Fauna Spotters logo

Julia Creek Dunnart (Sminthopsis douglasi)

The Julia Creek dunnart is a small, nocturnal carnivorous marsupial restricted to cracking clay grasslands in Queensland. It is listed as Vulnerable under the EPBC Act and Queensland legislation, and is difficult to detect because it occurs at low densities, has fluctuating populations and shelters below ground during the day. Clearing can remove both the tussock cover and the cracking clay refuges used for shelter and nesting.

Conservation status

Commonwealth, EPBC Act 1999: Vulnerable.

Queensland, Nature Conservation Act 1992: Vulnerable.

Breeding season

The timing below describes a typical seasonal pattern for central and north-west Queensland. Rainfall can shift breeding, recruitment, movement and detection, and the species may be captured outside these periods.

For survey and clearing planning, April to October is generally preferred over the wettest months. Fine weather improves access and reduces the chance of separating lactating females from young, but survey success remains rainfall dependent.

  • Breeding: Breeding generally begins in September and extends to mid-February. Females may raise two litters in a good season.
  • Young: Juvenile recruitment begins from December. Lactating females without attached young have been captured from November to April, and young remain dependent for about 10 weeks.
  • Adult movement and mate searching: By July and August, males from that season are actively looking for mates. The species otherwise moves within home ranges and may change daytime resting sites frequently.
  • Wet season shelter and flood exposure: Rain can close soil cracks and force animals to use vegetation. Heavy flooding may harm females carrying pouch young and less mobile juveniles.
  • Highest clearing risk: Clearing and ground disturbance are most likely to harm the species from September to April because breeding, pouch young, dependent juveniles and wet-season shelter use overlap. Risk is highest from December to February when young are recruited and cracks may be closed by rain.

Identification

The Julia Creek dunnart is the largest species in the genus Sminthopsis. It is cryptic, mostly nocturnal and may be encountered only briefly at night or in live traps.

Adults have brown fur peppered with grey above and buff white fur below. Rufous fur occurs on the cheeks and at the base of the ears. The face has a conspicuous dark stripe extending from between the ears towards the pointed snout, with dark hairs forming rings around the eyes and occurring on the outer edge of the ears.

The thick, tapered tail is about as long as, or slightly shorter than, the head and body. It is often dark haired towards the tip and fattened at the base. The feet are buff yellow to pinkish white.

Head and body: 110 to 135 mm.

Weight: Usually more than 26 g in adults; up to 70 g.

Hind foot: Usually more than 20 mm in adults.

Identification: Distinguishing similar dunnarts

The species occurs with the stripe-faced dunnart, Sminthopsis macroura, and the fat-tailed dunnart, Sminthopsis crassicaudata. An adult Julia Creek dunnart is generally larger, has a hind foot longer than 20 mm, dark hairs around the eyes, dark hairs on the ear tips and often a dark tail tip.

Juvenile Julia Creek dunnarts can overlap in weight and hind foot length with adult stripe-faced dunnarts. Individuals weighing 26 g or less should be aged from the premolar teeth where identification is uncertain. A fully descended third premolar that is larger than the other premolars indicates an adult, while a third premolar equal to or smaller than the second indicates a juvenile.

  • Sminthopsis macroura is generally 15 to 26 g as an adult, usually has a hind foot shorter than 20 mm, and usually lacks dark hairs at the tail tip.
  • Sminthopsis crassicaudata is generally 10 to 20 g, has no obvious facial stripe, and has a hind foot shorter than 20 mm.
  • Planigale species are much smaller, have a notably flatter head and lack the curled skin fold above the tragus characteristic of Sminthopsis.

Identification: Field signs and survey encounters

During the day, look for small soil cracks and holes in cracking clay plains that may be used as temporary resting sites. Radio-tracked animals were located in cracks or holes 10 to 50 cm below ground, but the animals did not reuse the same resting site during the short tracking period.

There is no documented call or distinctive track description in the supplied sources that should be relied on for identification. Evidence from predator feeding remains can be useful, including jaw bones in eastern barn owl pellets and remains from feral cat or fox stomach contents. Adult Julia Creek dunnart mandibles are generally at least 20 mm long.

For live detection, use targeted Elliott trapping, camera trapping, thermal imaging or corroborated searches of predator remains. Low capture rates do not demonstrate absence because trapping success varies from zero to 6.75 percent at a known site and may be below 1 percent in areas where the species is present.

