Grey-bellied Dunnart (Sminthopsis fuliginosa)

Conservation status

Commonwealth, EPBC Act 1999: Endangered.

Breeding season

The seasonal pattern below is based mainly on Kangaroo Island observations and related Sminthopsis biology. Exact timing may vary between populations and years.

The seasonal pattern is indicative rather than a substitute for a site specific survey. The supplied sources do not establish torpor or hibernation for this species.

• Breeding: Mating is inferred to occur from mid September to early October and again from November to December. An adult female carrying eight pouch young was recorded in October.
• Young: Young are expected in the pouch or dependent after mating. Related Common Dunnart young leave the pouch at about 65 days, but a species specific timing is not available.
• Juvenile dispersal and movement: Juveniles and subadults made up more than half of live captures from January to May in one survey. Dispersal from the natal range is poorly understood. Adults may use several den sites over periods of weeks.
• Adult activity and den use: The species is nocturnal and uses preferred ground dens throughout the year. Camera and radio tracking recorded repeated visits and occasional communal den use.
• Highest clearing risk: Clearing risk is highest from September to December because mating, pouch young and early dependent young are likely during this period. Disturbance can also harm animals or dens at other times because den use occurs year round.

Identification

This is a very small dasyurid with sooty grey upperparts, paler underparts and a slender, bicoloured tail. Identification can be difficult because dunnarts are cryptic, nocturnal and similar in size and shape to other small dasyurids.

The Kangaroo Island form is 80 to 100 mm long and weighs up to 26 g. It has dark sooty grey fur on the back, light grey fur below, a slender pointed muzzle, wide square shaped ears and a thin tail that is longer than the body and clearly bicoloured.

Identification: Separating similar animals

The Kangaroo Island form resembles the Common Dunnart, Sminthopsis murina, and was originally misidentified as that species. On Kangaroo Island it is the only dasyurid, which assists field identification, but mainland identification can require closer examination and, in some cases, genetic sampling to separate forms within the Sminthopsis griseoventer and Sminthopsis fuliginosus complex.

Identification: Field detection and signs

The supplied sources do not provide diagnostic descriptions of tracks, scats, calls or eyeshine. The animal is cryptic and has been detected by pitfall trapping and camera trapping, including at low density sites. Camera traps are useful where trapping is unsuitable, and predator scats or owl casts may provide supplementary evidence of small mammal presence, although they do not by themselves confirm this species.

• Use experienced personnel for handling and identification because small dunnarts can resemble other dasyurids and introduced House Mice.
• Treat a suspected record as requiring confirmation when the location is outside Kangaroo Island or when more than one dunnart species may occur.
• Use appropriate permits, animal welfare procedures and government survey guidance before trapping.

Distribution: National range and taxonomy

The name Grey-bellied Dunnart is used inconsistently across the supplied sources because the taxonomy of the Sminthopsis griseoventer and Sminthopsis fuliginosus complex remains unresolved. The strongest site specific information concerns the Kangaroo Island form, which is endemic to South Australia.

The Kangaroo Island form occurs only on Kangaroo Island, South Australia. The nominate grey bellied form is reported from the Eyre Peninsula in South Australia and south western Western Australia. A related Boullanger Island form occurs on Boullanger Island in Jurien Bay, Western Australia, although its taxonomic separation from mainland animals has been questioned.

• Confirm the accepted taxon and relevant survey method with the relevant state authority before relying on a mainland record for regulatory assessment.
• Do not assume that a record labelled Grey-bellied Dunnart is the Kangaroo Island form.

Distribution: Known strongholds and subpopulations

For the Kangaroo Island form, all records after the historical 1969 to 1976 records have been from western Kangaroo Island. Important areas include Flinders Chase National Park, including the Gosselands section, Ravine des Casoars Wilderness Protection Area and privately owned native vegetation north and east of these areas.

• The form has also been detected near the West End Highway and on private land, so protected area boundaries should not be used as a substitute for habitat assessment.

Distribution: Population and trend

The sources do not provide a reliable national population estimate. The available evidence indicates a small population, restricted distribution and continuing decline in distribution, habitat quality and mature individuals for the Endangered Kangaroo Island form.

1999 to 2001 targeted survey: 22 Kangaroo Island dunnarts were captured at six sites in Flinders Chase National Park.

