Heleioporus australiacus flavopunctatus, Mahony & Penman & Bertozzi & Lemckert & Bilney & Donnellan, 2021

Mahony, Michael J., Penman, Trent, Bertozzi, Terry, Lemckert, Frank, Bilney, Rohan & Donnellan, Stephen C., 2021, Taxonomic revision of south-eastern Australian giant burrowing frogs (Anura: Limnodynastidae: Heleioporus Gray), Zootaxa 5016 (4), pp. 451-489 : 479-483

Heleioporus australiacus flavopunctatus subsp. nov.

Southern Owl Frog

Figs 6 , 7 , 8 , 13 .

Holotype. NMV D56515, an adult male collected Yalmy Road, 1 km ENE of Waratah Road junction, East Gippsland , Victoria -34.65° 150.53°, on 25 January 1983 by Mark Hutchinson.

Specimens examined. See Table 1 and Supplementary Table S1 for details of specimens examined.

Measurements of holotype. (in mm), SVL 81.4, HW 36.5, HL 25.9, ED 11.5, EN 5.9, IND 7.43, TL 33.9, FL 42.0, IMT 5.5.

Description of the holotype. The holotype (NMV D56515) is a large (SVL 82.4) adult male ( Fig. 13F ). The body is rounded and robust, head wide (HW/HL 1.41), with short and robust limbs (TL/SVL 0.41). Arms and legs are thick and muscular with the girth of the arms exceeding that of the legs. The head is rounded when viewed from above and deep when viewed laterally. The eyes are large and prominent (ED/EN 1.95). The pupil is vertical ( Fig. 13F ). Tympanum large and circular slightly smaller than diameter of the eye, covered by skin with small tubercles. Sternum divided posteriorly.

Dorsal surfaces, including limbs and head, granular with many small tubercles. Upper flanks with larger tubercles ( Fig. 13B ). A glandular ridge, comprising a series of tubercles with a larger tubercle at the axil, extends from the side of the face below the midline of the eye to the axil. Scattered tubercles also occur on the chin, chest, and posterior and inner surfaces of the limbs. Most tubercles with a central, fine, black keratinized spine. Ventral surfaces are smooth. The cloaca is almost completely surrounded by round or elongated large tubercles and skin folds ( Fig. 13H ). Smaller rounded tubercles below the cloaca.

Hands are wide with short, rounded fingers and without webbing. The fingertips rounded and ventrally resemble a tubercle but are not disc shaped, and two larger tubercles occur at the base of each finger. Palmar surface with a prominent inner-metacarpal tubercle at the base of finger one, with large tubercles at the base of each finger and smaller tubercles on the palm and finger joints ( Fig. 13I ). Slight lateral fringe on fingers two, three and four. Proximal half of finger one is almost twice as thick as the other fingers. Fingers sequence in length 3> 1> 2> 4.

Distinct, large, black, keratinized, conical, nuptial spines on the upper surface of the first three fingers. On the first finger a large spine occurs at the junction of the first phalange, and three smaller spines occur along the second phalange. Six spines, on second and third fingers, smaller than on first finger.

The foot is broad, and the plantar surface is smooth. Upper surface slightly lighter tone than dorsum, without tubercles or spines. Toes short and robust, with rounded tips, and in size 4> 3> 5> 2> 1. No webbing between the toes. Large non-keratinized inner metatarsal tubercle (IMT) much longer than the fifth toe, rounded on the apex, contrasting in colour to the plantar surface of foot ( Fig. 13J ). Large and forwardly inclined pointed tubercles at the base and at the joints on each toe. Tubercles between the junctions of all toes except 4 and 5 ( Fig. 13J ).

Pre-orbital papillae in the anterior corner of the eye is bilobed. The vomerine teeth in two prominent transverse series extending from the midline to the posterior median border of the choanae.

Variation. Variation in body dimensions is summarized in Table 5. A large (SVL females mean = 76.6 mm, range 64–87; males mean = 78.3 mm, range 59–91), rotund frog with stout body and short, stout limbs (TL/SVL 0.32–0.44). Head broad (HW/HL 1.15–1.5) and short and rounded when viewed from above. Viewed from the side, the head is deep and blunt, the eyes prominent (ED/EN 1.07–2.35). The snout is short, with the nostrils closer to eyes than tip of snout. Tympanum large, diameter approximately two-thirds the length of eye, circular and distinct, often with small tubercles. The inner metatarsal tubercle is elongated and of uniform width in most individuals but occasionally can be broader and flap like in some specimens. Pre-orbital papillae in the anterior corner of the eye either smooth or lobed-shaped and with a variable number of tubercles among individuals ( Fig. 8E–H ). Males do not have a vocal sac.

