Publications
NSW National Parks and Wildlife Service, July 2001
ISBN 0 731 36213 6
The Corroboree Frog (Pseudophryne corroboree) (Anura: Myobatrachidae) was described by Moore (1953) from a single specimen collected at Round Mountain (Colefax 1956). The collection site is now within Kosciuszko National Park. Until recently, only one species of Corroboree Frog was recognised (Cogger 1992). However, Wells and Wellington (1985) provided a brief argument for recognising the Northern Form as a separate species, which they named P. pengilleyi after Dr R. Pengilley who undertook substantial research on the species in the 1960's. More recently, Osborne et al. (1996) provided a detailed geographic analysis of variation in the morphology and calls of P. corroboree and recommended that P. pengilleyi be recognised. This recommendation has been generally accepted by other authorities (eg. Tyler 1997). Pseudophryne pengilleyi is also now recognised as a distinct species in the TSC Act (on Schedule 2, Vulnerable).
All reference to P. corroboree in this Recovery Plan refers to the Southern Corroboree Frog. Pseudophryne pengilleyi refers to the Northern Corroboree Frog, a species with a more extensive distribution in the Fiery Range and Brindabella Range (Osborne 1989) (Figure 1).
The Southern Corroboree Frog is distinctive and easily recognised because of its striking dorsal colour pattern consisting of bright yellow longitudinal stripes alternating with black stripes (Cogger 1992). The ventral surface is boldly marked with black and yellow and white blotches. A large flat femoral gland is present on each hind limb, and the inner metatarsal tubercle is low and round. Adults reach a length of between 25 and 30 mm. There are a number of differences between the Southern and Northern Corroboree Frogs, including considerable genetic divergence (Roberts and Maxson 1989; Osborne and Norman 1991), differences in colour-pattern and morphology (Pengilley 1966; Osborne et al. 1996) and skin biochemistry (Daly et al. 1990).
All Pseudophryne lay large eggs that are individually surrounded by tough, transparent capsules. When hydrated the egg capsules swell to a relatively large size. The eggs of Southern Corroboree Frogs are amongst the largest in the genus (Tyler 1989); measuring approximately 3.4 mm in diameter, with a capsule that swells to up to 8.0 mm in diameter when hydrated.
The tadpoles are generally well advanced when they hatch, measuring about 15 mm in total length. Hatching generally occurs at about Gosner (1960) stage 27. Osborne (1991) provides a simple key to identify tadpoles that may occur sympatrically with both Southern and Northern Corroboree Frogs. Considerable skill is required to correctly identify the species at the tadpole stage. It is not possible to distinguish between the tadpoles of the two species of Corroboree Frogs although Osborne et al. (1996) noted that P. pengilleyi had a greater number of blotches on the tail-fin. The mouth parts of the two species are identical.
The Southern Corroboree Frog has a limited geographic distribution (Figure 1). All known historic sites of occurrence are shown in Figure 2. The species occurs only in Kosciuszko National Park from Smiggin Holes in the south, and northwards to the Maragle Range, about 5km west of Cabramurra. This constitutes a linear range of 51km. The broadest part of the range (24 km) occurs near Mount Jagungal. The species occupies a relatively narrow altitudinal strip between about 1300 and 1760 m, occupying an area of about 400 km2 (Osborne 1989). The Southern Corroboree Frog is separated from populations of the Northern Corroboree Frog (Figure 1) by the comparatively dry and wind-swept Kiandra and Coolamine Plains. These areas consist of broad stretches of cold, treeless plain that are devoid of sheltering tall heath and woodland. The plains have slightly lower annual precipitation than the nearby mountain ranges (Adomeit et al. 1987). This may act as a barrier to the dispersal of the frogs, effectively isolating the two species.