2009 follow up: Three individuals were recorded at some of those sites.

2017 to 2018 camera surveys: The form was detected at five sites in western Kangaroo Island.

2018 to 2019 private land surveys: It was detected on 42 occasions at six private land sites in western Kangaroo Island.

Habitat: Vegetation communities

The species is associated with native vegetation that provides a developed ground layer, prey and shelter. Habitat information is best established for the Kangaroo Island form and is less certain for mainland populations.

The Kangaroo Island form depends on woodland and mallee heath. Potential habitat includes low, open mallee woodland dominated by Kangaroo Island mallee ash and coastal mallee, and brown stringybark woodland.

• Include native vegetation remnants and regrowth in pre clearing habitat assessments, even where the canopy is open or the site is within private property.
• Treat occupied and potential habitat within the known distribution as habitat that may require protection or targeted survey.

Habitat: Structure and climate

Occupied and potential habitat commonly has annual rainfall greater than 600 mm and includes a mixture of post-fire successional stages. Early to mid stage regrowth, approximately 5 to 20 years after fire, is identified as important potential habitat.

• Record the age and structure of regrowth, shrub density, ground cover, fallen timber and available den sites during site inspections.
• The supplied documents provide no elevation limit for this species.

Foraging habitat: Diet

The Grey-bellied Dunnart is an opportunistic, ground forager. It hunts mainly at night among litter, low vegetation and other ground cover where invertebrate prey is available.

Dunnarts eat mostly ground dwelling invertebrates. Reported prey for the Kangaroo Island form includes spiders and ants, with beetles, scorpions, centipedes and grasshoppers also recorded. Small reptiles and other small vertebrates may occasionally be taken.

Foraging habitat: Foraging areas and range

The sources do not identify particular food plants. Foraging habitat is likely to be concentrated in native woodland, mallee heath and post-fire regrowth that support ground invertebrates and provide cover. Radio tracking of the Kangaroo Island form indicated home ranges about 200 to 300 m wide, but the sources do not give a separate foraging distance or density.

• Retain connected native ground cover, litter and low shrub structure where practicable during clearing and construction.
• Avoid treating recently burnt or regenerating areas as unused habitat. Early to mid stage regrowth is identified as potential habitat for the Kangaroo Island form.

Breeding habitat: Breeding biology

Breeding information is limited for the Grey-bellied Dunnart, particularly for the taxon as currently treated in different sources. Available observations indicate rapid reproduction and more than one breeding attempt may occur in a year.

The Kangaroo Island form is likely to have reproductive biology similar to related Sminthopsis species. Females may be polyoestrous, and juveniles have been recorded in more than one cohort within the same year. A female carrying eight pouch young was recorded in October 2021.

Breeding period: Likely mating periods include mid September to early October and November to December, based on juvenile ages recorded in surveys.

Litter size: Presumed to be 6 to 8 young; eight pouch young were observed in October 2021.

Gestation: Approximately 12 days in the closely related Common Dunnart. A species specific value was not provided.

Pouch period: Young of the closely related Common Dunnart leave the pouch at approximately 65 days. A species specific value was not provided.

Age at maturity: Not provided in the supplied sources.

Lifespan: Most males of related Sminthopsis species may rarely survive beyond their first year after breeding. Females probably live for one or two breeding seasons.

Breeding habitat: Breeding habitat

Breeding animals require native woodland, mallee heath or related native vegetation with ground level cover and secure den sites. The recorded female with pouch young was trapped in an area burnt in 2019 to 2020, showing that recently burnt habitat can still support breeding when shelter remains available.

• Inspect ground cover, dense low vegetation, litter, fallen material and existing dens before clearing native habitat.
• Avoid disturbing occupied den sites, particularly from September to at least March, when mating and young are most likely under the available evidence.

Sheltering habitat: Dens and ground cover

Dunnarts use ground level shelters rather than tree hollows or elevated roosts. Shelter sites can be reused and may be shared by several animals.

Radio tracked Kangaroo Island dunnarts used several den locations over six weeks, with some individuals returning to preferred dens continuously or on consecutive days. Camera trapping recorded communal use of dens by up to four dunnarts and, on one occasion, two mice.