  • Search cracking clay refuges, particularly where Mitchell grass or other tussock cover is dense.
  • Treat a suspected dunnart record from a camera image, hair sample or predator remain as provisional unless the method can separate it from sympatric dunnarts.
  • Do not disturb or collapse soil cracks during searches, and check live traps early before temperatures rise.

Distribution: Range and bioregions

The Julia Creek dunnart is endemic to Queensland. Its known distribution is centred on the Downs Country of central and north-west Queensland, with records extending into the Gulf Plains and Desert Uplands bioregions.

Most records are from the Julia Creek, Richmond and Winton areas. The species has been recorded across Mitchell Grass Downs grasslands and adjacent parts of the Gulf Plains and Desert Uplands bioregions.

The formerly known range expanded substantially after surveys in the 1990s and early 2000s. Records now extend more than 200 km north to south from earlier known localities, with a possible additional record near Barcaldine and a later locality near Mount Margaret, west of Julia Creek.

  • State: Queensland only.
  • Core geographic area: approximately 8,000 km2 between Julia Creek and Richmond, with the broader known range extending towards Winton, Hughenden, Cloncurry, Barcaldine and adjacent Desert Uplands areas.
  • The species has not been confirmed from apparently suitable areas on the Barkly Tableland in the Northern Territory.

Distribution: Strongholds and population pattern

Important known populations occur in Bladensburg National Park, Moorrinya National Park, Toorak Research Station and the Proa or Yorkshire Downs area. The Julia Creek aerodrome also contains a relatively undisturbed Mitchell grass remnant that has supported a live record.

Occurrences are patchy and abundances are low. Population numbers can change markedly with rainfall, ground cover and recruitment, producing boom and bust patterns. All known populations are considered important because the species has a restricted and fragmented distribution.

Known locations in surveys: 25 locations across the Mitchell Grass Downs and adjacent Desert Uplands were reported in the recovery plan.

Recorded density: Approximately 0.16 to 0.38 individuals per hectare in a recent density study.

Known long-term trend: No reliable whole-range population estimate or trend is available; abundance fluctuates with climate and recruitment.

  • Bladensburg National Park has supported relatively high numbers compared with other known sites.
  • Moorrinya National Park supports a smaller but significant population.
  • Toorak Research Station, private grazing properties, stock reserves and road reserves may support habitat or populations outside protected areas.

Habitat: Vegetation and soils

The species is associated with treeless or lightly timbered tussock grasslands on heavy cracking clay plains. Habitat suitability depends on both soil structure and the amount and condition of ground cover.

Preferred habitat is Mitchell grass tussock grassland on ashy or stony cracking clay soils. Associated grasslands may include Mitchell grass, Astrebla species, Flinders grass, Iseilema species, wire grass, Aristida species, and blue grass, Dichanthium species.

The species can occur where ground cover is sparse to dense, but habitat assessments indicate a preference for ground cover about 40 to 50 cm high with 40 to 65 percent cover. Dense Flinders grass cover has also been identified as a favoured habitat feature.

  • Prioritise treeless or lightly timbered rolling plains with deep cracking clay soils.
  • Retain areas with a high density of soil cracks and holes, especially where grazing pressure and prickly acacia density are low.
  • Assess both ashy and stony clay soils because the species has been recorded on both types.

Habitat: Climate and site condition

The species occurs in a region with strongly seasonal summer rainfall. Rain stimulates grass and forb growth, increases ground cover and can increase prey availability. As clay dries, the soil cracks and provides below-ground refuge.

Heavy rain can close cracks, cause flooding and restrict access to sites. Drought can reduce ground cover and prey availability. No specific elevation limit is given in the supplied sources.

  • Record recent rainfall, flooding, drought, fire and grazing condition when assessing habitat.
  • Map watering points, stock movement routes and prickly acacia because they can reduce habitat suitability through increased grazing pressure or woody weed invasion.
  • Treat intact grassland remnants, stock routes and wide road reserves as potential habitat rather than assuming that only national parks are occupied.

Foraging habitat: Diet and feeding sites

Julia Creek dunnarts forage at night across tussock grasslands and other suitable grass-covered clay plains. They are opportunistic carnivores and may move considerably within a stable home range.

The diet includes crickets, cockroaches, silverfish, slaters, other arthropods and arachnids. Small reptiles and possibly small mammals may also be taken. The supplied sources do not identify a required food plant.

