Revised Recovery plan for the Carpentarian Rock-rat Zyzomys Palatalis
Helen Puckey, John Woinarski and Colin Trainor
Parks and Wildlife Commission of the Northern Territory, January 2003
- 1.1 Description of the species
- 1.2 Distribution
- 1.3 Population Size and Structure
- 1.4 Habitat critical to survival
- 1.5 Life history and ecology
- 1.6 Reasons for listing
- 1.7 Current threats
- 1.8 Existing conservation measures
The Carpentarian Rock-rat is one of five species of rock-rat in the Australian endemic genus Zyzomys. All species are similar in gross morphology and diet, though the Common Rock-rat Z.argurus is substantially smaller than the other four species. Four species are restricted to northern Australia (three of which have very limited ranges): the fifth species, Central Rock-rat Z.pedunculatus, occurs only in central Australia.
The rock-rats are characterised by "compact, harsh-furred bodies, pronounced roman-nosed appearance, rounded ears, and protruding eyes" (Watts and Aslin 1981), and especially by the fragile carrot-shaped tail, in which fat is deposited. They also have notably long whiskers. The Carpentarian Rock-rat is very similar in appearance to the Arnhem Land Rock-rat Z.woodwardi and the Kimberley Rock-rat Z.maini. It is grey-brown above and white below. The tail is relatively hairy, with especially long hairs towards the tip. Body weight averages about 120g. As with other Rock-rats, Z.palatalis is known to lose their tails, fur and skin very easily.
The three large rock-rats in northern Australia occupy geographically disjunct distributions in rugged sandstone ranges: the Kimberley Rock-rat occurs only in the north Kimberley, the Arnhem Land Rock-rat is restricted to the escarpment and plateau of western Arnhem Land, and the Carpentarian Rock-rat occurs only in a very small area of the Gulf of Carpentaria hinterland. Morphological similarity between the three species suggests that the group may have had a more continuous distribution in the past, presumably when their major food-bearing plants were more widespread.
Currently, the Carpentarian Rock-rat is known from just four locations, all within a radius of 35km. All sites are within Wollogorang pastoral station, in the Northern Territory (adjacent to the Queensland border). Further occurrences beyond the known sites are possible, as the Gulf area has not been subject to comprehensive fauna survey however, specific searches for Carpentarian Rock-rats targeting apparently suitable habitat have been largely unsuccessful (Churchill 1996, Puckey 1999 unpublished data).
There are no fossil records. No known populations have become extinct, possibly because the species was only discovered as recently as 1986 (Kitchener 1989).
The actual population size of Carpentarian Rock-rats is largely unknown. However, extrapolation of population size from density estimates indicate that the population was 696 at Moonlight Gorge and 450 at Banyon Gorge in 1996 (Trainor 1996). No such estimates are available at the other two sites. Carpentarian Rock-rats generally have a 1:1 sex ratio in the population and females give birth to 1-3 young per litter, those females living into their second year producing up to four litters (Trainor 1996).
The Carpentarian Rock-rat is strongly associated with monsoon thickets (dry rainforest) occurring on rocky slopes within large gorge systems. Plant species characteristic of this habitat include Pouteria sericea, Terminalia subacroptera, T. volucris, Celtis phillipensis, Ficus leucotricha, F. virens, Cissus reniformis, Caesalpinia bonduc and Gyrocarpus americanus. As the entire distribution of Carpentarian Rock-rats is restricted to only four of these rainforest patches they clearly represent habitats critical to the survival of the species as defined under section 201A of the Environment Protection and Biodiversity Act 1999. However, it is possible that other rainforest patches may also be critical to the survival of the species.
The floristics, distribution and conservation of monsoon rainforests in the Northern Territory have been considered in detail. Russell-Smith (1991) mapped and classified rainforest patches, recording seven floristic groups (from 16 defined for the Northern Territory) from the Gulf area. Several of these groups are not associated with rocks, and are unlikely to support Carpentarian Rock-rats. However, three groups are rock-associated and comprise many plant species whose fruits are known to be eaten by rock-rats. The habitat provided by these groups may therefore be necessary to ensure the long-term future of the species, for example through re-introduction or re-colonisation. Further studies are necessary to identify any additional areas of habitat critical to the survival of the species and to define the boundaries of these areas.
Liddle et al. (1994) mapped the distribution of individual monsoon rainforest plant species. These maps show that many rare monsoon rainforest plants occur as relict populations in gorges of the Gulf area. Russell-Smith and Bowman (1992) described the conservation status of monsoon rainforests in the Top End of the Northern Territory, and noted that monsoon rainforest patches in general, and groups occurring in the Gulf region in particular, had a high frequency of disturbance from fire, feral animals and weeds.
The Carpentarian Rock-rat is a nocturnal rodent, sheltering during the day in cracks between boulders, caves and crevices. Much of its diet comprises fruits and seeds of fleshy-fruited plants, including Terminalia subacroptera, T. carpentariae, Ficus spp and Pandanus aquaticus. Its large incisors allow it to chew through the woody nuts of many of these species to access the kernel, and distinctively chewed nuts are a characteristic sign of its presence. This diet is common to all the rock-rats, and information from the more intensively-studied Arnhem Land Rock-rat (Begg and Dunlop 1980, 1985) is probably generally applicable to the Carpentarian Rock-rat.
As with all rock-rats, reproductive output is characteristically low, with females bearing only four nipples. However, growth of young is unusually rapid (Watts and Aslin 1981), perhaps allowing more frequent breeding events. Reproduction in the Arnhem Land Rock-rat has been said to be seasonal (mostly from March to May: Begg 1981) or a seasonal (Calaby and Taylor 1983). Breeding in the Carpentarian Rock-rat has been reported in most months.
There are only very limited data on longevity, sex ratios, territoriality and age structure of Carpentarian Rock-rats. There appears to be a substantial seasonal fluctuation in the abundance of Carpentarian Rock-rats, with relatively low numbers reported from January to June. It is unclear whether this is due to movements away from the habitats sampled in studies based on permanent grids, seasonal changes in trapability or real changes in population numbers.
The Carpentarian Rock-rat is listed as endangered due to its extremely limited total range; small number of fragmented populations within that range; presumed small total population size; specific habitat (and dietary) requirements; and demonstrated deterioration in the condition of that habitat generally (Russell-Smith and Bowman 1992) and specifically for at least one of its four known sites (Banyan Gorge).
Current putative threats common to all known sites are frequent and hot fires and possibly grazing by cattle, both of which can damage the integrity of remnant rainforest patches. Studies demonstrated that the Arnhem Land Rock-rat Zyzomys woodwardi declined substantially following wildfire (Begg et al. 1981). Feral predators (cats) are known to inhabit the area and could have a negative effect on the small and isolated populations of Carpentarian Rock-rats. The actual effects of these putative threatening processes remain a significant gap in our knowledge of Carpentarian Rock-rats.
There are currently 17 Carpentarian Rock-rats being maintained in a captive breeding colony at the Territory Wildlife Park. PWCNT is currently in the process of establishing a fire management program for each of the known sites under a cooperative management agreement with the land holder. All areas surrounding the four known habitats are currently de-stocked and intended to remain that way for the immediate future.