• Check under dense ground vegetation, fallen timber, litter and other concealed ground structures before mechanical clearing.
• Mark and protect den locations where detected, and avoid storing materials or placing machinery over likely shelter sites.
• Do not assume a den is vacant because an animal is not visible during daytime checks.

Sheltering habitat: Fire refuge

Some dens appear to provide protection during fire, including in areas that were completely burnt. The supplied sources do not provide den dimensions, burrow depth or a confirmed nest construction description for the Grey-bellied Dunnart.

• Retain unburnt patches, ground cover and potential den structures within and beside work areas where fire has recently affected the site.
• Use camera monitoring or carefully managed trapping when occupancy needs to be established.

Threats: Clearing and fragmentation

The main documented risks are habitat loss, declining habitat quality, severe fire and predation. The small range and restricted population of the Kangaroo Island form make local clearing or repeated disturbance potentially significant.

Historical records from central and eastern Kangaroo Island were associated with areas undergoing intensive land clearing. The Endangered listing is based in part on restricted distribution and continuing decline in habitat quality and mature individuals.

• Avoid clearing native woodland, mallee heath and early to mid stage regrowth within the known range without a targeted assessment.
• Retain habitat connections and avoid isolating small remnants, particularly around known western Kangaroo Island records.
• Survey private native vegetation as well as reserves because the Kangaroo Island form has been recorded on both.

Threats: Fire and predators

The Kangaroo Island form is susceptible to severe fire, although some individuals can persist in burnt areas where dens provide refuge. Feral cats are identified in the supplied material as a relevant conservation management issue, and predation risk can change after fire.

• Plan clearing and construction fire controls to prevent escapes and avoid removing denning cover unnecessarily.
• Control or manage feral predators in accordance with approved recovery and biosecurity programs where relevant to the project.

Threats: Roads, disease and hydrology

The sources record animals near the West End Highway but do not quantify road mortality or establish roads as a population level threat. They also do not provide species specific evidence for hydrology change as a threat. Field hygiene is relevant to conservation work on Kangaroo Island, where a Phytophthora fieldwork hygiene protocol is cited.

• Treat road construction, widening and repeated vehicle access as potential sources of direct disturbance and habitat loss, but document any road mortality rather than assuming its scale.
• Apply project hygiene procedures when moving soil, vehicles or equipment between native vegetation sites, particularly on Kangaroo Island.
• Assess drainage changes where they could alter native vegetation or ground cover, while recognising that the supplied sources do not quantify a species specific hydrology response.

Survey methodology: 1. Check habitat and survey conditions

Survey design must account for the species' low detectability, nocturnal activity and uncertain taxonomy. The available guidance is strongest for the Kangaroo Island form and related grey-bellied dunnart records, so confirm the identification and apply state survey requirements for the project area.

Inspect all native woodland, mallee heath and dense native shrub habitat within and near the proposed disturbance area.

Give particular attention to areas with annual rainfall above 600 mm and a mix of post-fire stages, including early to mid-stage regrowth about 5 to 20 years after fire.

Record dense ground cover, logs, rocks, burrows, den sites and connected native vegetation that may provide shelter or movement habitat.

Treat occupied and potential habitat as requiring assessment where the project occurs within the known range of the relevant taxon.

Activity pattern: Nocturnal

Typical body size: 80 to 100 mm long and up to 26 g for the Kangaroo Island form

Reported home range width: Approximately 200 to 300 m for the Kangaroo Island form

Survey methodology: 2. Camera trapping

Use motion-activated camera traps in suitable habitat, especially where the animal is cryptic or expected to occur at low density.

Place cameras beside likely travel routes, dense ground cover, den entrances, logs and other shelter features identified during the habitat inspection.

Run cameras during the species' nocturnal activity period and retain images of all small dasyurids for independent identification.

Use camera trapping alongside, rather than as a replacement for, other methods where the species may be difficult to detect.

Survey methodology: 3. Live trapping and sign searches

Use pitfall trapping where permitted and where the substrate and ground cover make the method safe and suitable for small dasyurids.

Conduct daytime searches for suitable shelter, tracks, scats, den sites and other signs before placing traps or cameras.

Collect predator scats, owl casts or remains where this can be done lawfully and without disturbing active nests or dens.

