The Bea)>le. Rcetircis of the Northern Territory Museum of Arts and Sciences. 1991 8( 1); 1 -10 THE FROG FAUNA OF MELVILLE ISLAND, NORTHERN TERRITORY. MICHAEL J.TYLER', MARGARET DAVIES' AND GRAEME F. WATSON'' ‘Department of Zoology. University of Adelaide. Box 498 G.P.O., Adelaide. S.A. 5001. Australia. -Department of Zoology. University of Melbourne. ParkviUe. Vic. 3052. Australia. ABSTRACT Fourteen species of frogs were found on Melville Island. Details are provided of their distribution, biology atid relative abundance. The advertisement calls of Crinia remota. Splienophtyne adelphe and Uperolciu itnindala are documented. The frog fauna of Melville Island is compared with the faunas of Darwin, the Cobourg Peninsula and Groote Eyiandt. The significance of the apparent absence of Litoria rubella and the comparatively low numbers of Cycloratui aii.stralis is explored. Keywords: Amphibia, Hylidae, Leptodactylidae, Microhylidae. Cycloratui australis. Litoria hicolar. Litoria caerulca. Litoria nticrohelos. Litoria nasuta. Litoria pallida. Litoria roiliii. Litoria toriiieri. Crinia remota. Limnodynastes convexiuscidus. Linmodynastes ornatus. Notaden melanoscaphus. Uperoleia inundata. Splienophtyne adelphe. Melville Island, fauna list, species distribution. INTRODUCTION to compare the fauna with the adjacent Cobourg Peninsula reported by Cogger and Lindner Because frogs possess a semi-permeable (1974), and with Groote Eyiandt in the Gulf of skin they are unable to cross marine barriers Carpentaria reported by Tyler ei al. (1986). other than by passive transportation. Accord¬ and Darwin, reported by Tyler and Davies ingly. insular Australian frog faunas represent (1986). peripheral, contintental populations, most of which were isolated in the Holocene or Late MATERIAL AND METHODS Pleistocene by marine transgressions upon the continental plate. Interest in island frog faunas The collection obtained has been deposited tends to focus upon the nature and extent of any in the Northern Territory Museum of Arts and divergence from mainland populations that Sciences (NTM) and the South Australian they exhibit. Museum (SAM). There is a large number of islands situated Specimens were collected at sites in close upon the Australian continental shelf. While proximity to unsealed roads and tracks on the some of these lack freshwater sources, and can western half of the island (Fig. 1). By day we be assumed to be unable to sustain frog popu¬ sought sites that we considered potentially lations, the majority have not been subjected productive, whilst at night we visited these to surveys designed to establish the existence sites and also slopped frequently elsewhere to and diversity of frog faunas. In fact the Irog listen for advertisement calls. Sample sizes fauna of the largest island. Melville Island off obtained rcllccted interest in laxa rather than the coast of Darwin, is unknown. our ability lo obtain specimens. In January 1990 we spent 10 days on Mel¬ Methods of measurement and the abbrevia¬ ville Island to determine the nature of the frog tions used in the text follow Tyler (1967). fauna. Beyond that documentation process, we Advertisement calls were recorded in the wished particularly to obtain data enabling us field using a Sony TC-D5PRO cassette re- M.J. Tyler, M. Davies and G.F. Watson Figs 1-2. 1. Melville and Bathurst Islands. Bold lines are roads and tracks along which collecting and recording was conducted. 2, (InscrtI Northern portion of the Northern Territory showing the physical relationships of the faunal sampling sites compared in Table 3. cordcr (tape speed 4.74cTri sec ') and a Beyer MELVILLE ISLAND M88 dynamic microphone. After locating a calling individual, the microphone was placed Melville Island is located approximately 60 close to the male and a series of calls recorded; km north of Darwin and is separated from an attempt was then made to collect the re¬ Bathurst Island by the narrow Apsley Strait. corded individual. Wet bulb air temperature Situated south of latitude 11° the island ap¬ was recorded at the terrestrial