A review of 60 years of fire management for threatened fauna and flora at Two Peoples Bay Nature Reserve, Western Australia
Megan Dilly A , Sarah Barrett
A , Sarah Comer A * , Allan H. Burbidge B , Alan Danks C , Judith M. Harvey D † , Angas J. M. Hopkins D † and Graeme T. Smith E †
A
B
C
D
E
† Judith M. Harvey, May 2024. Angas J. M. Hopkins, deceased July 2016. Graeme T. Smith, deceased June 1999. J. M. Harvey, A. J. M. Hopkins, and G. T. Smith were authors of ‘Fire history of Two Peoples Bay Nature Reserve’ and A. J. M. Hopkins and G. T. Smith were authors of ‘Fire: effects and management implications’, which were written for a special bulletin of on the natural history of Two Peoples Bay Nature Reserve.
Handling Editor: Mike Calver
Abstract
Two Peoples Bay Nature Reserve in Western Australia has a long history of ecological studies and adaptive fire management. This provides an excellent opportunity to assess the effects of fire management, including fire exclusion, on the ecosystems and threatened species of an important nature reserve.
To review the fire history of the Reserve and the complexity of managing fire for the conservation of threatened species and communities.
In this paper, we reviewed data from personal consultations, historical records of fire management, analyses of fire regimes, long-term Djimaalup/noisy scrub-bird monitoring, camera-trap surveys, botanical surveys, and quadrat analysis, dating from before to after a large fire in 2015.
Fire sensitive ecosystems at the Reserve are identified. Senescing flora species recruited following the 2015 fire and fire-stimulated species were recorded for the first time. The exclusion of fire was a key factor in the conservation of the Djimaalup/noisy scrub-bird, but has implications for the conservation of other species.
While introduced fire was excluded from the granite headlands for >60 years to conserve fauna habitat, this may not have been an optimal strategy for other conservation dependent fauna, and fire sensitive communities.
The exclusion of fire was an effective management tool for 60 years to conserve fauna habitat, initially driven by conservation of the Djimaalup/noisy scrub-bird, but adaptive management must consider the range of species present as well as a changing climate. Long-term studies and monitoring of threatened species are invaluable to allow informed decisions on adaptive fire management.
Keywords: fire history, fire sensitive communities, habitat management, long-term fire exclusion, noisy scrub-bird Atrichornis clamosus, obligate seeder, strategic fuel reduction.
Introduction
The Southwest Australian Floristic Region (SWAFR sensuGioia and Hopper 2017) is a region with a fire-prone Mediterranean climate. Fire as a recurrent feature and fire regime, including fire interval, intensity and extent, are key drivers of ecological processes (Abbott and Burrows 2003; Miller and Dixon 2014). Planning for fire management is complex because of the need to protect people and infrastructure, ecosystems, and threatened species and is complicated by the nature of fire ecology, the complexities of species’ response to fire, climate change, and logistical challenges (Abbott and Burrows 2003; Burrows 2008; Bowman et al. 2012; Fontaine et al. 2012; Russell-Smith et al. 2020).
Changes to fire regimes across southern Australia include increases in fire size, frequency, and severity, which have resulted in declines in the extent of long unburnt vegetation present in reserves in southern Australia, a significant threat to fire sensitive flora and fauna (Legge et al. 2022; Doherty et al. 2024; Driscoll et al. 2024). Inappropriate fire regimes are a significant threat to biodiversity on the south coast of Western Australia. For at least five threatened bird species of the south coast, change in the fire regime is a significant threat to population viability (Barrett et al. 2009; Department of Parks and Wildlife 2014). For mammals including the quokka (Setonix brachyurus) and Ngilgaitch/Gilbert’s potoroo (Potorous gilberti), fire regime is an important factor with time since fire, scale, and intensity influencing populations and occupancy (Department of Parks and Wildlife 2016; Bain et al. 2023; Friend et al. 2025). Fire exclusion areas are important to retain fauna habitat for mammal and bird species that require dense cover. The responses of heathland fauna to fire regimes can be highly variable (Keith et al. 2002; Clarke 2008; Ward et al. 2020; Povh et al. 2023). This is relevant for a number of endemic species found in the SWAFR. For example, the Djimaalap/noisy scrub-bird requires the complex structure provided by long unburnt vegetation and well-developed leaf litter, while Booderitj/western bristlebird (Dasyornis longirostris) habitat can become unusable if it becomes too tall and dense (A. Danks and S. Comer, unpubl. data).
