% of SL. Vertebrae 25 (Central Kimberley H. kimberleyensis sp. nov. 4b Predorsal usually extensively scaled, midline some- times with few scales. Operculum with large scales. Second dorsal fin with numerous white spots, surrounded by dark pigment forming wavy bands. Caudal fin with dark spots, forming 4-6 wavy bands. Bar at base of pectoral fin covering whole base. Body depth at anal fin origin 17-18% of SL. Vertebrae 26 (West Kimberley) H, ejuncida sp. nov. Hypseleotris compressa Figure 2 Eleotris compressus Krefft, 1864: 184 (Clarence River and Port Denison). Hypseleotris simplex - Allen, 1975: 95, fig. 12 (in part, WAM P25036-003). Material Examined AM 1.22746-001, 7 (27-35 mm SL), Lawley River, Western Australia; WAM P.16562- 16563, 2 (58 mm SL), Murchison River, Western Australia WAM P25-36-003, 35 (13-37 mm SL), Prince Regent River, Western Australia. Diagnosis A species of Hypseleotris closely related to H. cyprinoides of the western tropical Pacific (exclusive of Australia), but differing from it by lacking inter- orbital scales anterior to the middle of the eyes. It differs from other Western Australian members of the genus by a combination of characters which include the presence of preopercular pores; predorsal scales extending anteriorly to 247 A Review of the Gudgeon Genus Hypseleotris above middle of eyes; presence of ctenoid scales between first dorsal fin base and upper attachment of opercular membrane; and a modal count for the second dorsal fin of 1,9. Description Two to five preopercular pores. Adults usually with two pores connected by short tube above dorsoposterior margin of eye. Predorsal scales reach forward to above middle of eye; cycloid or ctenoid. Cheek with four to six rows of small embedded cycloid scales. Operculum with medium-sized cycloid scales. Gill opening extends forward to below posterior preopercular margin. First dorsal VI. Second dorsal 1,9-10. Anal usually 1,10-11. Typically with 1 or 2 more anal than dorsal rays. Pectoral usually 14-15. Longitudinal scale count 25-29. Predorsal scales 14-18. Transverse scales (TRDB) 9-10. Postdorsal scales 8-9. Gill rakers on outer face of first arch 3-4+1 + 19-11 + 12-16. Vertebrae 26. Sides of body often with about 7 or 8 brown vertical bars, often forming X-shaped marks on midside. Base of caudal fin with a vertically elongate dark brown spot just below midside. Dorsal fins with 2 black stripes. Second dorsal with round white spots posteriorly, surrounded by black. Live and fresh adult specimens viewed from above usually have a distinct dark mark near the posterior end of the second dorsal fin. Figure 2 Hypseleotris compressa^ female, 29 mm SL, Prince Regent River, Western Australia. Sexual Dimorphism Males typically have more elongate posterior second dorsal and anal rays, a higher first dorsal fin, and the two dorsal fins are closer together. Males reach a larger size than do females, and often develop a fleshy hump on the top of the end of the head. Breeding males are often more brightly coloured than females. Distribution and Habitat Hypseleotris compressa is common in lower reaches of rivers from eastern Victoria, New South Wales, Queensland, Northern Territory, and Western Aus- tralia. The Western Australian distribution extends from the Murchison River 248 Douglass F. Hoese and Gerald R. Allen (approximately 27‘^40'S) northwards in coastal streams and across the Kimberley region to the Northern Territory border. The species is most abundant in flowing waters and is often associated with aquatic vegetation. Juveniles are frequently found in swift-flowing water or in estuaries. The species sometimes occurs in full strength sea water, and this tolerance undoubtedly contributes to its wide distri- bution. Hypseleotris aurea Figure 3 Eleotris aurea Shipway, 1950: 75 (Murchison River). Shipioayia aurea — Whitley, 1954a: 152 and 155 (description of new genus). Material Examined AM 1.22743-001, 6 (43-48 mmSL), Murchison River, Western Australia; WAM P.3273 (holo- type), 1 (49 mm SL), Murchison River, Western Australia; WAM P22122-004, 14 (33-40 mm SL), Murchison River, Western Australia. Diagnosis A species of Hypseleotris closely related to H. regalis of the Prince Regent Reserve, West Kimberley. It differs from this species by possessing an additional vertebra (26 versus 25) and smaller scales. Ranges of counts for longitudinal scales, transverse scales, predorsal scales, and postdorsal scales arc 34-43 v. 24-26, 10-13 v. 8-9, 16-21 v. 14-16, and 