Food availability is linked to rainfall and vegetation growth. Individuals can be in better condition after the wet season when insects and other arthropods are more available. Locust irruptions may provide an additional food pulse.

  • Retain native tussock cover and litter that support arthropod prey.
  • Avoid assuming that bare clay is unsuitable for foraging because animals may forage across sparsely covered areas while using cracks or vegetation for daytime shelter.
  • Record long-haired rat irruptions because high rodent numbers may affect both prey availability and survey capture success.

Foraging habitat: Movement during foraging

The species is highly mobile and may move between temporary daytime refuges and nocturnal feeding areas. Short-term radio-tracking recorded consecutive-day movements of 31.7 to 111.6 m between resting sites.

Longer-term home ranges have been estimated at 0.25 to 7.13 ha, and one study recorded 0.5 to 8.0 ha. These figures support retaining connected habitat around any capture, camera or predator-remain record.

Short-term movement: 31.7 to 111.6 m between consecutive daytime resting sites in a radio-tracking study.

Home range: 0.25 to 7.13 ha in the recovery plan; 0.5 to 8.0 ha in longer-term trapping data.

  • Avoid treating an isolated capture location as the full area of habitat use.
  • Include adjoining tussock grassland, soil cracks and dense ground cover in impact assessments around records.
  • Maintain movement routes between habitat patches where possible, particularly across cleared works areas and infrastructure corridors.

Breeding habitat: Breeding biology

Breeding is seasonal and overlaps the wet season. Females use the same cracking clay grassland habitat as adults, but successful recruitment depends on shelter, prey and ground cover during and after rainfall.

Females have eight teats and can rear up to eight young in a litter. They may raise two litters during the extended breeding season, particularly when rainfall produces suitable food and cover.

Females may return to oestrus as young are weaned. Young remain dependent on the mother for about 10 weeks, and can relinquish the nipples at approximately 35 to 40 days.

Breeding season: September to mid-February in the wild, with some lactating females recorded from November to April.

Litter size: Up to eight young, matching the female's eight teats.

Gestation: Approximately two weeks in captivity.

Young dependence: Approximately 10 weeks, with young able to relinquish the nipples at about 35 to 40 days.

Sexual maturity: Females 17 to 27 weeks; males 28 to 31 weeks in captivity.

Lifespan: Usually 2 to 3 years, with wild males reported up to 12 months and females up to 24 months.

Breeding habitat: Breeding habitat features

Breeding habitat needs nearby daytime shelter, dense ground cover and reliable nocturnal access to arthropod prey. Females carrying pouch young may shelter in cracks, holes or temporary vegetation cover.

Heavy rainfall can close soil cracks and flood low areas. During such conditions, dependent young may be more exposed if females are forced to use vegetation or if young become separated from the mother.

  • Retain dense tussock and forb cover around soil cracks and holes during the September to April breeding and young-rearing period.
  • Avoid clearing or intensive earthworks near occupied habitat during the breeding season where practicable.
  • Inspect proposed work areas for suitable cracks, dense cover and predator remains before disturbance, and use targeted survey methods when habitat is present.

Sheltering habitat: Soil refuges

Daytime shelter is mainly provided by natural cavities in cracking clay soils and, when cracks close after rain, by dense vegetation. Shelter sites can be temporary and may change from night to night.

Dunnarts shelter in cracks and holes in cracking clay plains during dry conditions. Radio-tracked animals were located at depths of 10 to 50 cm. Habitat modelling identifies soil crack and hole density as an important feature because these cavities provide refuge from heat, fire and predators.

The species may not reuse the same resting site. In one short-term tracking study, none of the radio-collared animals reused a daytime crack or hole during the tracking period.

Recorded resting depth: 10 to 50 cm below ground.

Resting-site reuse: No reuse was recorded during one short-term tracking study.

  • Do not fill, compact, collapse or drive over cracking clay refuges without assessing the area for dunnart use.
  • Map soil crack density and the distribution of holes as habitat features, not just as incidental soil observations.
  • Retain unburnt or lightly disturbed refuge areas where they occur within or beside a work footprint.

Sheltering habitat: Vegetation cover during wet conditions

After rain, clay soils swell and cracks may close. The species is then thought to shelter under or within vegetation, including dense grass and forb cover. Individuals have also been observed perching in the tops of forbs during flooding rain.

Vegetation cover helps conceal dunnarts from predators during the wet season. Clearing, heavy grazing and fire can remove this cover even where the soil remains suitable.