calling sites (no pears as an isolated portion of the Cobourg specimens reported here were calling in water) Peninsula (Fig. 2). using an electronic thermister thermometer The island is of low relief, with undulating (Takara Digimulti Model D611) with a resolu¬ laterite rises and dissected low plateaux up to tion of 0.1 °C. 100 m above .sea level. The most common .soils Recordings were analyzed on a DSP 5000 are Sandy Red Massive Earths (Gn. 2.11., 14 digital Sona-Graph (Kay Elemetrics) with play¬ seiisu Northcote et al. 1975). Adjacent to back on a Nakamuchi Dragon cassette tape the numerous perennial springs, soils are poorly recorder. Overall variations in tape speed (both drained and acidic. Narrow corridors of dense recording and playback) are estimated to be rainforest Oungle) Hank the streams, whereas less than 0.5%. Frequency responses of all on the greater portion of the island the vegeta¬ audio-electronic components are close to lin¬ tion is predominantly open eucalypt forest, tall ear within the relevant frequency range (2000- open shrubland or low woodland. 5000 Hz). The built-in set up No. 10 was used The wet season extends from October or for analysis on the Sona-graph. November to March or April. The recording of 2 Frog Fauna of Melville Island rainfall commenced in 1963, and pooled data three or four others at each of the sites. None indicate a mean rainfall of 1973 mm falling on was seen or heard around the houses at Picker¬ 9-21 raindays (mean) (Haines 1986). Of the taramoor, where we expected it to be abun¬ sites visited, Pickertaramoor had a mean rain¬ dant. fall of 1600 mm. Snake Bay 1563 mm, and Litoria microbelos Cogger Three-Ways 1964 mm (Van Cuylenberg and Dunlop 1974). Mean daily maximum and mini¬ Material. NTM R16091. SAM R35I62-64, mum temperatures are 18.2 and 32.9'’C. Yapalika; NTM R16092-98, SAM R35165-70, Melville Island is sparsely populated and Snake Bay. settlements are concentrated on the western Distribution and Abundance. This species half of the island. There is no land access to the was located at two sites where it was associ¬ east coast. ated with L. hicolor. Vast numbers of individu¬ als were heard calling. SYSTEMATICS Remarks. At each of the sites, groups of potential predators were congregated around Family Hylidae the calling frogs. At Yapalika L. nasuta was Cyclorana australis (Gray) found perched on dry grasses up to 0.5 m above the ground: these frogs were not calling but Material. NTM R16048-.50, SAM R351I5, apparently foraging with nearby L. microbelos 5-10 km N of Pickertaramoor; SAM R35117, the likely prey. At Snake Bay as many as five Poonellic Rd; SAM R351I6. nr Pickertara¬ spiders (Dolomecles sp. ?) would be found moor; NTM R16051-52, R16082, 5 km N of gathered within 10 cm of frogs that were Poonellie. calling from floating debris on the surface of a Distribution and Abundance. The speci¬ swamp. Dolomecles is a known predator of mens collected were the only ones sighted on frogs (McKeown 1943). the island. Wc were surprised at the scarcity of the species, for at localities in northern Austra¬ Litoria nasuta (Gray) lia that we have visited C. australis is invari¬ ably abundant and one of the most conspicuous Material. NTM R16063-65, SAM R35157- components of the terrestrial fauna. 59, 1.5 km N of Three-Ways; NTM R16067, 5 Remarks. A small chorus was heard at km NW Pickertaramoor; SAM R35I60, Snake Pickertaramoor during rain on the night of Bay; SAM R35161, Yapalika. 