Extensive surveys were undertaken across three summers (1996-1998) to determine the current extent of the Southern Corroboree Frog (Hunter et al. 1997 and in prep.). Of the approximately 160 potential breeding sites surveyed during this period (which included 60 sites where the species was recorded to be formerly present by Osborne 1989), only 63 sites were found to still support the frogs (Figure 3). Of the 60 former sites surveyed, only eight were found to still have frogs (Figure 3). At sites where frogs still occur, the numbers of adult males remaining was very low. Fifty of the 63 sites had fewer than six adult males present. Only one large population remains (95 adult males in 1998). The overall geographic range of the species has now contracted (Figure 3), and includes extensive areas where the frogs are now either extinct, or in very reduced numbers (particularly Pretty Plain, Happy Jacks Plain, Finns Swamp, Whites River and the Smiggin Holes and the Guthega area south of the Snowy River).
 |
| Figure 1. The distribution of P. corroboree and P. pengilleyi in relation to existing nature conservation reserves. Darker stippling indicates the range of each species. |
All known and historical populations of the Southern Corroboree Frog occur within Kosciuszko National Park, an area managed by the NPWS. The security of this National Park tenure is governed by the provisions of the National Parks and Wildlife Act 1974 (NPW Act). The land is zoned 8a - National Park. Existing populations are managed within the provisions of the Kosciuszko National Park Plan of Management. Small populations of the Southern Corroboree Frog previously occurred in the Perisher-Smiggins and the Guthega resort areas within the park (Osborne 1988).
 |
| Figure 2. Map of part of Kosciuszko National Park showing sites where P. corroboree was recorded in 1986/87 by Osborne (1989) (all sites shown on the map) and the results obtained from resurveying the same sites during 1996 to 1998 (Closed circles, sites with P. corroboree in 1986/87 which still supported the frogs in 1996-1998). |
 |
| Figure 3. Distribution of all remaining sites where P. corroboree was recorded at least once during extensive surveys conducted during 1996 to 1998. Triangles represent sites where P. corroboree were found during extensive surveys during 1996-1998; open circles, sites where P. corroboree previously occurred and are no longer considered to be present. The species has disappeared from parts of its former range near Smiggin Holes, Happy Jacks Plain, Pretty Plain and Tooma Dam. Numbers of adult males at most sites shown were extremely small. |
Southern Corroboree Frogs utilise two distinct habitat types: a breeding season habitat associated with pools and seepages in sphagnum bogs, wet tussock grasslands and wet heath; and a terrestrial non-breeding habitat in forest, sub-alpine woodland and tall heath adjacent to the breeding area (Plate 1). During the summer, the adult frogs breed in shallow pools and seepages within the breeding area, before returning to the adjacent woodland and tall moist heath at the end of the breeding season (Pengilley 1966). Osborne (1988 and unpublished) found that, following breeding, adults are capable of dispersing over 300 metres into the surrounding woodland.
There have been no published studies of the breeding habitat of Corroboree Frogs, but detailed quantitative information is provided by Osborne (1990). Breeding sites usually consist of temporary pools and seepages, in areas with level, to gently sloping topography. Sites used by the frogs occur on granitic and volcanic substrates, but porous rock types, including shale and limestone, are generally avoided. The vegetation present at breeding sites varies considerably, ranging from bog and wet-heath at higher altitudes to wet sod-tussock grassland in some low-lying valleys.
Breeding sites are characterised by the presence of spreading rope-rush Empodisma minus, peat moss Sphagnum cristatum, and the shrubs Baeckea gunniana and Epacris paludosa. Several other plant species, including sod-tussock grasses Poa spp., candle heath Richea continentis, the rush Restio australis and the sedge Carex gaudichaudiana commonly occur at breeding sites. The breeding pools are characteristically shallow, and have relatively large surface areas, low water flow rates, and have a long duration (Osborne 1990). This allows the water in the preferred pools to become warmer during the day, possibly enhancing tadpole development. The breeding pools are the most sensitive feature of the breeding habitat of the Southern Corroboree Frog and protection of the catchment area of the pools and seepages is essential for their long-term persistence as breeding sites.