10-13 v. 8 lor H. aurea and//, regalis respective- ly. It differs from H. compressa on the basis of characters indicated in the diag- nosis section for that species and from other Western Australian Hypseleotris on the basis of a combination of characters which include its relatively high longi- tudinal and postdorsal scale counts and deeper body, the depth at pelvic fin origin 14-20% of SL. Description No head pores. Prcdorsal scales reach forward to above middle of eye. Cheek with four rows of small embedded cycloid scales. Operculum covered with medium-si/.ed cycloid scales. Gill opening extends forward to below posterior pre- opercular margin. First dorsal VI. Second dorsal usually 1,10. Anal usually 1,11 with 1 more anal ray than dorsal ray. Pectoral 14-16, usually 15-16. Longitudinal scale count 34-43. Transverse scale count (TRDB) 10-13. Predorsal scales 16-21. Postdorsal scales 10-13. Gill rakers on outer face of first arch 2-3+H-8-9 = 11-13. Vertebrae usually 26. Pelvic fin short, reaching half to two-thirds of distance to anus, 13-17% of SL. Body relatively deep, body depth at anal origin and at pelvic origin 20-23% of SL. Snout and head above eye dark brown in adult males. A black bar at pectoral base, entirely above pectoral rays. Body with 6-9 vertical brown bands, width about equal to eye diameter; bands more pronounced 249 A Review of the Gudgeon Genus Hypseleotris anteriorly and dorsally, often fading on caudal peduncle. A small dusky spot at base of caudal rays below midside. Second dorsal with 2 or 3 rows of small dark spots on basal half of fin, surrounding white to clear spots. Distal tip of fin dusky, sometimes whole fin dusky. Extreme tip of anal usually white. Figure 3 Hypseleotris aurea, male, 36 mm SL, Murchison River, Western Australia. Sexual Dimorphism Adult males have a prominent forehead hump extending from above the upper end of the opercular margin forward to the snout. The posterior dorsal and anal rays are not prolonged in males, nor were any other sexual differences observed. Distribution and Habitat Hypseleotris aurea is known only from the Murchison River, which flows into the Indian Ocean approximately 500 km north of Perth, Western Australia. It has been collected from small, quiet pools near the North-West Highway crossing at Galena (approximately 27^49'S, IH^dl'E). The habitat was characterized by moderately turbid water and a boulder substratum littered with dead tree branches. Hypseleotris regalis sp. nov. Figure 4 Hypseleotris simplex - Allen, 1975: 95 (in part, Prince Regent Reserve). Holotype WAM P25028-009, male, 32 mm SL, Wyulda Creek, about 2 km above junction with Roe River (approximately 15®26'S, 125'^37'E), Prince Regent Reserve, West Kimberley, Western Australia, G, Allen, 17 August 1974. 250 Douglass F. Hoese and Gerald R. Allen Paratypes AM 1.22744-001, 5 (26-34 mm SL), collected with holotype; WAM P25028-007, 12 (23- 36 mm SL), collected with holotype; WAM P25040-004, 14 (22-37 mm SL), Youwanjela Creek (approximately 15°34'S, 125^25'E), a small tributary of the Prince Regent River, West Kimber- ley, Western Australia, B. Wilson, 21 August 1974. Diagnosis A species of Hypseleotris closely related to H. aurea from the Murchison River of Western Australia. The characters which differentiate these species are dis- cussed in the diagnosis for H. aurea. It differs from H. compressa by the absence of preopcrcular pores and other features mentioned in the diagnosis for that species. Finally, it is separable from //. kimberleyensis and //. ejuncida on the basis of a combination of characters which includes a longitudinal scale count of 24-26; a moderately slender body, the depth at pelvic origin 16-20% of SL; pre- dorsal scales extending to behind eyes; modal gill rakers on outer face of second arch 10; colour pattern consisting of about 15-20 chevron marks on side; and pelvic fin origin positioned under posterior opercular margin. Description No head pores. Predorsal scaled forward to above middle of eye. Cheek with 2 or 3 rows of small embedded cycloid scales. Operculum covered with medium- sized cycloid scales. Lower attachment of gill opening below posterior preopercu- lar margin. First dorsal VI. Second dorsal usually 1,9-10. Anal usually 1,10. Typically an equal number of dorsal and anal rays. Pectoral usually 13-14. Segmented caudal rays 15; branched caudal rays 1 1, rarely 13. Longitudinal scale count 24-26. Transverse scale count (TRDB) 8-9. Predorsal scales 14-16. Post- dorsal