23.1.90. A specimen collected on the road on Distribution and Abundance. This species 21.1.90 regurgitated an adult L. rothii. appeared widely dispersed and was heard at a number of localities intermediate between those Litoria hicolor (Gray) listed above. Choruses of 1-2 dozen individu¬ als were heard on several occasions. Material. NTM R16083-85, SAM R35138- 40, Yapalika; NTM R16086-90, SAM R35165- Litoria pallida 70, Snake Bay. Davies, Watson and Martin Distribution and Abundance. The habitats of the two localities at which we found this Material. NTM R16070-71, SAM R35I29- species were both adjacent to permanent wa¬ 31,5 km NW of Pickertaramoor; NTM R16068- ter. At each site very large numbers of indi¬ 69, SAM R35132-33, 1.5 km N of Three-Ways. viduals were present. Distribution and Abundance. We heard and collected the species only at the two locali¬ Litoria caerulea (White) ties listed. Following rain large choruses con¬ gregated around temporary pools there. Material. One specimen was collected ap¬ proximately 24 km E of Three-Ways, and one Litoria rothii (de V^is) approximately 10 km N of Three-Ways. The specimens have been retained alive. Material. NTM R16060, Yapalika; NTM Distribution and Abundance. L. caerulea R16059,1.5 km N of Three-Ways; NTM R16061, is patchily distributed on the island. Wc saw SAM R35I23-24, 5 km N of Poonellie; NTM only the two specimens collected and heard R16062, SAM R35125, Snake Bay. 3 M.J. Tyler, M. Davies and G.F. Watson Distribution and Abundance. Although sive permanent marsh at Snake Bay. where widely distributed, we encountered no more approximately 12 males were calling from than six individuals at any locality. refuges in shallow water. Two were also heard (but not collected) at a freshwater swamp Litoria tornieri (Nieden) adjacent to the beach at Pooncilie. Material. NTM R16077-81. SAM R35I46- IJmnodynastes ornatus ((Jray) .31,5 km NW of Pickertaramoor; NTM R16072- 76. SAM R3.31.32-.56. I..3 km N of Threc- Material. NTM R16056. 1 km N of Three- Ways. Wavs; NTM R16057, SAM R35122; Paru Road; Distribution and Abundance. This species NTM R16058. Three-Ways. appears to exist in dense, discrete and widely Distribution and Abundance. The number separated communities. of specimens collected reflected the scarcity of the species. However we did not hear calling Family Leptodactylidae and our vouchers therefore were obtained for¬ Crinia remota Tyler and Parker tuitously. Material. NTM R 1610.3-04. SAM R.351.34- Noladen melanoscaphus Hosmer 37. Yapalika: NTM R16102. 1.5 km N of Three-Ways: NTM R16105. Snake Bay. Material. NTM R16055. SAM R35118-19, Distribution and Abundance. The species Yapalika. An additional eight specimens from also was heard at the springs that form the the same locality have been retained alive. source of the Takamprimili Creek approxi¬ Distribution and Abundance. We encoun¬ mately 3 km west of Pickertaramoor. It was tered this species only once following heavy very common in Hooded grasslands within a 5 rainfall at Yapalika. km radius of Yapalika. Remarks. Frogs were calling on 19.1.90 Call. Table 1 shows a comparison between from clear water no more than 2 cm deep. the published values of call attributes of C. Amplexus was observed, and one amplectant remota (Tyler and Parker 1974) and those pair produced fertile eggs. obtained from a call of an individual at Yapalika. Clearly the calls are very similar and confirm Uperoleia innndata the identity of the species as C. remota. Tyler, Davies and Martin IJmnodynastes con vexiusculns Material. NTM R16099-101, SAM R35126- (Macleay) 28. Yapalika. Distribution and Abundance. Occasional Material. NTM R16053. SAM R35120. calling individuals were heard at various points Yapalika: NTM R16054. SAM R35I21, Snake between Three-Ways and Snake Bay. but it Bay. An additional four specimens from Snake was only at Yapalika that we heard a large Bay were retained alive. chorus. Distribution and Abundance. The site at Call. Table I shows a comparison between which we heard most specimens was the exten¬ the published values of call attributes of U. table I. Comparison of a call of Crinia and Uperoleia from Melville Island with published values for C. remota and U. innndata. The published values for U. innndata shows the ranges and means (in italics) of call attributes. Species No. of Duration Pulse repetition Dominant Effective Pulses rate Frequency Temperature (ms) (pulsc.s/s) (Hz) (°C) C. remota ' 14 720 18.1 4250 Crinia. Yapalika M.I.- 16 675 7? 4 4280 26.3 U. innndata' I1-/2.5-I4 I27-/45..5-I60 80.()-S5.9-89.7 2.40()-26(;0-28()() 24.0-26.0 Uperoleia. Yapalika M.l.' 12 144.1 76..4 2370 26.3 (' Tyler and Parker (1974), this study. 