Males and females are significantly different in body size, and males have a significantly larger HW, HL, TL, FL, ED and EN ( Table 6). Males in breeding condition have enlarged forelimbs which exceed the girth of the hindlimbs ( Penman et al. 2006a). Males with a series of conical black nuptial spines on first, second, and occasionally third fingers; largest spine with bony core on metacarpophalangeal knuckle of first finger; smaller but multiple spines on second and third fingers. These nuptial spines are absent on females although smaller spines on the upper surface of the fingers do occur in females. In males in breeding condition many of the tubercles on the dorsal surface, flanks, upper surfaces of the limbs, on the ventral surface around the pectoral girdle and under the chin, the glandular area from the corner of the jaw to under the tympanum, and on the posterior of the thigh and around the cloaca have fine black keratinized spines. Females may also have spines, but they are smaller and not so widely distributed across the body as in males.

Colour and pattern. Colouration and texture of the skin does not differ between the sexes. Dorsal surfaces of the body, head, limbs and hidden surfaces of the thigh and groin are brown to bluish black without larger blotches or patches. The tympanum, dorsal surfaces of the hands, feet and soles are a lighter shade ( Fig. 6I–P ). The metatarsal tubercle is cream to brown, with other tubercles cream coloured. There are often poorly defined yellow or cream patches on the upper surfaces of the femur and humerus. Upper surfaces of hands and feet with lighter white mottling, with a yellow spot on top of hand. There are many small raised rounded tubercles on the dorsal surface of the body that are slightly darker in tone on these surfaces and overall provide a slightly shagreened appearance.

The dorsal colouration extends to mid-way down the flanks where it gradually transitions to the cream colour of the ventral surfaces. Above the insertion of the arm is also cream coloured. There is a slightly lighter toned stripe from tip of snout onto the head, and the head is lighter tone than the back. On the lateral surfaces the tubercles are larger than on the dorsum and in many the centre has canary yellow to cream spots. The number of spots varies considerably (mean = 31 in females, 32 in males) and in numerous cases the spots coalesce to form larger spots. The mean size of the coloured spots on the flanks is 1.7 mm in females and 2.6 mm in males. The spots are larger towards the groin. At the transition with the venter the larger tubercles and intervening skin are cream coloured.

A prominent, yellow coloured, glandular ridge of larger interconnected tubercles extends posteriorly along the upper maxilla from below the eye, extending beneath the tympanum and terminating above the axil. The margin of the mandible has many fine tubercles and is yellow or white.

Ventral surfaces, including limbs, white to bluish white, tending towards transparent in the region of the chest. Throat with faint yellowish mottling, frequently washed with brown in males, immaculate in females.

Soles and palms lightly to darkly pigmented with the same colour as dorsum. The metatarsal tubercle is cream to brown, with other tubercles cream coloured.

The tubercles surrounding the cloaca with yellow or white spots that coalesce to form a ring with a small gap at the top and bottom in most specimens ( Fig. 7I–P ).

The iris is finely reticulated silver grey ( Fig. 13F ). The nictitating membrane is transparent with a sharp black line on the upper margin.

Male calls. The call consists of a single note with distinct pulses and meets the definition of the pulse repetition type with dense harmonics in the classification of Beeman (1986). The call is amplitude modulated with a gradual rise in amplitude energy to the third or fourth pulse followed by a gradual decline. We observed considerable variation in temporal attributes (see Fig. 9 ). Duration of the call is just under half a second (mean 0.45, range 0.23– 0.57 s), and comprises 4 to 6 discrete pulses, of short duration (mean 0.052, range 0.04– 0.06 s). In several individuals the length of the last pulse is longer (mean 0.06 s) than the preceding pulses. The pulse repetition rate varied considerably (mean 9.83, range 7.11–13.92), as did the call repetition rate (mean 28.33 s, range 16.00– 60.00 s) ( Table 8). The call is broadband with most energy contained within a low dominant frequency (mean 373.5 Hz, range 301.5–430.7 Hz), without evidence of frequency modulation.