  • Retain tussock bases, dense grass, forbs and litter around potential refuge areas.
  • Avoid slashing or grading dense cover during periods when cracks are closed and animals may be sheltering above ground.
  • Where fire is planned, assess the availability of nearby unburnt cover and soil refuges before ignition.

Threats: Clearing, earthworks and fragmentation

The main management threats are habitat removal and degradation, feral predators, prickly acacia and grazing. Rainfall extremes, fire and the species' low and fluctuating abundance can increase the effect of these pressures.

Clearing Mitchell grass grassland can remove tussock cover, prey habitat and the cracking clay cavities used for daytime shelter and nesting. Excavation, stripping, stockpiling, grading and vehicle movement can destroy or compact soil refuges.

The species has a fragmented distribution and may use habitat on private land, stock routes, reserves, road reserves and research stations. A failure to detect animals during a short survey should not be treated as evidence that a suitable site is unoccupied.

Minimum recommended coverage: Assess at least 10 percent of suitable habitat in a project area.

Example effort per 100 ha of suitable habitat: 10 one square metre habitat quadrats, 20 Elliott traps for 7 nights, and 5 cameras for at least 28 nights.

  • Avoid clearing intact cracking clay grassland where practicable, especially areas with 40 to 50 cm ground cover and 40 to 65 percent cover.
  • Survey and map suitable habitat before clearing, including soil crack density, tussock cover, grazing, weeds, fire and flooding history.
  • Stage clearing so that escape habitat remains connected and avoid leaving isolated strips or small remnants surrounded by disturbed ground.
  • Use the Queensland targeted survey guideline as a minimum planning reference: assess at least 10 percent of suitable habitat, with at least 50 percent of survey effort using Elliott-type live trapping.

Threats: Grazing and weeds

Heavy grazing can reduce ground cover, alter soil cracking and increase exposure to predators. Close proximity to stock watering points is associated with increased grazing pressure and lower predicted habitat suitability.

Prickly acacia, Vachellia nilotica, is a major invasive plant threat. Increasing woody weed density can reduce the open grassland structure used by the species and may alter livestock movements and fire behaviour.

  • Record grazing intensity, watering points, livestock tracks and prickly acacia density during habitat assessment.
  • Protect dense grass cover around known and potential populations through grazing management.
  • Prevent the movement of prickly acacia seed and soil on vehicles and earthmoving equipment.

Threats: Predators, fire and weather

Feral cats are the main introduced predator of concern, with foxes also relevant. Native barn owls take Julia Creek dunnarts, although the species is not considered a readily available prey item. Loss of ground cover can increase exposure to predators.

Fire can remove vegetation cover and alter the availability of soil refuges. Severe floods may drown pouch young or females carrying young, while drought can reduce cover and prey. Heavy rain can also close cracks and make young more vulnerable in vegetation.

  • Manage feral cats and foxes around important populations using approved, site-specific control programs.
  • Plan burns to retain suitable unburnt refuge and ground cover, and avoid creating large areas without shelter.
  • Record recent rainfall, drought, flood and fire when interpreting survey results or deciding whether habitat is occupied.
  • Do not rely on a single survey period after drought, flood or unusual rainfall because detectability and abundance can change markedly.

Threats: Roads and disease

A road-killed Julia Creek dunnart has been reported, so roads can cause direct mortality. The supplied sources do not quantify road mortality or identify road crossing structures that are effective for this species.

Disease is not currently identified as a major threat to wild populations in the supplied sources. A 2024 study documented reproductive disease, lymphoma and other age-related conditions in a captive colony, but these findings should not be extrapolated as evidence of a widespread wild disease threat.

  • Record and report any road-killed dunnarts because carcasses can confirm presence and help identify mortality hotspots.
  • Limit vehicle speeds and night driving through known habitat where practicable, particularly during the breeding and young-rearing period.
  • Use hygiene controls between sites to reduce the risk of transferring weeds and pathogens on trapping, survey and earthmoving equipment.

Survey methodology: 1. Habitat assessment and daytime searches

Survey the Julia Creek dunnart, Sminthopsis douglasi, as a cryptic, nocturnal mammal of Mitchell Grass Downs and related cracking clay grasslands. Use more than one method because capture rates vary widely with rainfall, season, habitat condition and population cycles, and a failed survey does not demonstrate absence.