Use sand trays or soil plots where sandy substrates are absent, if the method is suitable for the site and approved by the relevant authority.

• Obtain all required animal ethics, wildlife research and land access approvals before live trapping or collecting biological material.
• Provide shelter, shade and checking arrangements that prevent exposure, dehydration, predation and unnecessary stress.
• Use experienced personnel who can distinguish the target animal from other small mammals and identify when genetic confirmation is needed.

Survey methodology: 4. Timing and detection records

Survey at night for this nocturnal species and record the start and finish time, weather, moon conditions, recent rain, wind, temperature and fire history.

Avoid relying on a single survey night because low capture and detection rates have been reported for the Kangaroo Island form, including one individual during a 1990 island-wide survey and 22 individuals at six sites from 1999 to 2001.

Repeat surveys across suitable seasonal conditions where the first survey does not detect the species but habitat remains suitable.

Record the exact location, method, effort, observer, camera deployment period, trap effort, identification confidence and photographs or genetic samples where collected.

Survey methodology: Minimum effort and reporting

• Prepare a site-specific survey plan that states the area searched, number and type of cameras, number of trap nights, number and length of transects, search hours, weather and conditions, and the reason for selecting the effort.
• Do not describe a site as free of the species solely because one method or one night produced no detection.
• Have a qualified mammal ecologist review records and resolve uncertain identifications before clearing starts.
• Submit confirmed records, negative survey results where required, photographs and spatial data to the relevant state or territory wildlife database.
• Report any record that may represent a taxonomic form requiring genetic confirmation to the relevant wildlife authority and project approval body.
• Retain permits, field sheets, trap checks, camera records, animal welfare records and laboratory results with the project environmental file.

Before habitat disturbance: Avoid and minimise habitat loss

Plan the project on the assumption that suitable native vegetation may support the species, even when no animal is detected. The Kangaroo Island form is associated with woodland and mallee heath, dense shelter and a range of post-fire regrowth stages.

Avoid clearing native woodland, mallee heath and dense shrub cover wherever practicable.

Retain connected native vegetation between occupied or potentially occupied areas, and avoid isolating small patches with roads, compounds or linear infrastructure.

Retain logs, rocks, dense ground cover, burrows, den sites and other shelter features within areas that will not be cleared.

Use the smallest practicable disturbance footprint and keep machinery, spoil, lighting and vehicle movements out of retained habitat.

• Define site-specific exclusion zones around retained habitat and shelter features before mobilisation.
• Do not apply a generic buffer distance where the documents do not specify one; set the distance using the habitat assessment, approval conditions and advice from the relevant authority.

Before habitat disturbance: Approvals and work planning

Check the current national listing and taxonomic treatment before finalising the assessment because the grey-bellied, dusky and Kangaroo Island forms have had changing taxonomic treatments.

Refer the action under the EPBC Act if the proposed action may significantly impact the nationally listed taxon or its habitat.

Check state or territory threatened species listings, wildlife permits, animal ethics requirements and licences for trapping, handling, genetic sampling and relocation.

Prepare a fauna management plan that identifies survey methods, exclusion zones, clearing sequence, fauna spotter catcher responsibilities, stop-work triggers, treatment of injured animals and reporting pathways.

• Confirm whether the project is within the known range of the taxon being assessed rather than assuming that a record of a related form applies everywhere in Australia.
• Include the current approval conditions and relevant state wildlife database requirements in the fauna management plan.

Before habitat disturbance: Prepare the site

Complete pre-clearance surveys in suitable habitat before vegetation removal, using the approved methods and timing in the project survey plan.

Mark retained vegetation, shelter features, exclusion zones, access routes and no-go areas on the ground and in the machinery briefing maps.

Stage clearing so that the fauna spotter catcher can inspect each work area immediately before disturbance.

Install exclusion fencing only where it will direct animals away from machinery and towards retained habitat without trapping them or blocking necessary movement.

• Brief all operators on the species' small size, nocturnal activity, likely shelter sites and the requirement to stop when the fauna spotter catcher gives a stop signal.
• Schedule the most disruptive works to avoid periods identified by the ecologist as having high breeding, sheltering or post-fire habitat value.
• Do not burn, slash or mulch habitat before the pre-clearance inspection and fauna management controls are in place.