The aggressive call was recorded at three locations, two in the south at Broadwater State Forest, NSW, and another at Budawang in Morton National Park, NSW ( Hoskin et al. 2015). The aggressive call is a pulsatile-harmonic sound ( Beeman 1998), with a combination of tonal components with moderate to strong amplitude modulation ( Fig. 12 ). It is a single drawn out tonal note that sounds like a soft “whooop” or “moan” of variable duration of 0.35 to 1.48 s. The dominant frequency (mean 344 Hz) was the same as the advertisement call of the male making the call ( Table 10). There is no distinct harmonic structure or frequency modulation in the spectrogram and the sound is broadband (range 172–516 Hz). In addition to this call there is a longer drawn-out single note pulsatile call which we assume is also an aggressive call ( Fig. 12 ).

It is most common to find solitary males in nature, however, choruses of two to three males are encountered, and, in these cases, males are often spaced along a suitable breeding pool ( Penman et al. 2006a). Although rarely observed or reported, several lines of evidence indicate that males engage in combat over suitable ovipositional sites similar to that described in numerous other frogs (Wells 2010). Males have large nuptial spines and large muscular forearms which provides evidence of sexual selection as described in other frogs ( Katsikaros & Shine 1997, Wells 2007), and evidence of combat can be seen by the numerous scratches that occur on the dorsal surface of adult males, but which do not occur on females. The spines may also assist in restraining the female during oviposition. It is also possible that the large spines of males are used in defence against predators since the males will strongly grip and push the spines into a hand when handled.

Males also make a distinctive “distress call” sensu Wells (2007) ( Table 10, Fig. 14 ), that is elicited when adults are handled or attacked by a predator and is accompanied by an “unken reflex” or diematic display ( Umbers et al. 2015), a behaviour whereby the frog hyper-inflates the lungs, dips it head and raises up on its hind limbs to expose its thighs and cloacal region. Our observations are that in some cases the male will stand high on all limbs, with the mouth open widely “screeching” and jump towards the potential threat. This behavioural repetoir involves exposing the brighter spots of colouration around the cloaca and may also be used as a warning as Heleioporus produce copious glandular secretion (Umbers 2016). Penman et al. (2007b) observed that red-bellied black snakes ( Pseudechis porphyriacus ) prey on adults and this defensive display and the increase in body size hinders gape-limited predators. We have analysed one distress call from H. a. flavopunctatus subsp. nov. (Narooma, - 36.221 S), and this call has also been heard on numerous occasions in populations of H. a. australiacus in the Sydney Basin bioregion (FL, MJM, pers. obs.). This call is classified as a spectrally-structured pulsatile sound ( Beeman 1998, Köhler et al. 2017), and a count of individual pulses is not possible. The call is relatively long (5.4 s), and notes are emitted over a wide frequency band with two or three spectral peaks which appear to be harmonics, with small, stepped rises and drops in frequency (e.g., 1033 to 2239 to 3617 back to 2239 Hz), associated with increased amplitude on several occasions in the call ( Fig. 14 ). These frequencies are significantly higher than those produced in the advertisement call. Alternating amplitude modulation is apparent in the oscillogram, but without discrete energy peaks that could unambiguously be referred to as pulses.

Eggs and Larvae. The eggs and larvae have been described by Watson & Martin (1973), based on four egg masses from 12 km south of Walhalla, Victoria (NMV D5921). Daly (1996) reported the number of eggs per clutch and site of egg deposition for three clutches from Vincentia near Jervis Bay NSW, and Anstis (2017) included these in her description along with observations from locations that are now referable to H. a. australiacus . Eggs are unpigmented and deposited in a foamy mass ( Daly 1996, Anstis 2017). Eggs are about 2.6 mm in diameter (Martin & Watson 1973). Egg counts show a level of variation with values of 775 to 1239 (n = 2) Walhalla Victoria (Martin & Watson 1973), and 698 to 807 (n=3) from Vincentia, NSW ( Daly 1996). In all reports egg masses have been observed in shallow water, mostly commonly in low flow situation. Egg masses are found in a variety of scenarios including in standing water concealed in a burrow and in slowly flowing water ( Watson & Martin 1973), in a crevice on the edge of a pool (RB pers. obs.), and in shallow water beneath dense vegetation ( Daly 1996, see photo in Anstis 2017, p. 360).