Habitat: Mitchell Grass, Flinders grass, wire grass and blue grass tussock grasslands on cracking clay plains

Recommended search spacing: 50 to 100 metres between transects

Habitat assessment quadrat: 1 by 1 metre

  • Stratify the project area using current aerial imagery, vegetation mapping, soil information and field inspection, then target treeless or lightly timbered tussock grasslands on ashy or stony cracking clay plains.
  • Search for soil cracks and holes, dense Mitchell grass or Flinders grass cover, and other potential shelter sites along planned transects spaced 50 to 100 metres apart, or in representative 1 by 1 metre quadrats.
  • At trapping or camera sites, record dominant grass species, native ground cover, leaf litter, weed cover, soil crack number and width, and disturbance from grazing, fire and flooding.
  • Give particular attention to ground cover about 40 to 50 centimetres high with 40% to 65% cover, while recognising that the species can occur where cover is sparse or cracks are absent.
  • Do not disturb cracks, holes or vegetation during the search, as these features can provide daytime shelter or nesting sites.

Survey methodology: 2. Elliott trapping

Preferred survey period: April to October, with higher capture success often reported from April to July

Queensland guideline effort: 20 traps, 50 metres apart, for 7 nights, or 140 trap nights per 5 hectares

Earlier species profile effort: Four sites per 5 hectares, 25 traps per site, 20 metres apart, for 4 to 5 nights, or 400 to 500 trap nights

Typical capture rate: Often less than 1% where the species is known to occur, with 0% to 6.75% recorded at Bladensburg National Park

  • Use Type A Elliott or equivalent live traps where live capture, identification or biological sampling is required.
  • For sites up to 5 hectares, use four survey sites per 5 hectares, with 25 traps per site spaced at 20 metre intervals, for four to five consecutive nights, giving 400 to 500 trap nights.
  • For broader project areas, use the Queensland guideline effort of 20 Type A Elliott traps spaced 50 metres apart for seven nights, giving 140 trap nights across 5 hectares per 100 hectares of suitable habitat.
  • Bait with peanut butter and bacon, or peanut butter, rolled oats and bacon, and check and replenish bait daily because ants can remove it.
  • Re-bait completely after the third day and close or move traps with substantial ant activity.
  • Plan trapping for April to October, preferably April to July, in fine weather and outside rainfall periods, very hot conditions and very cold conditions.
  • Check traps from sunrise, before temperatures become hot, and protect animals from heat, cold, dehydration and aerial predators.
  • Close traps during likely rain, because as little as 5 to 10 millimetres can make access difficult and can expose animals to wet and cold conditions.
  • Do not keep an individual in a live trap for more than three consecutive nights, and provide a rest night when repeated recaptures occur.

Survey methodology: 3. Camera trapping

Minimum camera deployment: 5 cameras for 28 nights per 100 hectares of suitable habitat

Camera spacing: 100 metres

Operating period: Just before sunset to just after sunrise

Recommended camera view: Directly downwards, with a focal distance of about 70 centimetres

  • Use camera trapping alongside live trapping where animals occur at low density, where live trapping is unsuitable, or when long-haired rat irruptions may reduce Elliott trap success.
  • For 100 hectares of suitable habitat, deploy five cameras spaced 100 metres apart for at least 28 nights, giving 140 camera trap nights and covering about 5 hectares.
  • Set cameras directly above the ground, preferably with a reduced focal length of about 70 centimetres, high sensitivity, no time delay and at least three photographs per trigger event.
  • Use a scale bar or bait cage of known size in the image so body size, proportions and fur colour can be assessed against the two other dunnarts in the area.
  • Run cameras from just before sunset to just after sunrise, and deploy them after the Elliott trapping survey rather than concurrently.
  • Use a secure peanut butter, rolled oats and bacon bait station, firmly attached below the camera, and trim vegetation in the focal area to reduce false triggers.

Survey methodology: 4. Thermal imaging and indirect evidence

Thermal transect effort: More than 2.5 kilometres in suitable habitat

Useful predator remains: Dasyurid jaw bones at least 20 millimetres long

  • Use vehicle based thermal imaging on fixed width night transects in suitable grassland, followed by handheld thermal inspection and visual or capture confirmation where possible.
  • Survey more than 2.5 kilometres of transect when using the Queensland guideline effort, record distance and strip width, and define strip width from the maximum detection distance through the ground cover.
  • Work in minimal wind, no rain and moderate night temperatures, and allow soil surfaces to cool for more than 1.5 hours after dusk where the thermal unit has limited sensitivity.
  • Search owl pellets, predator scats, carcasses and feral cat or fox gut contents when available, and look for dasyurid mandibles at least 20 millimetres long.
  • Use wildlife detection dogs only with an experienced team, reference odour samples for S. douglasi and the two sympatric dunnarts, and corroborate indications with cameras or live trapping.
  • Treat hair samples as supporting evidence only, because hair analysis may not reliably separate all Sminthopsis species.