'Tyler et al. (1981)) 4 Frog Fauna of Melville Island Fig. 3. Splienniyhryne aciciphe. Fickcrtaranioor, Melville Island. iuundata (Tyler ct cil. 1981) and those obtained calling approximately 30 cm above the ground from acall ol'an individual at Yapalika. Clearly from the low branches of a shrub. Traditionally the calls are very similar and confirm the Australian microhylid frogs are considered to identity of the species as iinindata. be rainforest species. Our observations dem¬ onstrate that .S’, adelphe is an exception. Family Microliylidae As reported by Zwcifel (1985) the call con¬ Sphenopliryne adelphe Zvveifel sists of "a moderately rapidly repeated series (Figs 3, 5) of high-pitched peeps” and is illustrated in Figure 4. Our more detailed analysis of calls Material. NTM R16106-11. S AM R3.‘S 109,14, from three individuals (Table 2) indicates similar Pickertaramoor. values to those provided by Zweifel (1985) Distribution and Abundance. We heard from calls of two individuals (one call incom¬ the distinctive call of this .species at almost plete), with no temperature data. every site that we visited. At Pickertaramoor Number of notes per call was relatively we collected it on the lawns around the houses. consi.slent within individuals and varied from Densities were as high as 6/100 ml One of the 24-30 (mean = 26.8) between individuals. The individuals is .shown in Fig. 3. lowest value (7) given by Zweifel (1985) seems Habitat. As indicated by the distribution, .S'. to be extremely low and likely to be from an adelphe occurred in a wide variety of habitats incomplete or interrupted call. Even so our in open country. Specimens were heard at the data suggest a higher number of notes per call periphery of plantations of Finns earihaea, is typical of the species and consistent with an and in rainforest. individual reported by Zweifel (see Table 2), Call. At Pickertaramoor individuals were but not included in his analysis. Similarly, calling in the open on leaves around houses. values for call duration are longer in our calls All but two were upon the ground either ex¬ than in Zweifel's analysis; these differences posed or beneath leaves. One specimen was are due to the lower number of notes per call in calling from the summit of a heap of grass Zweifel’s sample. Note durations vary within clippings at the base of a tree, and another was each call with the middle notes tending to be 5 M.J. Tyler, M. Davies and G.F. Watson Fig. 4. A, Wave form display of a call of Sphenophryne adelphe (SAM R3.5109). Pickertaramoor, Melville Island. Wet- bulb air temperature 25.4'’C. B. Wave form display of the middle note (16th) of the call of S. adelphe shown in A. C. Power spectrum of call of S. adelphe shown in A. Tabic 2. Comparison of calls of Sphenophryne from Melville Island with published values for S. adelphe (Zwcifel 1985), showing ranges and means (in italics). Locality No. of notes Call Note Note Repetition Dominant Wet Bulb per call Duration Duration Rate Frequency Temperature (s) (s) (notes/sec) (ig (°C) Croker I.. NT 1-14.6-10' 1.9-.?.56-4.8' c. 0.12 3.6-4.0S-4.3 c. 4600 unknown Pickertaramoor #1 30 7.83 Lst .11 4.90 3840 25.4 (SAM R.35109) (n=4:29-29.5-30) mid .098 last .10 Pickertaramoor #2 24 6.03 1st .11 3.89 3760 25.8 (not collected) (11=5:24-24.2-25) mid .098 last . 11 Pickertaramoor #3 29 8.53 1st .18 3.35 1520-4080-4400- 25.1 (not collected) mid .12 last .14 ' Values for a second individual available to Zweifel (1985) but not included in his analysis had a greater call duration (>7 sec) and number of notes per call (>31) but a similar note repetition rate (4.4). -Three peaks of equal intensity. shorter (Table 2); but the range of durations men of S. adelphe, and observed, “it would be extends over the value estimated by Zweifel of intere.st to know it S. adelphe has the orange (1985). Note repetition rate is similar in both sroin and axilla color noted lor Australian studies. Even though they are smaller, the ^racilipes". We can establish that the S. adelphe dominant frequency of Melville Island indi¬ collected by us lacks