Litter, logs and dense ground cover in the understorey of snow gum woodland provides over-wintering habitat (Pengilley 1966). Plant species typical of this habitat include Bossiaea foliosa, Prostanthera cuneata, Oxylobium spp. and Phebalium spp.
 |
| Plate 1. Typical breeding habitat used by Southern Corroboree Frogs. Woodland and heath in the distance provides non-breeding season habitat for adults and juveniles. Photograph: Pavel German |
Frogs typically undertake annual movements from non-breeding habitats to ponds and other breeding sites (Duellman and Trueb 1986). Osborne (1988) conducted a two-year program of pitfall trapping Northern Corroboree Frogs at Ginini Flats in the ACT.
Ginini Flats is a large Sphagnum bog surrounded by dry Bossiaea foliosa heath and subalpine woodland. The results of this study indicated that at the end of the breeding season adult males and females undertake a seasonal movement away from the bog, upslope into surrounding heath and woodland. The dispersal distances measured at this time were up to 300 m. Pengilley (1992) also found that Northern Corroboree Frogs moved between woodland and low-lying heath and grassland breeding sites at Coree Flats in the Brindabella Range. It is likely that the Southern Corroboree Frog undertakes similar movements and the few observations available support this view.
The over-wintering sites of Southern Corroboree Frogs are not known. Pengilley (1966) found large numbers of adult and sub-adult Northern Corroboree Frogs near Coree Flats in the northern Brindabella range "under logs, leaf litter and vegetation in the forest". Osborne (1988) also searched for over-wintering frogs at Coree Flats and Ginini Flats, but found only 15 individuals, all sheltering beneath logs. The frogs were found at distances ranging from 15 to 155 m from the nearest breeding pools. It is likely that any ground-layer litter and debris, including logs, rocks, fallen bark and grass thatch, will provide over-wintering sites for the frogs. Because of the importance of suitable over-wintering habitat to the frogs, protection of habitat should include those areas adjacent to the breeding sites, which may provide shelter and food for sub-adults and adults throughout the year.
Like most frogs, Southern Corroboree Frogs have two stages in the life-cycle; an aquatic tadpole stage and terrestrial post-metamorphic juvenile and adult stage. However, they differ from other frogs in that their eggs are laid out of the water in a terrestrial nest, and the embryos develop to an advanced stage within the egg capsule before hatching.
Adult males move into the breeding areas in early summer, and call during January and February. The males call from small chambers in moss or other soft vegetation at the edges of the breeding pools. The frogs have three call types; an advertisement call, threat call, and courtship call (Pengilley 1971b). The advertisement call and courtship call are used to attract females to the males calling site, whereas the threat call serves as a warning to other males (Pengilley 1971b). Stimulating the males to give threat calls has been recommended as a reliable procedure for monitoring the number of calling males (Osborne 1991). Advertisement call intensity varies depending on the weather, with more calling occurring during warmer overcast conditions, and during late afternoon (Pengilley 1971b).
Females only enter the bogs briefly to lay their eggs in the terrestrial oviposition site, and then leave the breeding site. The males continue calling for a number of weeks, presumably to continue mating. They then leave the bogs during late February to return to the over-wintering habitat (Pengilley 1966; Osborne 1988). The eggs are laid in small clutches of 16 to 38 eggs from January to mid February (Pengilley 1973). The eggs are amongst the largest in the genus (Tyler 1989), measuring about 3.5 mm in diameter; the transparent capsules swelling to about 8 mm in diameter when hydrated (Osborne 1991). Tadpole development initially occurs within the egg and the relatively advanced tadpole emerges from the eggs at Gosner (1960) stage 27, measuring about 15 mm. At this stage the tadpoles are well-developed with the first sign of the hind limb showing as a small bump on the body near the base of the tail.