Survey methodology: Minimum effort and reporting

Minimum area assessed: 10% of suitable habitat in the project area

Minimum live trapping share: At least 50% of the survey effort

Breeding season: September to mid-February

  • Assess at least 10% of suitable habitat in the project area using the prescribed fauna methods, with at least 50% of that survey effort consisting of Elliott type live trapping.
  • For each 100 hectares of suitable habitat, aim for 10 habitat quadrats, 140 Elliott trap nights and 140 camera trap nights, scaled to the area and habitat strata.
  • Record survey dates, weather, rainfall, ground cover, soil condition, fire and grazing history, trap and camera locations, effort, bait, captures, photographs and all non-detections.
  • Identify small dunnarts carefully, especially animals weighing 26 grams or less, because juvenile S. douglasi can overlap with adult S. macroura.
  • Use hindfoot length, facial and ear markings, tail tip and body size for identification, and check premolar development when a small individual cannot be identified confidently.
  • Submit confirmed records, survey effort and reliable negative results to the relevant Queensland wildlife database or other relevant state wildlife database, and retain the voucher photographs and authority details with the project records.
  • Do not rely on a single short survey, a March trapping result or a non-detection to clear habitat, because capture success can be very low and can vary between years.

Before habitat disturbance: Avoid and minimise

Treat suitable cracking clay grassland as potential habitat even when the species has not been recorded on the project area. The species is patchy, occurs at low density and can be missed by standard surveys.

Key habitat features: Cracking clay soils, soil cracks and holes, native tussock grass cover and low prickly acacia density

Important known areas: Bladensburg National Park, Moorrinya National Park, Toorak Research Station, Proa or Yorkshire Downs and the Julia Creek aerodrome

  • Avoid clearing Mitchell Grass Downs, Gulf Plains and Desert Uplands tussock grassland where practicable, particularly areas with suitable soil structure, dense ground cover and known records.
  • Retain connected areas of cracking clay grassland rather than isolated small remnants, and keep retained habitat connected to nearby grassland, stock routes, reserves and other known or potential habitat.
  • Retain soil cracks and holes, dense native ground cover and low weed cover, because these features provide shelter from heat, fire and predators in dry periods and cover when cracks close after rain.
  • Keep new roads, laydown areas, water points and other disturbance away from retained habitat where practicable, because grazing pressure and watering points can reduce habitat suitability.
  • Control the spread of prickly acacia and other weeds before machinery enters retained habitat, and prevent soil, seed and weed transfer between sites.

Before habitat disturbance: Approvals and work planning

Preferred project survey and clearing period: April to October, subject to approvals, weather and animal welfare

Period to avoid where practicable: December to February, when rainfall, flooding and dependent young can increase risks

  • Check the current EPBC Act listing and refer the action to the Australian Government where the clearing, construction or infrastructure work may have a significant impact on the species or its habitat.
  • Check Queensland Nature Conservation Act requirements, scientific or animal ethics requirements and any permits needed for trapping, handling, movement, photography or biological sampling.
  • Prepare a species specific fauna management plan covering survey design, exclusion areas, clearing sequence, fauna spotter catcher roles, weather triggers, release sites, incident response and reporting.
  • Schedule clearing and trapping for April to October where practicable, avoiding the September to mid-February breeding season and the December to February wet season.
  • Stage clearing so that survey results can inform later work areas, and retain escape routes to nearby suitable habitat that will not be cleared.
  • Use a qualified ecologist to review the plan when survey conditions are poor, suitable habitat is extensive or the project has records of the species.

Before habitat disturbance: Site preparation

  • Complete the targeted survey before disturbance and repeat the pre-clearance inspection immediately before clearing if conditions, timing or work areas have changed.
  • Mark retained habitat, exclusion zones, access routes, trap lines, camera locations and any confirmed or suspected shelter areas on plans and in the field.
  • Install exclusion fencing only where it will direct animals towards retained suitable habitat and will not trap animals in the work area or block movement across grassland.
  • Brief all machine operators, supervisors and fauna spotter catchers on the species appearance, nocturnal habits, habitat features, stop-work rules and handling limits.
  • Check that machinery, vehicles, traps, cameras and fencing are clean and do not spread prickly acacia, pathogens or contaminated soil into retained habitat.