Males have been observed calling in different positions and not always in a burrow as is the case in all other Heleioporus ( Lee 1967, Bailey & Roberts 1981, Anstis 2017). Littlejohn & Martin (1967) observed one male calling from a partially flooded burrow at the edge of a stream and another from a more open position in the middle of a pool partly concealed by plant debris. Other observations are of males calling from shallow water in open positions ( Bilney 2015), and in shallow water under dense reeds ( Daly 1996; see photo in Anstis 2107). The common feature is that the males call from shallow water, and that some form of concealment is favoured, although not necessarily in a burrow. Outside of breeding times our observations are that males will often sit with the head held high with the front legs straight. Movement is often by walking rather than hopping.

Morphology of the tadpoles from Vincentia, NSW, was described by Anstis (2017). Larval development is aquatic, with an unspecialized morphology; the external gills are prominent with a median anus ( Anstis 2017). The tadpoles are large reaching a total length of 75 to 85 mm ( Gillespie 1990, Daly 1996, Anstis 2017). The minimum duration of the larval period in nature is 92 days ( Daly 1996), however, development often extends over winter from egg masses laid in autumn and the duration in these cases would be over six months ( Daly 1996, Anstis 2017).

The literature suggests that some morphological variation may exist in the eggs and tadpoles between the two subspecies ( Penman et al. 2004). Ova from a clutch from Walhalla, Victoria were reported to be unpigmented (Littlejohn & Martin 1973), and those from the Sydney Basin region to be lightly pigmented ( Moore 1961, Daly 1996). Whether this difference is consistent between the subspecies has not been resolved, however, Anstis (2017) indicated that ovarian eggs in museum specimens collected from Victoria are pigmented, with a similar observation reported for specimens from Eden in far south-eastern NSW ( Penman et al. 2004), and it is therefore possible that the lack of pigmentation reported by Littlejohn & Martin (1973) represents some form of population or developmental variation.

There is also some variation in the larval tooth row formula. Lee (1967) reported four upper and three lower labial tooth rows, based on two specimens from Kanangra Walls (Australian Museum collection, registration numbers not reported) in the Sydney Basin bioregion, NSW (deep within the range of H. a. australiacus ). However, Watson & Martin (1973) reported six upper and three lower labial tooth rows based on specimens from Walhalla, Victoria. This difference may be due to ontogenetic changes during development, since Anstis (2017) observed that tadpoles at Gosner stages 26 to 27 had four upper and three lower rows, and that a fifth upper row develops by stage 28 and many tadpoles eventually have six upper rows. The oral disc has been illustrated by Littlejohn & Martin (1973) and Anstis (2017). There is an anterior gap in the papillary border ( Watson & Martin 1973, their Fig. 3 A , Anstis 2017).

Etymology. The name flavopunctatus refers to the occurrence of yellow spots on the lateral surfaces and around the cloaca.

Distribution. Occurs from the Parma Creek Nature Reserve, NSW (latitude -34. 9886°), south to near Walhalla (-37.9412°) in eastern Victoria. See the account of the distribution for H. a. australiacus above for a discussion on the northern distribution limit of H. a. flavopunctatus subsp. nov. It is restricted to the coastal slopes east of the Great Dividing Range. Populations occur across a range of altitudes from near sea level (Broadwater State Forest -37.0088°, 149.8983°, 172 m elevation) to montane areas (> 800 m elevation at Bondi State Forest, NSW, -37.2090°, 149.2934°).

Conservation status. Based on the IUCN Red List criteria ( IUCN 2016, Mace & Lande 1991) we assess H. a. flavopunctatus subsp. nov. to be endangered A2(c)B2(a)(b) since the Area of Occupancy (AOO) is less than 2000 km 2, and there is evidence of population reduction from mapping of records over three generations ( Fig. 15 and Supplementary Information 2). Although H. a. flavopunctatus subsp. nov. is known over a relatively large range (Extent of Occupancy, 31,022 km 2), the number of location records and the reports of population sizes are low ( Table 11). The Atlas of Living Australia has records for 10 locations in the decade from 2010 to 2020, and 23 in the previous decade. Gillespie (1990) reviewed the conservation status of the Victorian, and southern NSW populations and noted that prior to 1983 only five specimens had been recorded in Victoria from three localities ( Fig. 15 ). In the period between 1983 and 1990 he listed 11 records, the majority of which were for single individuals. This was despite extensive fauna surveys in the East Gippsland region in Victoria, across various seasons over a three-year period, which included several field methods such as pitfall trapping (5400 pitfall nights), surveys of small streams and other waterbodies and spotlighting along tracks ( Gillespie 1990). This situation is also the case in southern NSW. Fewer than eight records of populations have been made in the past ten years, and most records are prior to this period ( Lunney & Barker 1986, Kavanagh & Webb 1998, Daly 2019, Lemckert et al. 1998, Penman et al. 2004) ( Fig. 15 ). In a five-year survey of the herpetofauna of the coastal forests near Bega for a total of 260 person days between August and December in 1980 and for 30 person days each December from 1981 until 1984, Lunney & Barker (1986) reported only one record. Two populations occur within reserved land in southern NSW, at Ben Boyd National Park (ALA) and Impressa Moor in Nadgee Nature Reserve ( Lunney & Barker 1986), and in eastern Victoria only one population is within reserved land, in the Snowy River National Park ( Gillespie 1990). The sub-species has been observed on several occasions in Ben Boyd National Park, NSW in the past ten years (G. Malolakis NSW NPWS, pers. comm.), and a nearby population at Broadwater State Forest has been observed consistently for>12 years (Forestry Corporation of NSW R. B. observations).