During habitat disturbance: Daily controls and machinery coordination

Clearance must proceed in small, inspected stages under the direction of the fauna spotter catcher. The animal is small, ground dwelling and nocturnal, so machinery operators may not see it without active supervision.

Inspect the work area each day before machinery enters and repeat the inspection after any overnight pause, changed work front or new stockpile.

Use a fauna spotter catcher who has authority to stop machinery immediately when an animal, den, burrow, shelter feature or unexpected habitat risk is identified.

Keep machinery at walking speed during initial vegetation removal and maintain direct communication between the spotter catcher, supervisor and operators.

Clear from the retained habitat towards the already disturbed area where this is safe and consistent with the approved fauna management plan.

Keep excavations, trenches, pipes and pits covered or fitted with escape ramps at the end of each shift.

• Stop work if the spotter catcher cannot maintain a safe view of the work front or if dust, darkness, noise or machinery movements prevent effective searching.
• Record the date, area cleared, machinery used, personnel present, weather, habitat features removed and all fauna observations.

During habitat disturbance: Handle habitat features carefully

Search dense ground cover, logs, rocks, burrows and den sites before they are moved, crushed, mulched or buried.

Inspect hollow-bearing trees and fallen timber before removal, although the available documents identify ground shelter and den use more clearly than tree hollow use for this dunnart.

Where approved, move suitable logs, rocks and other shelter features into nearby retained habitat rather than removing them from the site.

Protect retained woodland, mallee heath, native shrub cover and post-fire regrowth from vehicle damage, soil compaction and accidental fire.

• Do not enter or block a suspected burrow or den without direction from the fauna spotter catcher and the approved fauna management plan.
• Treat communal or repeatedly used den sites as occupied until inspected and cleared by an authorised person.
• Keep fauna handling and relocation separate from routine earthworks, stockpiling and waste disposal.

During habitat disturbance: If an animal is found

Stop work immediately within the animal's vicinity and keep people, vehicles and dogs away from it.

Allow the animal to move into nearby retained habitat if it can do so safely and without being chased or exposed to machinery.

Only an appropriately authorised and experienced person may capture, handle, identify, mark, sample or relocate the animal.

Relocate an animal only where authorised and only to nearby suitable retained habitat identified in the fauna management plan.

Place an injured or distressed animal in a secure, quiet, temperature-appropriate container and arrange prompt assessment by a veterinarian or licensed wildlife carer.

Do not handle a pouch young, lactating female or animal showing signs of heat stress unless necessary to prevent immediate harm and the handler is authorised.

• Stop work and seek advice from the project ecologist and relevant wildlife authority if the animal cannot be safely captured, identified or released.
• Stop work across the affected work area if repeated detections, an active den, a breeding female or unexpected habitat use indicates that the approved controls are not adequate.
• Record the location, time, condition, sex if known, photographs, handling time, release location, personnel, permit details and outcome for every animal encounter.

After habitat disturbance: Complete and verify clearance

Post-clearance checks must confirm that no animals, active dens or trapped fauna remain in the work area and that retained habitat controls are still functioning. Use the results to adjust the work method before the next stage.

Carry out a final inspection of the cleared area, excavations, stockpiles, drainage structures and retained habitat edges after machinery leaves.

Check that no animal has been left in a trench, pipe, pit, tree debris pile or isolated patch of vegetation.

Remove temporary fencing only after the fauna spotter catcher and site supervisor confirm that it will no longer protect fauna or direct movement safely.

Keep access controls, covers and escape devices in place for any open excavation or construction area that remains active.

• Document the final cleared boundary, retained habitat, habitat features moved, fauna encounters and unresolved risks before signing off the work area.
• Reinspect the area after heavy rain, fire, flooding, major earthworks or any change that could expose or trap fauna.

After habitat disturbance: Monitor retained and restored habitat

Use camera traps or other approved methods to check retained habitat where the species was detected, where a den was found or where clearing may have changed access to shelter.

Monitor connected native vegetation and post-fire regrowth around the work area for continued shelter and movement opportunities.

Compare detections, habitat condition, weed cover, predator activity and disturbance against the baseline survey and approval requirements.