Hatching is thought to occur during autumn and winter during periods of high rainfall or snow melt. The tadpoles wriggle to the nearby pools where they live for the remainder of the larval period, feeding on algae and other organic detritus in the water. The tadpoles show little growth during winter, when temperatures at the breeding sites are very low and snow often covers the ground. The water in the pools at this time may be below 1°C and the surfaces of exposed pools regularly freeze if exposed to the air on cold nights (Jacobson 1963). At the end of winter, when snow has melted from the breeding sites, the tadpoles continue growing slowly until metamorphosis in early summer.
Reliable autumn precipitation is critical to the over-winter survival of the tadpoles. It appears that if the nests are not submerged during this period, then the eggs either do not hatch, or hatch prematurely in the nest site, with the tadpoles becoming stranded in the dry vegetation surrounding the nest. For example, in the autumn and winter of 1997 precipitation was not high enough to cause the flooding of nests at many sites and there was extensive mortality of tadpoles of both species of Corroboree Frogs (D. Hunter, University of Canberra, unpublished observations). The pre-metamorphic period is also critical even if the tadpoles manage to move to the pools because they are vulnerable to pool-drying at this time.
Very little is known about the life history of the frogs after they leave the pools as juveniles. Pengilley (1966, 1971a, and pers. comm.) suggested that they remain in moist vegetation near the breeding pools for several months, where they feed on a wide variety of small invertebrates. As they grow larger, the juveniles leave the breeding area and move into the adjacent non-breeding habitat where it is thought they remain until they are adults. The diet of sub-adults and adults consists mainly of small ants and, to a lesser extent, other invertebrates (Pengilley 1971a). Food intake is greatly reduced during winter, with many individuals apparently not feeding. Frogs that have a predominance of ants in their diet are considered to be specialised (Duellman and Trueb 1986).
The particular reproductive strategy of the Southern Corroboree Frog — low clutch size and prolonged period of larval development — is suggestive of an evolutionary adaptation to an environment that is predictable, with little long term variation. The breeding pools must have water in them from the time when the tadpoles hatch and enter the pool (thought to be late autumn and winter), to mid-summer (December and January) when the newly emerged froglets leave the pool (Pengilley 1966; Osborne 1991).
Pengilley (1966) suggested that individuals reached sexual maturity in about three years (ie. one year as an embryo and tadpole, and two years as a juvenile). In addition, based on the results of a mark-recapture study of Northern Corroboree Frogs, Pengilley (1966, 1992) suggested that there was a high mortality of adults between years, with few adults that had already mated being recaptured in successive years. However, there was a severe drought during the middle of his study which may have increased mortality. Recent research undertaken by David Hunter indicates that the survival rate of Northern Corroboree Frogs is higher than found by Pengilley and that adults may live to eight years. Now that the techniques for marking and ageing frogs have been validated (D. Hunter unpublished data) this research will be extended to include the Southern Corroboree Frog.
Amphibians typically experience high levels of mortality during their pre-metamorphic life-history stages (Wassersug 1975). However, the level of pre-metamorphic mortality may be highly variable between species and appears to be related to the life-history characteristics of individual species (Duellman & Trueb 1986). Whether mortality is greatest during embryonic or larval development also appears to be related to the life history characteristics of a species (Duellman and Trueb 1986). Species that lay their eggs terrestrially, such as those within the genus Pseudophryne, often have greater embryonic survivorship than those species that lay their eggs in an aquatic environment (Duellman & Trueb 1986).
Pre-metamorphic mortality in Pseudophryne has been investigated previously by Woodruff (1975) and Pengilley (1966, 1992) who focused primarily on embryonic mortality. Woodruff (1975) found embryonic mortality in the laboratory to be less than 5% for Pseudophryne bibroni, P. dendyi and P. semimarmorata, although he did observe much higher mortality in the field as a result of desiccation when a breeding season was followed by an abnormally dry period. Pengilley (1992) studied field embryonic mortality in P. corroboree and P. pengilleyi (then recognised as P. corroboree). He also found relatively low embryonic mortality, except during a drought year when the majority of egg clutches exhibited 100% mortality. These results support the generalisation that species which lay their eggs terrestrially exhibit low embryo mortality except when faced with catastrophic events.