scales usually 8. Gill rakers on outer face of first arch 3-4+1+9-10 = 12-15, usually 12-14. Vertebrae 25. Pelvic fin moderately long, 18-22% of SL. Body slender, body depth at anal origin and depth at pelvic origin 16-20% of SL. Snout short, less than eye diameter. Mouth small, forming an angle of about 45° with body axis; jaws ending under posterior nostril. First dorsal origin about an eye diameter behind pelvic origin. Pelvic origin below posterior opercular margin. First dorsal fin low, with a rounded margin, height less than body depth; fourth and fifth spines longest. Second dorsal fin subequal in height to first dorsal fin, with posterior rays longest, particularly in males. Caudal short with a rounded to truncate margin. Pectoral rays, except upper 2 or 3 and lower 1 or 2, branched once. Anal rays sometimes with a second branch near tip. Pelvic fins long and pointed, reaching to anus. Caudal peduncle long, length greater than second dorsal fin base. Head dark brown, usually darker than body in males. A black bar at base of pectoral fin extending vcntrally to opposite third to seventh pectoral ray. Body with 15-20 thin black chevron marks along midside. A black spot on pos- terior end of caudal peduncle below midside. Basal one-third of dorsal and anal fins dark brown to black (sometimes with white spots on second dorsal fin) 251 A Review of the Gudgeon Genus Hypseleotris followed distally by a thin white stripe, followed distally by a broader black stripe; extreme distal tips white. Figure 4 Hypseleotris regalis, male holotype, 32 mm SL, Wyulda Creek, Western Australia. Sexual Dimorphism Males appear to reach a larger size, although only six females were examined (23-30 mm SL) and 21 males (27-36 mm SL). In males the first dorsal membrane reaches almost to the base of the second dorsal fin, but the two fins are widely separate in females. Males have the posterior dorsal and anal rays prolonged, with the last ray longer than the third ray. In females the last ray is about two-thirds length ol the third ray. The caudal fin margin is rounded in males and truncate in females. The latter sex is generally lighter in colour. Distribution and Habitat Hypseleotris regalis is known only from the Upper Roe River and Youwangela Creek, in a remote section of the Prince Regent Reserve. The two populations arc identical with regard to coloration, although there are slight differences in fin-ray, scale, and gill raker counts. The holotype and majority of paratypes were collect- ed with rotenone ov^er rock boulder substratum in a quiet, clear pool approxi- mately 4 X 10 m with a maximum depth of about 2 m. Etymology The species is named regalis (Latin for ‘regal’ or ‘royal’) with reference to the name of the type locality area, Prince Regent Reserve. Hypseleotris kimberleyensis sp. nov. Figure 5 Holotype WAM P25454-009, male, 34 mm SL, Barnett River near Barnett Gorge (approximately 16®32'S, 126°08T), Central Kimberley, Western Australia, B. Hutchins and A. Chapman, 30 July 1975. 252 Douglass F. Hoese and Gerald R. Allen Paratypes AM 1.22742-001, 3 (31-35 mm SL), collected with holotype; WAM P25454-007, 6 (27-34 mm SL), collected with holotype; WAM P25872-006, 4 (28-34 mm SL), Manning Creek Gorge (approximately 16°38'S, 125‘^55'E), Centred Kimberley, Western Australia, G. Allen and G. Evans, 15 September 1977. Diagnosis A species of Hypseleotris closely related to//, ejuncida from the Prince Regent River of Western Australia. These species differ from other Western Australian members of the genus by a combination of characters which includes the absence of preopercular pores, longitudinal scale count 28-32; body relatively slender, the depth at pelvic fin origin 14-18% of SL, and the pelvic fin origin under or behind the pectoral fin insertion. Hypseleotris kimberleyensis differs from H. ejuncida on the basis of the following combination of characters: predorsal and operculum without scales; second dorsal and caudal fins without distinct spotting or wavy bands; dark bar on pectoral fin base developed only on dorsal portion, usually above uppermost pectoral ray; body depth at anal fin origin 13-17% of SL; and vertebrae 26. Description No head pores. Head entirely naked". No median nape scales before first dorsal fin. Body scaled, a naked patch under first dorsal fin from upper pectoral base to second dorsal origin. Ventral surface of belly naked. Gill opening extends forward to below a point just in front of posterior preopercular