National Multi-species Recovery Plan for the Carpentarian Antechinus Pseudantechinus mimulus, Butlers Dunnart Sminthopsis butleri, and Northern Hopping-mouse Notomys aquilo 2004-2008
Australian Government Department of the Environment and Heritage, 2004
Part C: Known and potential threats
- Biology and ecology relevant to threatening processes
- Identification of threats
- Populations and Areas under threat
Limited knowledge of the ecology and general biology of all three species is a major constraint on conservation planning, identification of threats and management of those threats.
Based mainly on detailed studies of its close relatives (e.g. Watts and Aslin 1981), the northern hopping-mouse is probably mostly granivorous, taking seeds of grasses, herbs and shrubs. Many granivorous birds are known to have declined extensively across northern Australia (Franklin 1999), largely because this resource has been affected by vegetation change caused by altered fire regimes, grazing by livestock and incursions by exotic weed species (particularly pasture grasses). Many granivorous rodents have probably suffered similar patterns of decline for similar reasons, although the information available is generally more sparse (Woinarski 2000).
For granivorous animals, the most critical resource changes appear to be related to spatial heterogeneity, with seed resources most likely to be available year-round if the consumer can access a broader range of variability (such as that created or enhanced by fine-scale burning) (Fraser et al. 2003; Woinarski et al. in press, a). Seed availability is also affected by changes in phenology, vegetation structure and floristic composition, with each of these capable of being affected by imposed fire regimes, grazing and occurrence of exotic plants (Crowley and Garnett 1999, 2001; Woinarski et al. in press, a).
The northern hopping-mouse is also highly communal, with many individuals occupying a single burrow system. This characteristic may render the species particularly susceptible to predation by feral cats, which may wait at burrow entrances and over several nights consume entire colonies. To some extent, this threat may be partly limited by the explosive exit of hopping-mice from their burrows and some degree of concealment of burrow entrances (Johnson 1964; Dixon and Huxley 1985).
The two other species considered here are more solitary, and probably consume a broad variety of invertebrate and small vertebrate food, and hence may be somewhat less affected by habitat change and predation. The very limited information available suggests that Butlers dunnart shelters, at least at times, under fallen woody debris. This may render the species susceptible to direct mortality in hot fires, and to a reduction in suitable shelter sites following extensive hot fires.
Given the lack of specific information, some assessment of threats likely to be affecting these three species may be interpolated from evidence of a more general decline in native mammal assemblages across much of northern Australia (Woinarski et al. 2001; Watson and Woinarski 2003; Pardon et al. 2003; McKenzie and Burbidge 2002). Such declines appear to be affecting species with very diverse ecologies. Given the extensiveness of the patterns of decline and the broad range of mammal species involved, the causes of decline must also be extensive and non-specific. There are three processes that may fit this bill: predation by feral cats; altered vegetation patterning (of floristic composition, structure and patchiness, and due to either changed fire regimes or the impacts of feral or managed livestock); or disease. There is no evidence that points conclusively to the primacy of any of these factors.
Of these three putative threatening processes, there is by far the most information available about the impacts of fire. Recent studies have shown that mammal communities in northern Australia can be readily moulded by variation in fire regimes (e.g. Woinarski et al. in press, b), with many species showing very marked responses to fire frequency, fire intensity and fire scale, mediated through direct impacts upon vegetation patterning. There is now much evidence showing that current fire regimes are markedly dissimilar to those that were probably most pervasive under Aboriginal land management (Bowman 1998; Vigilante 2001; Yibarbuk et al. 2001; Preece 2002), and hence that there has been broad-scale change in habitat suitability for many mammal species across northern Australia.
However, there is little or no specific information about the response of the three mammal species considered here to individual fires or fire regimes, because these species have been subject to so little research. A limited correlative study of the northern hopping-mouse suggested that it preferred spatially heterogeneous heathlands, with such patchiness probably promoted largely by fine-scale burning (Woinarski et al. 1999). In other heathlands in northern Australia, frequent fire (typically with recurrence of less than 4-5 year intervals) has been shown to reduce floristic diversity, and particularly so for the shrubby native pea species (Russell-Smith et al. 1998, 2002), with which the northern hopping-mouse appears to be associated. Hence, a preferred fire regime for this species may be fine-scale burning with return intervals of at least 5 years, and the extent of any departure from this regime will be related to the extent of detrimental impact.
However, this conclusion is tentative and needs to be examined through more rigorous study. There are no data on the responses of carpentarian antechinus and butlers dunnart to fire. Other ground-dwelling dasyurids in northern Australia have been shown to be disadvantaged by frequent extensive hot wildfires (e.g. Oakwood 2000). On such slight evidence, a conservative fire management (infrequent fires and/or fine-scale “cool” fires) may be preferred by butlers dunnart and carpentarian antechinus, but such a presumption should be subject to more rigorous study.
There is no information on the impacts of predation by cats upon these three species. Such impacts may be expected to be most severe on the colonial burrow-nesting northern hopping-mouse rather than the solitary Butlers dunnart and carpentarian antechinus, with the latter probably given some additional protection from predation because of its preferred rugged rocky habitat.
There is no information on any impacts of disease upon any of the three species considered here.
Grazing by livestock and/or feral stock may have some impact upon these three species, but there are no data to evaluate such a claim. Neither the northern hopping-mouse nor butlers dunnart is known from any lands used for pastoralism, but feral stock (cattle, buffalo and pigs) occur, generally at low densities, across much of their known range. The apparent stronghold of carpentarian antechinus (the Pellew islands) is largely free of stock, except for Vanderlin Island (where cattle occur at generally low densities). The few recent records of carpentarian antechinus in Queensland occur largely on pastoral lands, but pastoralism is unlikely to affect this species, even on such properties, given that it is largely restricted to rugged rocky areas.
In contrast to the pervasive threats of altered vegetation patterning, predation by feral cats and disease, more acute factors may (also) threaten the northern hopping-mouse and carpentarian dunnart. The largest known populations of northern hopping-mouse occur on Groote Eylandt and the mainland of north-eastern Arnhem Land. Broad-scale strip-mining occurs in both areas, for manganese and bauxite respectively, and this mining probably affects some populations of northern hopping-mice. However, in both regions, the mining activity is concentrated mainly away from the sandy substrates preferred by northern hopping-mouse; and on the mainland of north-eastern Arnhem Land, a population of northern hopping-mouse is protected from mining within the Nanydjaka IPA. All recent records (i.e. since 1966) of Butlers dunnart have been from the Tiwi Islands. Over the last decade, an extensive forestry operation has been developed on the larger of these islands, Melville, with approval for clearing 25,000 ha of native forest for plantation of exotic timber species. This constitutes about 3% of these islands, and may be expected to have a similarly proportioned impact on the population of Butlers dunnart.
The carpentarian antechinus was first recorded from Queensland during a study of the impacts upon biota of sulphur dioxide emissions from the Mt Isa copper smelter (Griffiths 1998). In that study, capture data were too few to derive any relationship with emission levels, but captures of carpentarian antechinus occurred only in the two sites with lowest levels of emissions (>20 km from emission source), and not in similar environments at four sites with higher emission levels.
All populations of each of the three species are subject to at least some of the pervasive threatening processes of vegetation change through altered fire regimes (and/or impacts of livestock) and/or predation by feral cats.
Populations of northern hopping-mice are subject to some impacts from strip-mining on north-eastern Arnhem Land and Groote Eylandt. Populations of Butlers dunnart are subject to some impacts of vegetation clearance for forest plantation on Melville Island.