Figures

FIGURE 3. Clustering analysis of SNP variation in eastern Heleioporus. A) PCoA ordination of SNP diversity, B) STRUC- TURE barplot of k=2 showing ancestry, C) STRUCTURE barplot of k=3 showing ancestry. Numbers refer to localities detailed in Table 1 and Fig. 1.

FIGURE 6. Colour pattern profiles for adult individuals of A) plate from Shaw & Nodder (1795); B–H) Heleioporus a. australiacus, and I–P) H. a. flavopunctatus subsp. nov.

FIGURE 7. Examples of the colour patterning around the cloaca in eastern Australian Heleioporus. A–H) Heleioporus a. australiacus; A) AMS R27631 Darkes Forest NSW, B) SAMA R45377, C) Ourimbah NSW, D) AMS R78853 Pittwater NSW, E) SAMA R12064 Gosford, F) ANWC A0357 Wadbilliga, G) AMS R95406 Darkes Forest, H) AMS R104952 (neotype) Maroota; and I–P) H. a. flavopunctatus subsp. nov., I) Yambulla NSW, J) Yambulla NSW, K) Narooma NSW, L) NMV D50515 (type) Yalmy Road, Vic., M) NMV D33671 6 km S Ben Cruachan, Vic., N) NMV D33672 6 km S Ben Cruachan, Vic., O) no voucher state forest near Eden, NSW, P) AMS R136722 Coolangubra State Forest, NSW.

FIGURE 8. Variation in the pre-orbital papillae in anterior corner of eye of eastern Australian Heleioporus. Heleioporus a. australiacus—A) AMS R104952 (neotype) Maroota State Forest, B) AMS R5958 North Sydney, C) AMS R9470 Wollemi National Park, D) AMS R78853 Commodore Heights. H. a. flavopunctatus subsp. nov.—E) AMS R136719 Coolangubra State Forest, F) ANWC A0357 Wadbilliga, G) AMS R136730 Coolangubra State Forest, H) ANWC A1504 Bondi State Forest.

FIGURE 9. Differences in call repetition rate and pulse repetition rate related to temperature in eastern Australian Heleioporus. A) fast call, Morton National Park, NSW (60 calls per minute; 19°C), B) moderate, Ku-ring-gai Chase National Park, NSW (36 calls per minute, 17°C), two males, a and b, alternating calling; C) slow, Watts Creek, Vic. (24 calls per minute, 8°C).

FIGURE 12. Aggressive calls in eastern Australian Heleioporus. A) H. a. australiacus, Ourimbah State Forest, NSW. B) Enlargement of one call showing dense pulses, and C) enlargement of a call from H. a. flavopunctatus subsp. nov. from Morton National Park, NSW.

FIGURE 13. Images of eastern Australian Heleioporus types. A–E) Neotype of Heleioporus a. australiacus. AMS R104952 from Maroota State Forest, NSW. F–J) Heleioporus a. flavopunctatus subsp. nov. NMV D56515 from Yalmy Road, 1 km E.N.E. of Waratah Road junction, Gippsland, Victoria. Photograph in life P. Robertson.

FIGURE 14. Heleioporus a. flavopunctatus subsp. nov. aggressive call and distress call. A) Distress call from Broadwater State Forest NSW, and B) Aggressive call from Broadwater State Forest, NSW.

FIGURE 15. Historic distribution of eastern Australian Heleioporus by decadal sampling periods from the ALA (accessed May 2020). Circles show where records have declined in the far south east of the distribution of H. a. flavopunctatus subsp. nov. and in the Greater Blue Mountains area in the north west of the distribution of H. a. australiacus.