GeezerVison

Aims and progress

The NKMP aims to:

• deliver a robust estimate of koala populations and
• build a long-lasting capability to monitor and assess trends in koala populations.

Get involved

NKMP have developed two apps for koala monitoring.

Koala estimates

CSIRO’s data-driven population model has been built based on tens of thousands of data points spread across the species range. As of May 2025:

• The latest data-driven estimate for the size of the listed (NSW, ACT, QLD) koala population is between 398,000 and 569,000.
  • This estimate is substantially greater than previous NKMP estimates, arising from the incorporation of additional data from across a wider area and model improvements.
• The latest population estimate for the unlisted koala population (VIC, SA) is between 303,000 and 381,000 individuals.
  • This is also higher than previous NKMP estimates and varies for the same reasons as the listed population.

These changes in the population estimates do not necessarily indicate a change in actual koala populations, but rather reflect more data being available and a refinement of the modelling approach from previous NKMP estimates. Like any data-driven model, the estimates themselves and associated confidence bounds will change as more data are added.

The NKMP model will include data from the NSW Government, collected as part of efforts to collect baseline data on koalas populations across the state, in the future. The next population update is anticipated to include more NSW data to help improve the confidence bounds of the estimate.

These population estimates will provide a post-2022 listing baseline. The intention of the NKMP is to establish consistent data-driven modelling that will be able to track population trajectories over time.

Over time the NKMP will develop data-driven models of listed population trend to monitor the recovery of the endangered koala. This new survey data and associated population estimate identifies that the baseline population of koala may be higher than previously believed.

The next national koala population estimates are due in mid 2026. The intention is to provide population estimates at the regional level where sufficient data are available. Interim updates may also be released where appropriate.

Estimate (date)	Method	National population range	Listed population range (QLD, NSW, ACT)	Unlisted population range
NKMP 1st annual estimate (2023)2	Data-driven NKMP model	287,830 – 628,010	117,050 – 244,4401	170,780 – 383,570
NKMP 2nd annual estimate (2024)3	Data-driven NKMP model	224,000 – 524,000	95,000 – 238,0001	129,000 – 286,000
NKMP 3rd annual estimate (2025)4	Data-driven NKMP model	729,000 – 918,000	398,000 – 569,0001	303,000 – 381,000

1. Population estimates will be improved and may change, as additional data added over time. Changes in the population estimates year-on-year do not necessarily indicate a change in actual koala populations, but rather a refinement of the modelled estimate.
2. See here for a more detailed description of the NKMP’s approach to a national koala model. (PDF, 610kb).
3. See here for more detailed description of the March 2024 population estimate. (PDF, 1.3Mb).
4. Technical details of that latest population estimate will be available here in the form of a peer-reviewed paper scientific paper when published (expected late 2025 early 2026).

The NKMP progress to date

Strong collaboration

• All relevant jurisdictions engaged in co-design and capability building.
• First Nation, Technical Science and Citizen Science Communities of Practice established to help co-design the program.
• 50 agreements made with states, tertiary institutions and other relevant stakeholders.
• A broader community of koala modellers established and actively working to refine the National approach.

Innovative statistical modelling

• NKMP analytical framework developed, including:
  • analytical approach and software package for bringing together disparate sources of data peer-reviewed and published (Foster et al 2023a)
  • new method for designing large-scale monitoring programs developed and published (Foster et al 2023b).
• The third estimates of koala populations generated using the robust data-driven NKMP model.
• Data handling procedures refined.

On ground survey effort

• A Network of NKMP monitoring sites established and priority monitoring sites selected.
• Koala surveys were conducted at 1148 sites, covering a total transect distance of 2102 km, following the national survey design. Information collected on the location and abundance of 81 non-koala species, including 196 sightings of the endangered greater glider.

Supporting public resources developed

• Two koala sighting apps released: Koala Counter (for collecting survey data) and Koala Spotter (for reporting opportunistic sightings).
• Guidance on koala survey methods developed, tested with key experts and published.
• Integrated and collected koala rescue and habitat data to inform better koala rescue release decisions.

FAQ

What information does the NKMP use?

The NKMP pulls together information on the presence and density of koalas from systematic surveys undertaken throughout their range by CSIRO research teams and by project partners in government, NGOs and universities. We also utilise opportunistic sightings of koalas made by the citizens that are recorded in publicly available databases. All these data on koala abundance are combined with information on the presence of koala habitat trees, soil conditions, rainfall and temperature in an integrated model that produces a prediction of where and how many koalas are likely to occur.

Why has the estimate changed so much in 2025?

The change in the population estimate most likely reflects a combination of increased survey effort, more sensitive survey methods, the inclusion of additional data and model improvements. The new survey methods include drones equipped with thermal imaging capability, and acoustic recordings from which male koala bellows are detected using koala call recognising software.

Does this mean that koalas have increased significantly since the summer bushfires?

Environmental conditions have been more favourable for koalas since the 2019-20 summer bushfires, though whether this has resulted in a substantial population increase is unclear at this stage. It is likely, however, that a large part of the change in the estimated koala population is attributable to increased survey efforts from the NKMP, including more sensitive surveying techniques (for example, acoustic recording and thermal drones). This has provided greater understanding of koala distribution and density, rather than demonstrating an increasing population over time per se.

Geographically, where are there significantly more koalas than previously thought?

Through the more extensive monitoring it has come about that koalas are more continuously distributed through some of the forested areas of the Great Dividing Range than previously thought.

How does the 'model' estimate the number of koalas in areas where we haven’t surveyed?

In areas that haven’t been surveyed, we can’t know for certain whether koalas are present or at what density. We can, however, predict their likely density (if they are indeed present) based on their habitat preferences that have been calibrated from areas where they do occur. Before assigning a likelihood of koala presence in unsurveyed areas, we check for nearby observations of koalas, then use a spatial smoothing function for prediction away from these known populations. Where our surveys are sufficiently close together, this enables the model to essentially “join the dots” indicating occupied koala habitat. We are careful not to extend our predictions beyond what can be justified by the known locations of koalas.

Is my data included?

All citizen-generated data are considered for use in modelling population numbers, though to avoid using duplicate sightings of the same koala, and to correct for spatial bias in observations, a strict process of data thinning, filtering and verification is implemented.