Skip to main content
Dryad logo

Poor quality monitoring data underestimate the impact of Australia’s megafires on a critically endangered songbird


Crates, Ross et al. (2021), Poor quality monitoring data underestimate the impact of Australia’s megafires on a critically endangered songbird, Dryad, Dataset,


Aim: Catastrophic events such as south-eastern Australia’s 2019/20 megafires are predicted to increase in frequency and severity under climate change. Rapid, well-informed conservation prioritisation will become increasingly crucial for minimising biodiversity losses resulting from megafires. However, such assessments are susceptible to bias, because the quality of monitoring data underpinning knowledge of species’ distributions is highly variable and they fail to account for differences in life-history traits such as aggregative breeding. We aimed to assess how impact estimates of the 2019/20 megafires on the critically endangered regent honeyeater Anthochaera phrygia varied according to the quality of available input data and assessment methodology.

Innovation: Using Google Earth Engine Burnt Area Mapping, we estimated the impact of the megafires on the regent honeyeater using six monitoring datasets that differ in quality and temporal span. These datasets are representative of the variable quality of monitoring data available for assessing fire impact on 326 other threatened species; most are poorly monitored and few have standardised, species-specific monitoring programs. We found that assessments based on Area of Occupancy (AOO), Extent of Occurrence (EOO) and public sightings underestimated the fire impact relative to recent, targeted monitoring datasets; a MaxEnt model, sightings from a national monitoring program and nest locations since 2015. Using an impact threshold of 30% of habitat burned, regent honeyeaters would not meet this criteria using estimates derived from EOO, AOO or public sightings, but would exceed the cut-off based on estimates derived from the targeted monitoring data that account for population density.

Main conclusions: To ensure that conservation prioritisation has the greatest capacity to minimise biodiversity losses, we highlight the need to improve targeted, threatened species monitoring. We demonstrate the importance of using recent, standardised monitoring data to estimate accurately the impact of major ecological disturbances, particularly for declining, nomadic species undergoing range contractions.


Study species

The regent honeyeater is endemic to Australia’s eastern seaboard and was abundant and widespread as recently as 60 years ago (Franklin et al. 1989). Extensive land clearing has led to a rapid population decline, with fewer than 350 individuals estimated to persist in the wild today in a range exceeding 600,000 km2 from Victoria to southern Queensland (Crates et al. 2021). Regent honeyeaters nest primarily in association with flowering events in a small number of Eucalyptus tree species, which show very high spatio-temporal variation in flowering phenology (Birtchnell and Gibson 2006; Franklin et al. 1989). Regent honeyeaters have evolved a highly nomadic life-history to track these dynamic nectar resources. Individuals can travel hundreds of kilometres, and typically nest in loose aggregations when flowering conditions allow. Aggregative nesting could help optimise settlement decisions, antipredator defence and the cultural transmission of information amongst conspecifics (Crates et al. 2017a). The entire population represents a single genetic management unit but the core remaining population persists within the greater Blue Mountains area of central / eastern New South Wales (Crates et al. 2019a,b).

Fire severity mapping

We used the Australian Google Earth Engine Burnt Area Map (GEEBAM, Commonwealth of Australia 2020), derived remotely from Sentinel 2 satellite imagery. GEEBAM calculates the difference between pre-fire (April 2018 to April 2019) and post-fire (November 2019 to May 2020) Normalized Burn Ratio using near infrared and shortwave infrared spectral data. Fire severity classes reported in GEEBAM include low (little change), medium (crown unburnt), high (crown partially burnt), very high (crown fully burnt) and unclassified (i.e. non-native vegetation or areas outside of the fire footprint). Further details of the fire severity mapping are available at

Regent honeyeater monitoring datasets

We used six monitoring datasets based on varying degrees of data quality (Table 1). Since 2015, we have used these datasets to establish a standardised, targeted and range-wide monitoring program for the regent honeyeater – the National Regent Honeyeater Monitoring Program (NRHMP). The aim of the NRHMP is to increase the quality and quantity of monitoring data available for the regent honeyeater, with the ultimate goal of informing targeted conservation action to help prevent the species’ extinction in the wild. Developed over the past 6 years through extensive field surveys, habitat modelling and expert elicitation, the NRHMP now surveys over 1300 sites stratified in known or potential breeding areas throughout the species’ contemporary breeding range (Figure S1) during the Austral spring and early summer (Crates et al. 2019). The NRHMP sampling regime aims to account for both the nomadic life-history and the breeding biology of the regent honeyeater (Crates et al. 2017b). The six monitoring datasets based on wild birds are:

1) Extent of Occurrence (EOO): minimum convex polygon of verified regent honeyeater sightings since 1990, sourced from BirdLife Australia and used for IUCN classification.