Very little research has been conducted on tadpole mortality in Pseudophryne. This is mainly due to the logistical problems associated with accurately sampling the aquatic life-history stage. The tadpoles frequently hide in thick pond vegetation or bury themselves in silt at the bottom of the pools (D. Hunter and W. Osborne, pers. obs.). Pengilley (1966) determined the mortality rate of the Southern Corroboree Frog tadpoles in three pools based on the number of eggs originally present. The results varied with mortality rates of 50%, 86.5% and 93.6% being calculated for the different pools. Pengilley (1966) suspected that the mortality he detected was predominantly due to predation by dragonfly larvae (although this statement was not qualified). The highly ephemeral nature of the breeding sites for both species of Corroboree Frogs also renders them susceptible to pool drying (Osborne 1990). These observations suggest that there is the potential for high levels of mortality at the tadpole stage, which is consistent with the mortality levels observed in most amphibians which possess an aquatic larval stage (Duellman & Trueb 1986).
During three years of field-based research, Osborne (1990) saw no evidence of predation on Corroboree Frogs. Although tadpoles of many frog species are subject to very high rates of predation from fish and invertebrates (Heyer et al. 1975; Caldwell et al. 1983; Duellman and Trueb 1986), these organisms are scarce in pools used by Corroboree Frogs. Vertebrates, such as herons and snakes, which may prey on juvenile and adult frogs, are also very uncommon at high montane and subalpine altitudes (Green and Osborne 1994) and are unlikely to influence populations in this area. Other potential predators of adult and sub-adult frogs include predatory beetles and reptiles which are common in subalpine woodland, however it is likely that the frogs are unpalatable to these smaller predators (W. Osborne and D. Hunter, unpublished observations).
Parasites of frogs include protozoans, helminths, leeches, mites and the larvae of flies (Tyler 1976; Duellman and Trueb 1986). Whilst many parasites are not lethal to their hosts, massive infestations by sporozoans have been observed to cause high mortality in adults and tadpoles in other countries (Duellman and Trueb 1986). In Australia, Pseudophryne spp. are parasitised by the dipteran genus Batrachomyia or frog flies, protocephalid cestodes and oxyuroid nematodes (Pengilley 1966, Tyler 1976). Pengilley (1966) thought that heavy infestations of Batrachomyia spp. may have caused the death of juvenile and sub-adult P. pengilleyi, however this was not observed directly. Pengilley noted that the low percentage of adults infected with small larvae may have indicated that adults are less susceptible to infection. Batrachomyia sp. was found to parasitise both P. pengilleyi and P. corroboree but was not observed in adults from the subalpine zone (Pengilley 1966). Since the decline in the size of the frog population, the number of parasitised frogs has apparently been reduced markedly. Of 226 frogs examined in 1986/1987 by W. Osborne and M.Lau, only one individual (a P. pengilleyi) was found to be parasitised (W. Osborne, unpublished data). Although there has been little direct research on the topic, it is unlikely that parasites have had a major role in the declines of these populations.
The onset of rapid declines in undisturbed populations of frogs in high altitude areas has given grounds for concern that the declines may be attributable to a particularly virulent pathogen. Initially it was suggested that the cause of the population crashes was most likely to be a virus (eg. Laurance et al. 1996), however recent evidence from dying frogs collected in pristine areas strongly implicates a skin fungal disease (Chytridiomycosis) in the deaths of frogs in both Australia and South America (Berger and Speare 1998).
Further discussion on this issue is provided in Section 4.2.3
If the continuing decline of populations can be prevented, it is highly likely that, with correct management of habitat, local populations will recover. However, it seems unlikely at present that the species will recover throughout its geographic range.