margin. Pelvic origin behind pectoral insertion. First dorsal VI. Second dorsal usually 1,10. Anal usually 1,10. Pectoral rays 14-15. Segmented caudal rays 15; branched caudal rays 11, rarely 12, Longitudinal scale count 29-32. Transverse scale count (TRDB) usually 9. Postdorsal scales 7-9. Gill rakers on outer face of first gill arch 1-3+H-7-9 = 9-12. Vertebrae 26. Pelvic length 18-22% of SL. Body depth at anal origin 14-17% of SL. Body depth at pelvic origin 13-17% of SL. Snout short, subequal to eye. Mouth small, forming an angle of about 50^ with body axis; jaws ending under posterior nostril. First dorsal origin about an eye diameter behind pelvic origin. First dorsal fin low, with rounded margin, height less than body depth; fourth and fifth spines longest. Second dorsal fin subequal in height to fii'st dorsal fin; posterior rays sometimes prolonged ranging from two-thirds or equal to length of third ray. Caudal fin with slightly rounded to truncate margin. Pectoral rays seven to 1 1 branched, with one branch point each; middle rays longest. Second dorsal and anal rays branched once, except for first segmented ray, which is unbranched; other anterior rays sometimes with a second branch point near tip. Pelvic fins long and pointed, reaching about to anus. Caudal peduncle long, length greater than second dorsal base. Head dark brown, darker than tan body. Head of females lighter, often with dark brown on top of head only. A black bar at base of pectoral fin from upper margin to opposite first to fifth pectoral ray. Body with 4-7 narrow dark brown bars anteriorly from first dorsal fin to middle of second 253 A Review of the Gudgeon Genus Hypseleotris dorsal fin, more distinct anteriorly. Scales edged in dark brown, edgings darkest on midside forming a row of X-shaped marks just below midside. A small dark brown spot on posterior end of caudal peduncle just below midside. Median fins dark brown to black in males, dusky in females. Second dorsal and anal some- times with a thin whitish longitudinal stripe just below middle of fin; distal margin usually lighter than rest of fin, but without a distinct white margin. No white spots on fins. Pectoral and pelvic fins dusky to clear. Caudal fin sometimes with two thin vertical wavy lines near base. Figure 5 Hypseleotris kimberleyensis, male holotype, 34 mm SL, Barnett River, Western Australia. Sexual Dimorphism Adults of both sexes lack a prominent hump on the forehead. Of the five females examined the largest is 31 mm SL, and the nine males range from 30 to 30.5 mm SL. Males have the two dorsal fins separated by about 2 or 3 scale rows; in females the separation is slightly wider, about 3 or 4 scale rows. Males collected in spring (September) are characterized by prolonged posterior dorsal and anal rays, but in males collected during winter (July), these rays are not prolonged. The caudal margin is rounded in males and distinctly truncate in females. The anterior 3-4 bands on the side are distinct in males, but in females 6-7 bands are evident. In addition, males are typically darker than females. Distribution and Habitat This species is known only from the Barnett River system in the vicinity of Mount Barnett Station (16’^40'S, 125‘^57'E) which is situated on the Gibb River (Derby to Wyndham) Road, approximately 260 km north-east of Derby, Western Australia. The type specimens were collected from clear, rocky pools with moder- ate flow. Etymology The species is named kimberleyensis with reference to the type locality, the Kimberley district of Western Australia. Douglass F. Hoese and Gerald R. Allen Hypseleotris ejuncida sp. nov. Figure 6 Holotype WAM P25032-009, male, 46 mm SL, Gundarara Creek, about 2 km above junction with Prince Regent River (approximately 15°49'S, 125®37'E), Prince Regent Reserve, West Kimber- ley, Western Australia, G, Allen, 21 August 1974. Paratypes (all collected with holotype) AM 1,22745-001,4 (34-37 mm SL); WAM P25032-002, 8 (23-43 mm SL), Diagnosis A species of Hypseleotris closely related to H. kimberleyensis. The reader is referred to the diagnosis for the latter species for a discussion of the differences between these two species and other Western Australian members of the genus. Hypseleotris ejuncida differs from H. kimberleyensis on the basis of the follow- ing combination of characters: predorsal and operculum usually extensively scaled; second dorsal fin with white spots and wavy bands; caudal fin with dark spots forming 4-6 wavy bands; dark bar