Investigate any decline in detections or loss of shelter features with the project ecologist and relevant authority.

• Set monitoring duration and frequency in the fauna management plan because the supplied documents do not specify a universal post-construction monitoring period for this species.
• Do not interpret non-detection as proof of absence without considering nocturnal activity, low density, weather, survey effort and the taxonomic identification issue.

After habitat disturbance: Reporting and corrective action

Submit required fauna records, injured animal outcomes, relocations, permit returns and monitoring results to the approval body and relevant state or territory database.

Report any breach of an exclusion zone, unexpected mortality, repeated den destruction or failure of fauna controls to the project environmental officer and regulator as required.

Maintain weed, erosion, vehicle access, lighting and waste controls around retained habitat for the life of the project.

Apply approved offsets or other compensatory measures where residual impacts remain after avoidance and mitigation.

• Update the fauna management plan when monitoring identifies a new risk or when construction methods change.
• Keep spatial records of retained habitat and restoration areas so future maintenance does not remove shelter or connectivity.

Rehabilitation actions: Restore native cover and shelter

Rehabilitation should recreate the ground structure and native vegetation used by the species, not just replace canopy cover. The strongest available habitat information identifies woodland and mallee heath, dense shelter and a mixture of post-fire successional stages.

Replant or allow regeneration of locally appropriate native woodland, mallee and heath species that match the pre-disturbance community.

Recreate dense native shrub and ground cover, including a mix of young regrowth and older vegetation rather than a uniform stand of plants.

Retain and reinstate logs, coarse woody debris, rocks, leaf litter and other ground shelters in connected groups.

Protect restored areas from compaction, uncontrolled access, grazing and weed invasion during establishment.

• Use local provenance plants where required by the approval or restoration specification.
• Do not introduce plants or substrates that are not appropriate to the local vegetation community or that could increase fire, weed or predator risks.

Rehabilitation actions: Manage fire, weeds and predators

Plan fire management to retain a mosaic of vegetation ages, including early to mid-stage regrowth about 5 to 20 years after fire where this matches local habitat requirements.

Avoid broad, high-intensity burns that remove all ground shelter from a restored area.

Control weeds that reduce native shrub cover, ground shelter or movement between retained patches.

Manage introduced predators and other pest animals in accordance with approved programs and state requirements, especially where monitoring shows increased predator use of cleared or burnt areas.

• Do not use fire as a routine rehabilitation treatment without a site-specific ecological assessment and approval.
• Record fire history, weed treatment, predator control and resulting habitat condition in the restoration monitoring data.

Rehabilitation actions: Measure rehabilitation success

Monitor plant survival, native ground cover, shrub density, logs and shelter features, weed cover, erosion, predator activity and connectivity with retained habitat.

Use motion-activated cameras and other approved survey methods to test whether dunnarts use restored habitat after vegetation and shelter have established.

Set success measures in the restoration plan using the pre-clearing habitat condition and approved reference sites rather than an invented universal density or occupancy target.

Continue maintenance and monitoring until the approval body's success criteria are met and the area is unlikely to become a trap or barrier for fauna.

• Record the date, survey method, effort, weather, camera locations, detections and habitat condition at every monitoring event.
• Review the restoration method if no dunnart use is recorded, shelter is lost, weeds dominate, predators increase or restored areas become isolated.
• Use genetic or expert review where photographs cannot distinguish the target taxon from a related dunnart.

Sources

• Grey-bellied Dunnart: <i>Sminthopsis fuliginosa</i>, researchers.cdu.edu.au: https://researchers.cdu.edu.au/en/publications/grey-bellied-dunnart-isminthopsis-fuliginosai/
• Department of Climate Change, Energy, the Environment and Water, Australian Government DCCEEW: http://www.environment.gov.au/biodiversity/threatened/species/pubs/92894-conservation-advice-27082023.pdf
• [DOC] consultation-document-sminthopsis-riseoventer-aitkeni.docx, Australian Government DCCEEW: https://www.agriculture.gov.au/sites/default/files/documents/consultation-document-sminthopsis-riseoventer-aitkeni.docx
• Survey guidelines for Australia's threatened mammals, Australian Government DCCEEW: https://www.agriculture.gov.au/sites/default/files/documents/survey-guidelines-mammals.pdf