Inappropriate fire intervals have adverse impacts on plant species that are killed by fire, and which have canopy-stored seed (Keith et al. 2014). Most fire-driven population extinctions of plants are linked to short fire intervals (Barrett et al. 2009). Frequent fire is a significant threat to obligate seeders with long juvenile periods. Less commonly, long fire intervals have been linked to population decline in serotinous species due to senescence of plants and lack of regeneration in the absence of disturbance (Bradstock et al. 1996; Keith 1996; Enright et al. 1998; Burrows and Wardell-Johnson 2003; Yates and Ladd 2005). Additionally, there are adverse effects on populations of species with soil-stored seed when the inter-fire period becomes longer than the viability of the seed (Keith 1996). However, few studies have assessed mortality in long unburnt vegetation and whether this is offset by inter-fire recruitment (Gosper et al. 2013).
Two Peoples Bay Nature Reserve (Fig. 1) occupies 4637 ha on the south coast of Western Australia, east of Albany. The Reserve includes the granite inselberg of Maardjitup Gurlin/Mt Gardner, an isthmus to the west of Maardjitup Gurlin/Mt Gardner, which leads to dunes and wetlands in the lowlands, and a low lateritic plateau in the north-west (Hopkins et al. 2024a; McArthur et al. 2024; Tyler et al. 2024). Two Peoples Bay Nature Reserve is an important conservation area, created in 1966 to conserve the Endangered Djimaalup/noisy scrub-bird (Atrichornis clamosus), which only occurred within the Reserve at that time (Department of Conservation and Land Management 1995; Danks 1997; Chatfield and Saunders 2024). The Reserve also supports populations of other rare species including the Endangered Booderitj/western bristlebird and Dading/western whipbird (Psophodes nigrogularis), the Vulnerable quokka, and the Critically Endangered Ngwayir/western ringtail possum (Pseudocheirus occidentalis). The Critically Endangered Ngilgaitch/Gilbert’s potoroo was rediscovered on Maardjitup Gurlin/Mt Gardner in 1994 (Sinclair et al. 1996; Danks et al. 2011). Where known, Indigenous names for biota and locations are from Abbott (1983, 2001, 2009) and Knapp et al. (2024).
Two Peoples Bay Nature Reserve in a regional context (Reserve boundary shown in red). Where known, Noongar names are given before the European name and are from Knapp et al. (2024).
Little is known about Aboriginal burning practices at Two Peoples Bay, although Aboriginal people used fire as a resource management tool in the Albany region (Hallam 1975). Lullfitz et al. (2021) suggest that regimes involving infrequent burning by Noongars in plant communities on old, climatically buffered, infertile landscapes (OCBIL) (Hopper 2009, 2023; Lullfitz et al. 2021) were conducive to conservation of fire sensitive members of the Proteaceae. This is supported by Knapp et al. (2024) who suggest that the use of fire in Two Peoples Bay was circumspect and that traditional practices would have avoided burning the habitats of rare species on Maardjitup Gurlin/Mt Gardner.
From the early 1900s, Two Peoples Bay increased in popularity as a place for fishing, camping, and picnicking. By the 1920s, people began building rough shacks to enhance their enjoyment in the area (Chatfield and Saunders 2024). These activities would have increased the incidence of fire in the area, but there are no records on the frequency or extent of fires before the 1960s. However, communication with local people contributed to an understanding of attitudes to fire during the first 60 years of the 20th century. Local fishers were reputed to have lit occasional fires to allow easy access to their favourite fishing spots and kangaroo shooters burnt the bush to attract their quarry to the fresh regrowth in a similar way to Aboriginal people before them. Although land clearing for farming purposes began on a small scale in the 1930s, the impact on the Two Peoples Bay area was more significant after World War II with the increase in recreational activity and development (Chatfield and Saunders 2024). An important influence was the pumping station on the Angove River, which required a continuous supply of dry wood to feed the boilers between 1913 and 1953 (Chatfield and Saunders 2024). In forested sections, woodcutters fired the bush to kill trees prior to felling and cutting the dried wood the following year. A local legend tells of a pyromaniac who worked at the pumping station before World War II (D.A.P. West, pers. comm.). In 1939, the voluntary Lower Kalgan Bushfire Brigade was formed and, after maintaining control over fires throughout the war, has remained active to the present day. The brigade relies on local farmers for its membership. In the early days, it was customary to burn small patches of thick scrub before summer, along roads and in uncleared areas, to provide fire breaks and prevent wildfires (E. Poole, pers. comm.).