During habitat disturbance: Daily controls and machinery interface

Use a controlled clearing system that gives animals a safe route away from machinery and allows the fauna spotter catcher to stop work quickly. The spotter catcher must have authority to stop the activity whenever an animal, shelter feature or welfare risk is encountered.

  • Hold a daily pre-start meeting covering the approved clearing footprint, weather, escape routes, retained habitat, fauna risks and the spotter catcher's stop-work authority.
  • Inspect the work area at first light and before each new clearing stage for animals, fresh signs, soil cracks, holes, carcasses, predator remains and changes in water or ground conditions.
  • Clear in small, staged sections towards retained habitat, keep machinery speed low enough for the spotter catcher to inspect the active edge, and do not work ahead of the approved inspection area.
  • Keep the spotter catcher in direct communication with the operator and supervisor, and stop machinery immediately when the spotter catcher calls stop.
  • Do not clear during rain, when rain is forecast and traps or animals may become wet, or when heat, cold, flooding or poor visibility could compromise animal welfare.
  • Keep vehicles and workers out of retained grassland unless access is approved, and prevent light, noise, dust and night works from extending into retained habitat where practicable.

During habitat disturbance: Habitat features and fauna handling

Known daytime shelter depth: Animals have been recorded in soil cracks and holes at depths of 10 to 50 centimetres

Activity pattern: Nocturnal

  • Treat soil cracks and holes as occupied shelter features and do not insert tools, fill them, collapse them or use them as vehicle routes during clearing.
  • Inspect grass tussocks and dense vegetation before removal, especially where heavy rain has closed soil cracks and animals may be sheltering above ground.
  • Do not excavate, dismantle or relocate shelter features unless this is covered by the approved fauna management plan and directed by the authorised fauna spotter catcher.
  • Only an appropriately authorised and experienced person may capture, handle, mark, sample or relocate the species, and handling must be limited to what is needed for identification, welfare and approved conservation action.
  • Place a captured animal in a secure, dark, ventilated container suitable for its size, protect it from heat, cold and dehydration, and minimise handling time.
  • Release an uninjured animal at dusk into nearby retained habitat with suitable cracking clay grassland, native ground cover and shelter, at a location approved in the fauna management plan.
  • Do not release an animal into a cleared work area, beside machinery, near active roads, into dense prickly acacia or into habitat that is flooded, burnt or otherwise unsuitable.
  • Take injured or heat stressed animals to an experienced wildlife veterinarian or authorised wildlife carer as soon as practicable, and record the outcome.

During habitat disturbance: Stop work, records and incidents

  • Stop work in the immediate area if a Julia Creek dunnart is found, if an occupied shelter feature is exposed, or if the approved escape route is blocked.
  • Stop work if flooding, rain, extreme temperatures, fire, machinery failure or an unauthorised activity creates an animal welfare or habitat risk.
  • Keep the animal, feature and machinery undisturbed until the fauna spotter catcher or project ecologist gives a clear instruction to proceed.
  • Record the date, time, location, habitat, weather, work activity, identification evidence, animal condition, actions taken, release or treatment location, personnel and photographs.
  • Report deaths, injuries, repeated captures, unexpected records, clearing outside the approved footprint and any breach of a condition to the project environmental officer and relevant approval authority.
  • Update the clearing sequence and controls when a record shows that the species is using the work area or when retained habitat is less suitable than predicted.

After habitat disturbance: Post-clearance checks

The project must show that retained habitat remains functional and that clearing controls have not left animals, shelter features or connected habitat exposed to avoidable risks.

  • Inspect each cleared stage and its boundary after machinery leaves, including soil cracks, holes, tussocks, spoil piles, drains, fences and access tracks.
  • Check retained habitat for trapped animals, injured fauna, collapsed shelter, soil compaction, weed spread, erosion, sediment, dust and unplanned vehicle access.
  • Remove temporary exclusion fencing, bait stations, traps and other equipment only after confirming that animals are not confined and that removal will not damage retained habitat.
  • Complete a final fauna and habitat inspection before demobilisation and document all outstanding actions.