the orange patches, so viduals is generally lower than the Crokcr providing a simple key character to distinguish Island specimens, although the third individual the allopatric species. recorded had three equal peaks of energy in the Zweifel reported that the largest specimen power spectrum of its call (Table 2), and the that he examined was a female ot 21.9 mm S- upper value (4400 Hz) approaches that re¬ V. and stated that males attain at least 19.1 ported by Zweifel (1983). mm. The single female obtained by us was Remarks. In the original description Zweifel gravid at 18.0 mm, whilst the adult (calling) (1985) compared S. adelphe with the sibling males ranged 13.0-16.9 mm. Ranges of pro¬ species S. gracilipes (Fry), and noted that he portions employed by Zweifel (1985) are shown could not distinguish them on morphological below, with those cited by Zweifel in paren¬ grounds. Zweifel had not seen a living speci¬ theses; TL/S-V 0.40-0.49 (0.43-0.49), E-M/IN 6 Frog Fauna of Melville Island Fig. 5. Lateral views of heads of: A, male (SAM R35109). and B, female S. aclelphe (SAM R351 10). Note divergence in snout shape and tympanum size. 0..‘i()-0.92 (0.66-0.82), E/S-V 0.10-0.12 (0.10- of Litoria nihella. which is common and 0.12), E-N/S-V 0.04-0.07 (0.07-0.08). locally abundant over much of the northern The female collected by us differs morpho¬ half of the continent. We expected this spe¬ logically from the males. The tympanum in the cies to be there, were familiar with its call and specimen is extremely large and well defined arrived when we judged the climatic condi¬ (Fig. 5) whereas in the other material it is poorly tions ideal for its activity. But despite exten¬ defined and considerably smaller. In addition, sive field survey we did not hear or see this the snout is more pointed in profile and over¬ species at any site that we visited. hangs an undershot jaw (Fig. 5). The shape ol The absence of L. wotjithimensis is less the snout corresponds to the description of S. surprising. We did not locate any shallow gmcilipes of Zweifel (1985). It may be that rock-strewn creeks comparable to those where sexual dimorphism occurs elsewhere within we have encountered the species elsewhere. Spheiiophryiie, but we cannot confirm this. We were influenced to expect two species of Uperoleia largely because two species occur at Darwin and on Groote Eyiandt. Their DISCUSSION absence from the Cobourg Peninsula could be a consequence of inadequate collection dur¬ Our studies indicate that the frog fauna of ing the wet season. We are confident, how¬ Melville Island includes 14 species (Table 3). ever, that only one species occurs on Melville From our assessment of the habitat diversity Island. existing upon the island and the fact that our Litoria pallida was located on Melville sampling included all habitat types, this figure Island, but has not been reported from Co¬ coincides with our anticipations, based on work bourg Peninsula. The faunal survey of Co¬ on Groote Eyiandt and upon the adjacent main¬ bourg Peninsula by Cogger and Lindner (1974) land. antedated the description of L. pallida Davies Interest should centre more upon species that et al. (1983). There remains the possibility we failed to encounter on Melville island rather therefore that the species could have been than those that occur there. By far the most included unrecognised amongst the material conspicuous deficiency is the apparent absence reported as L. tonueri. Accordingly, the se- 7 M.J. Tyler. M. Davies and G.F. Watson Table 3. Faunal comparisons of four peripheral portions of the Northern Territory. Genus and Species Melville Cobourg Darwin' Groote Island Peninsula' Eyiandt' Cvclorana australis + + + Litoria hieolor + + + + Litoria caerulea + + + + Litoria microhelos + + + + Litoria nasiita + + + + Litoria pallida + - + - Litoria rothii + + + + Litoria ruhella - + + + Litoria tornieri + + + + Litoria wotjuhtmensis - + - + Crinia remota + +¥“ + Limnodvnastes cotn exiuscidus + + + Limnodvnastes ornaius + + + + Notaden melanoscaphiis + - + - Vperoleia immdata + - + + Uperoleia iithomoda - - + + Sphenophryne adelphe + -1-' - - 'Derived from Cogger and Lindner (1974) -Derived from Tyler and Davies (1986) 'Derived from Tyler el at. (1986) ¥“Further analysis is required to determine if these populations represent C. remora or C. hilini’iia 'Reported as ‘5. rohiisla' ries was examined by one of us (M.D.) but was Kerin), and Jeremy Russell-Smith who acted found to include only L. tornieri. as our guide during the early portion of our The presence of C. australis here contrasted visit. with its absence on Groote Eyiandt, where we Greg Sutton and John Gilham, of Melville expected to find it. We noted that it was scarce Forest Products Pty Ltd provided support and on Melville island, whereas it is abundant upon transport on the island, whilst in Darwin we are the adjacent mainland. We were unable to grateful to the North Australia Research Unit identify any factor that would provide an ex¬ of the Australian National University and par¬ planation for these differences. ticularly Jann King for facilities. There are various coefficients that can be used to express degrees of faunal similarity REFERENCES between discrete geographic units. This study is the second in an anticipated continuing Cogger, H.G. and Lindner, D.A. 1974. Frogs and series of studies investigating tropical, insular reptiles. In: H.J. Frith and J.H. Calaby (eds) amphibian faunas. Until a greater number of Fauna Survey of the Port Essiiifiton District. such studies has been completed we have not pp. 63-107. CSIRO Division of Wildlife Re- attempted to elucidate broader biogeographic .search Technical Paper 28. relationships and degrees of faunal similarity. Davies, M., Watson. G.F. and Martin. A.A. 1983. Redcl'inition of the Litoria laiopalniata spe¬ ACKNOWLEDGEMENTS cies group (Anura: Hylidae). Transactions of the Roval Societv of South Australia 107(2): 87-108. For penTii.ssion to visit Melville Island and Haines, M.W. 1986. An outline of the Melville to undertake our investigation there.we are Island plantation forestry project. Technical greatly indebted to the Tiwi Council. Our visit Memoranda 86/4. Conservation Commission was made possible by the support of the Aus¬ of the Northern Territory: Darwin. tralian Research Council and the Conservation McKeown, K.C. 1943. Vertebrates captured by Commission of the Northern Territory (CCNT). Australian spiders. Proceedintis of the Royal We thank particularly CCNT staff who helped Zoological Society of New South Wales (1942- 43): 17-20. us in a variety ot ways (Bill Freeland and Sarah 8 Frog Fauna of Melville Island Northcote, K.H.. Hubble, G.D., Isbell, R.F., Th¬ Tyler. M.J. and Parker, F. 1974. New species ompson, C.H. and Bettenay, E. 1975. A de¬ ofhylid and leptodactylid frogs from southern scription of Australian soils. C.S.I.R.O.: Glen New Guinea. Tran.sactions of the Royal Soci¬ Osmond. ety of South Australia 98(2):7l-78. Tyler. M.J. 1967. Papuan hylid frogs of llie genus Van Cuylenberg, H.R.M. and Dunlop. C.R. 1974. Hyla. Zoolofii.sclie Verliandelingen (96): 1-203. Land units of the seventeen mile plain, Mel¬ Tyler, M.J. and Davies, M. 1986. Prof’s of the ville Island. 1973. Animal Industry and Agri¬ Northern Territory. Conservation Commis¬ culture Branch. Department of the Northern sion of the NoUhern Terriiory:Alice Springs. Territory. Darwin. Technical Bulletin (14). Tyler. M.J.. Davies. M. and Martin, A.A. 1981. Wells, M.R.. Van Cuylenberg. H.R.M. and Dun¬ Australian frogs of the leptodactylid genus lop, C.R. 1978. Land systems of the western Uperoleia Gray. Australian .lonrnal of Zool¬ half of Melville Island. Northern Territory ogy, .Supplementary Series (14): 1-64. Land Conservation Unit. Territory Parks and Tyler, M.J.. Davies, M. and Watson, G.F. 1986. Wildlife Commission. Darwin. LC78/10mimeo. The frog fauna of Groote Eylandt, Northern Zweifel, R.G. 1985. Australian frogs of the family Territory, Australia. Zoological Journal of the Microhylidae. Bulletin of the American Mu¬ Linnean Society 88:91-101. seum of Natural History 182(3):265-388. Accepted 20 December 1990 9