A local resident recalls that there were fewer fires in the 1950s, but at least one wildfire in that decade occurred near the mouth of Gardner Creek and burnt one of the shacks. In 1958/1959, the brigade was issued with a new fire truck and they carried out a controlled burn near the Gardner Creek crossing (E. Webb, pers. comm.).
Changes to fire regimes throughout the south-west of Western Australia since European colonisation are thought to have been a critical factor in the decline of formerly widespread populations of several threatened vertebrates, including the Djimaalap/noisy scrub-bird, Ngilgaitch/Gilbert’s potoroo, and Booderitj/western bristlebird found at Two Peoples Bay (Smith 1985a; Department of Parks and Wildlife 2014; Burbidge et al. 2025). Fire driven alteration of habitat is considered one of the major factors in the decline of the Djimaalup/noisy scrub-bird. The small areas of habitat remaining unburnt at Two Peoples Bay were the sole refuge for this semi-flightless passerine (Smith 1977; Danks 1997).
The importance of the Maardjitup Gurlin/Mt Gardner Headland as a refuge from fire is well documented (Hopper 2000), with historical aerial photography showing the extensive network of bare granite ridges that act as natural firebreaks. The survival of species that are reliant on vegetation that is long undisturbed, such as the Djimaalup/noisy scrub-bird, provides evidence of the refugial nature of the headland and the long-term burning practices of the Noongar people (Knapp et al. 2024). It is likely that the rocky landscape and wet gullies protected areas of vegetation from fire, prior to active management of the Reserve (Danks et al. 1996). It was as a response to this thinking that the fire management plan drawn up for the Reserve in 1976 emphasised fire exclusion through strategic protection (Hopkins 1985). Between 1966 and 2015, fire management at Two Peoples Bay focused on maintaining core habitat for threatened species with no planned introduction of fire and suppression of unplanned fires (Department of Conservation and Land Management 1995; Barrett et al. 2009; Danks et al. 2011; Woinarski et al. 2015).
While the initial management of fire in the Reserve was driven by conservation management of the Djimaalaup/noisy scrub-bird and later Ngilgaitch/Gilbert’s potoroo, the role of fire in shaping habitat quality was only considered from a faunal perspective. Djimaalup/noisy scrub-birds were known to require long-unburnt habitat, with this connection to cover and the development of a well-developed leaf-litter invertebrate fauna well documented (Smith 1985a, 1985b; Danks 1997). For Ngilgaitch/Gilbert’s potoroo, significant habitat has been described as long unburnt habitat (>30 years), which allows for optimal vegetation density and well-developed hypogeal fungi (Department of Parks and Wildlife 2016; Friend et al. 2025; Syme et al. 2025).
Populations of two other species of threatened birds, Booderitj/western bristlebird and Dading/western whipbird also have conservation significance and have specific fire management requirements at Two Peoples Bay (Burbidge et al. 2025). Two Peoples Bay also has a significant population of quokka and combined with animals in the Waychinicup–Yilberup/Mt Manypeaks area, is considered to be a weakly differentiated genetic cluster of the southern forest population (Spencer et al. 2019). The quokka is considered fire sensitive, with slow recovery following fire in forest ecosystems (Department of Environment and Conservation 2013; Bain et al. 2023).
An awareness of the need to develop a mosaic of fire ages across the broader Two Peoples Bay–Yilberup/Mt Manypeaks area was identified in the early 2000s, and while the commencement of a mosaic burning approach to management of Yilberup/Mt Manypeaks was established in 2015, this has not commenced on the Maardjitup Gurlin/Mt Gardner Headland (Comer and Burbidge 2006).
An appreciation of the fire history and management of fire in Two Peoples Bay Nature Reserve is vital for understanding and guiding future management of its plant and animal communities.
Aims
The fire history of Two Peoples Bay Nature Reserve includes areas that have been subject to wildfires and prescribed burning as well as a large area managed for fire exclusion, which was then subject to a large and intense wildfire in 2015 that burnt approximately 70% of the headland. A controlled experiment was not possible as there were no analogous sites where fire exclusion was not a management priority. However, the fire history, coupled with a history of ecological studies in the Reserve, provides research opportunities to inform adaptive fire management. In this paper, we document the fire history of the Two Peoples Bay Nature Reserve, both prior to and after the creation of the Reserve, and appraise the effectiveness and conservation outcomes of fire management and exclusion including:
Was the exclusion of fire instrumental in the conservation of the Djimaalup/noisy scrub-bird and its habitat including leaf-litter invertebrates?