After habitat disturbance: Monitoring and reporting

Suggested monitoring period: April to July, with repeat monitoring suited to the approved project duration

Monitoring measures: Detection, survey effort, native ground cover, soil cracks, weeds, grazing, fire, flooding and predator activity

  • Monitor retained and rehabilitated habitat during the breeding and post-breeding periods, with repeat surveys timed for April to July where practicable because capture success is often higher then.
  • Use the same trapping, camera and habitat assessment methods at each monitoring event so changes in detection, ground cover, weeds, grazing and soil condition can be compared.
  • Record effort as trap nights and camera trap nights, not only the number of animals detected, and report zero detections with the associated weather and habitat conditions.
  • Submit confirmed records and survey results to the relevant state wildlife database and meet all EPBC Act, Queensland approval, permit and fauna handling reporting conditions.
  • Review the monitoring results with the project ecologist and change weed, grazing, predator, fire, access or rehabilitation controls if habitat condition declines.

After habitat disturbance: Offsets and residual impacts

  • Calculate residual loss of suitable habitat after avoidance, minimisation, rehabilitation and monitoring measures have been applied.
  • Where an approval requires an offset, design it around secure protection and improvement of suitable Julia Creek dunnart habitat, not only an area calculation.
  • Prioritise offset actions that retain cracking clay grassland, improve native ground cover, control prickly acacia and feral predators, and reduce damaging grazing pressure.
  • Tie offset management to measurable outcomes, a defined duration, responsible land managers, reporting dates and contingency actions.

Rehabilitation actions: Rebuild habitat structure

Rehabilitation should recreate the ground conditions and shelter structure used by the Julia Creek dunnart. Replacing vegetation alone is not enough if soil cracking, ground cover, weed control and predator pressure are not addressed.

Target ground cover reference: About 40 to 50 centimetres high and 40% to 65% cover

Native plant examples: Mitchell grass, Astrebla spp., Flinders grass, Iseilema spp., wire grass and blue grass

  • Reinstate deep, stable clay soil profiles and natural drainage after excavation, avoiding compaction that prevents the formation of cracks and holes.
  • Restore native tussock grassland using locally appropriate Mitchell grass, Astrebla spp., and other native grass species recorded in the surrounding habitat, including Flinders grass, Iseilema spp., wire grass and blue grass where suitable.
  • Aim for native ground cover that develops towards about 40 to 50 centimetres high and 40% to 65% cover, while using local reference sites to set realistic staged targets.
  • Retain or reinstate surface roughness, grass tussocks and small shelter features without creating hazards for fauna or machinery.
  • Do not introduce trees or dense woody vegetation into areas that were naturally treeless or lightly timbered grassland.

Rehabilitation actions: Manage weeds, predators, fire and grazing

  • Control prickly acacia and other weeds before they dominate the understorey, and inspect machinery and rehabilitation materials to prevent reinvasion.
  • Control feral cats and foxes through approved, humane and coordinated programs where monitoring shows predator pressure is affecting retained or rehabilitated habitat.
  • Manage grazing so that native ground cover and soil cracks are retained, and keep stock and watering points away from restoration areas while plants and soil structure recover.
  • Prepare a fire management regime that protects soil cracks, refuges and sufficient native ground cover, and avoid broad intense burns across all retained or rehabilitated habitat at once.
  • Use rainfall, ground cover, soil condition and dunnart detections to adjust grazing, fire, weed and predator actions rather than applying one treatment across all sites.

Rehabilitation actions: Monitoring and success measures

Key success indicators: Native tussock cover, functioning soil cracks and holes, low weed cover, controlled grazing and fire, predator management and connected habitat

Evidence standard: Report method, timing, effort, weather, habitat condition and detections for every monitoring event

  • Inspect rehabilitation at least after establishment, after major rainfall and during the approved monitoring period for plant cover, species composition, soil cracks, weed cover, erosion, grazing and fire impacts.
  • Compare rehabilitation plots with nearby reference grassland and record the area and condition of connected habitat, not only plant survival within the work footprint.
  • Use camera trapping, Elliott trapping, thermal imaging or predator remains searches where approved and safe, with survey effort reported so non-detection can be interpreted correctly.
  • Set success measures that include stable native tussock cover, suitable cracking clay soil, low prickly acacia cover, controlled grazing and fire, and evidence that animals can move between retained and rehabilitated habitat.
  • Continue corrective actions until the site supports the approved habitat targets and no longer creates a direct welfare or access risk for the species.

Sources