2) Area of Occupancy (AOO): 1 km x 1 km grids containing a verified regent honeyeater sighting since 1990, sourced from BirdLife Australia.

3) Public sightings: Location of wild regent honeyeaters detected by the general public and reported to BirdLife Australia between January 2015 and December 2019. We refined the database to remove duplicate records of the same individuals in each year. The final database contained 152 point locations for 290 individuals, with potential for duplicate records of the same individuals across different years.

4) MaxEnt species distribution model: A raster model with 1 km2 resolution developed using public sightings of regent honeyeaters between 2000 and 2010, and six month lagged rainfall data (Stojanovic et al. in revision). SDMs were individually run for six-month periods between July and December, or ‘time slices’ (n = 22). The final consensus model flags all cells modelled as suitable breeding habitat in at least 11 of the 22 seasons (i.e., probability of suitability greater than 50%). See Supplementary File S1 for further details on the modelling process.

5) Public and NRHMP sightings: Location of wild regent honeyeaters detected either by the general public or through the NRHMP between 2015 and 2019. We refined the database to remove duplicate records of the same individuals in each year. The final database contained 416 point locations for 899 individuals, with potential for duplicate records of the same individuals across years.

6) Nests: Location of regent honeyeater nesting attempts that reached the egg stage, located either through the NRHMP (n = 138) or by members of the public (n = 6) since August 2015 (Crates et al. 2019a). We also included three nests involving a captive-released regent honeyeater if the partner was of wild origin.

Spatial and statistical analysis

We used ArcGIS Desktop 10.8 (ESRI, 2020a) for all geoprocessing. Prior to spatial analysis, we projected all spatial data to EPSG: 3577 (GDA94 – Australian Albers), ensuring equal area between raster cells. We resampled the GEEBAM data to 40m resolution during projection, then converted from raster to polygon for later use in analysis. We clipped the EOO minimum convex polygon to the coastline and converted each pixel of the MaxEnt raster (i.e., 0 | 1; unsuitable | suitable) into a distinct polygon using a three-step procedure involving: i) raster to point conversion; ii) creation of a fishnet based on the raster dimensions; and iii) conversion of point feature to polygon with raster values, using the raster to point output for labelling the final polygons. We projected all point data for known nests, public sightings and NRHMP sightings from WGS84 to Australian Albers, buffered by 500m radius, and converted from circular polygons to square polygons. We used a 500m buffer to account for foraging and dispersal movements of regent honeyeaters (Geering and French 1998) and created square buffers around point data to ensure consistency and comparability between reporting of impacts based on raster data (1 km2 cells) versus vector distribution data. Because the sightings and nesting databases confer data at the individual level, some of the buffered polygons overlapped. We therefore created a complementary fire impact estimate for the sightings databases by dissolving the boundaries around overlapping buffered sightings locations.

To calculate the proportion of habitat within each dataset impacted by the 2019/20 bushfires, we overlaid each projected vector distribution dataset (EOO, AOO, the MaxEnt model, public sightings, NRHMP plus public sightings, and nests) with the GEEBAM fire severity mapping using a ‘Summarize Within’ function in ArcPro (ESRI, 2020b).

Using R version 3.4.3 (R Core Team 2017), we fitted a logistic regression model via package lme4 v1.1-21 (Bates et al. 2019) to compare the proportion of fire-affected 1km grid cells in each database.

Usage Notes

File 'PPN_burnt.csv' summarises the 1km x 1km grid cells in each moniotoring dataset and whether they were burnt or not (1 = yes 0 = no) and the breakdown of fire severity (low, medium, high & very high) in the fire-affected grid cells. This dataset is used to fit the logistic regression summarised in Table 3 of the manuscript. 

File 'Stacked_barchart.csv' summarizes fire severity in each monitoring dataset according to the proportion of fire-affected grid cells in each.

File 'Regent honeyeater fire impact assessment' provides code to run the logistic regression (Table 3) and plot the fire impact by dataset (i.e. stacked barchart in Figure 2).

Due to the sensitive nature of the raw data (involving breeding locations of a critically endangered species), these data are not available. Please contact the corresponding author ( for further information.

For further information on the spatial analysis, please contact Dr Jason Mackenzie (