covering entire pectoral fin base; body depth at anal fin origin 1 7-18% of SL; and vertebrae 25. Description No head pores. Top of head scaled forward to just behind eyes, midline of nape often scaled at least partially. Operculum with large cycloid scales. Cheek naked. Belly covered with cycloid scales. Body scales largely ctenoid, cycloid in patch under first dorsal fin and on head. Gill opening extends forward to below posterior preopercular margin. Pelvic origin just below or behind dorsal pectoral insertion. First dorsal VI. Second dorsal 1,9-10. Anal 1,9-1 1, usually 1,10. Pectoral rays 14-15. Segmented caudal rays 15; branched caudal rays 11-13, usually 11. Longitudinal scale count 28-31. Transverse scale count (TRDB) usually 8 or 9. Predorsal midline scales 2-20, scales at side of midline 14-20. Postdorsal scales usually 8. Gill rakers on outer face of first gill arch 2-3+1+8-9 = 11-13. Vertebrae 25. Pelvic length 17-22% of SL, usually 19-21%. Body depth at anal origin 17-18% of SL. Body depth at pelvic origin 16-18% of SL. Snout short, about equal to eye diameter. Mouth small, forming an angle of about 45*^ with body axis, jaws end under a point just in front ot posterior nostril. First dorsal fin origin about an eye diameter behind pelvic origin. First dorsal fin low, height less than body depth; fourth and fifth spines longest. Second dorsal slightly higher than first; posterior rays prolonged in males, longer than third ray; short in females, about two-thirds length of third ray. Caudal fin with rounded to almost truncate margin. Pectoral fin with 8-13 branched rays, usually 11-13; middle rays often with 2 or 3 branch points. Second dorsal and anal rays (except for first segmented ray) with one branch point each in small specimens; often with a second branch point near tip in larger specimens. Pelvic fins long and with pointed tip reaching almost to anus. Caudal peduncle long, length greater than second dorsal fin base. Head dark 255 A Review of the Gudgeon Genus Hypseleotris brown in males, lighter in females, darker than tan body. Dark brown bar at base of pectoral fin darker dorsally, covering whole pectoral base. Body with 3 or 4 dark brown vertical bars between pectoral base and second dorsal fin origin. Body scales edged with dark brown, forming a diamond-shaped pattern on body. A longitudinal irregular dark brown stripe along side just below midside, formed from intense dark edges to scales; stripe often obscure in large dark males. A prominent dark brown spot at end of caudal peduncle, extending ventrally from midside. Dorsal and anal fins dusky to black. First dorsal fin with a pale whitish median longitudinal stripe. Base of second dorsal fin with 1 or 2 rows of white spots, prominent in males, sometimes obscure in females; membranes between rays with dark pigment forming almost vertical thin stripes, which cross fin rays. Second dorsal and anal fins with a pale whitish margin. Caudal fin with 4 to 6 thin dark brown wavy vertical lines, more prominent in females. Pectoral and pelvic fins clear to dusky. Figure 6 Hypseleotris ejuncida, male holotype, 46 mm SL, Gundarara Creek, Western Aus- tralia. Sexual Dimorphism Adults of both sexes lack a hump on the forehead. The six females examined ranged in size from 23-36 mm SL and the seven males 30-47 mm SL. The posterior membrane of the first dorsal fin of males reaches almost to the second dorsal fin origin, but in females the membrane is separated from the second dorsal fin by two scale rows. In addition, the posterior dorsal and anal rays are pro- longed and the caudal fin is more rounded in males. Females are generally lighter coloured, with the anterior body bands more distinct than in males. Distribution and Habitat This species is known only from the Prince Regent River drainage, situated approximately 260 km north of Derby, Western Australia. The type specimens were collected with rotenone from a clear, quiet pool over sandstone bottom at depths to about 2 m. Etymology The species is named ejuncida (Latin for ‘slender’) with reference to the slender body shape. 256 Table 1 Vertebral counts of species of Hypseleotris from Western Australia. An asterisk indicates count of holotype. Douglass F. Hoese and Gerald R. Allen a CO o d VD Oh 0 ^ 3 O S| ct3 O O C Q < V 2 H 257 Table 3 Longitudinal scale count in Hypseleotris from Western Australia. An asterisk indicates count of holotyp' A Review of the Gudgeon Genus Hypseleotris CO c 3 CO O CO U (J c/3 a D 03 a s Oh