What were the impacts of this fire exclusion for other threatened fauna including Ngilgaitch/Gilbert’s potoroo, quokka, Booderitj/western bristlebirds and Dading/western whipbirds?
Did historical prescribed burning and fuel reduction in the strategic fuel reduction zone provide protection for the Maardjitup Gurlin/Mt Gardner headland from fires from the west?
Did increased on-site management, early detection and suppression capacity at Two Peoples Bay Nature Reserve contribute to fire exclusion from the headland?
How did fire exclusion impact flora on the Maardjitup Gurlin/Mt Gardner headland, and is there any evidence of senescence of non-sprouting canopy-stored seed species?
Materials and methods
History of fire
The fire history of Two Peoples Bay Nature Reserve since European settlement was compiled from Department of Conservation and Land Management records and the Department of Biodiversity, Conservation and Attractions (DBCA) GIS fire history database (accessed January and February 2024). These records include delineation of fire scars on aerial photography and the personal records and accounts of local people and agency staff responsible for managing the Reserve (see Supplementary material Table S1). In addition, on the earliest available aerial photography of the area, which was taken in March 1946 by the Department of Defence for mapping purposes, patterns of small mosaic burning can be seen. This photography shows the scars of numerous small fires along the coast and between the granite outcrops throughout Maardjitup Gurlin/Gardner headland. Larger fire scars are evident in the isthmus and western boundary areas (Fig. 2). This pre-1960 fire history is considered in the context of Djimaalup/noisy scrub-bird habitat.
Historical aerial photography of the Maardjitup Gurlin/Mt Gardner headland from (a) 1946 to (b) 1966 (Images from Department of Defence and CSIRO archives).
An assessment of plant biomass (all vegetative matter) was undertaken in the 1970s within the strategic fuel reduction zone, which is a strip of land approximately 0.6 km wide and 2.5 km long (~150 ha) located in the narrowest part of the isthmus that separates Maardjitup Gurlin/Mt Gardner headland from the rest of the Reserve. The assessment involved measuring biomass before and after a prescribed burn in 1976. The biomass was then re-measured six times within the first 10 years after fire. Vegetation was sampled by harvesting five 4 m2 quadrats within a 625-m2 plot, fenced to exclude western grey kangaroos (Macropus fuliginosus). Data on effects of grazing are from five 2 m2 quadrats outside the fenced exclosure.
The vegetation communities of Two Peoples Bay Nature Reserve (Hopkins et al. 2024b) were assessed using the presence/absence of dominant plant species and landforms to determine those identified as fire-sensitive communities on the South Coast (appendix IV – Barrett et al. 2009). Fire history (Table S1) was examined, and fire age mapping was overlain on the locations where fire sensitive communities occur to determine the fire intervals to which these communities have been exposed.
In April 2024, H. elliptica, which is a non-sprouting dominant species with canopy-stored seed in which senescence had been observed, was assessed in long unburnt vegetation (minimum of 51 years since fire) on the Maardjitup Gurlin/Mt Gardner headland. H. elliptica was selected as it has low susceptibility to Phytophthora cinnamomi (G. Freebury, pers. comm.) and plant death or decline is unlikely to be due to this pathogen. Three transects (200 m in length) were walked across the width of the H. elliptica population and all individuals recorded along the transect were assessed. H. elliptica individuals (n = 159) were assessed as live or dead, with the live plants further divided into three size classes (<0.5 m, 0.5–1.5 m, >1.5 m). All live plants were assessed for plant health by estimating the percentage of the plant that was dead, to the nearest 10%. Based on observations, dead plants are likely to be identifiable up to at least 5 years after death, but mortality may have been underestimated.
Hakea elliptica regeneration after fire
H. elliptica regeneration was assessed a minimum of three times in nine 10 m × 10 m quadrats in H. elliptica fringing shrublands on the Maardjitup Gurlin/Mt Gardner headland after the 2015 fire. A visual estimate of canopy cover of H. elliptica for the entire quadrat was recorded (to the nearest 5%). Quadrats were monitored once at the beginning of winter (May/June 2016, 6 months post-fire), once in spring (September/October 2016, 10 months post-fire), and once at the start of autumn (March 2017, 16 months post-fire).
Fire response of flora
Floristic composition was documented in 27 10 m × 10 m quadrats on the Maardjitup Gurlin/Mt Gardner headland. Five of these quadrats were established in the spring prior to the November 2015 fire and had pre-fire floristic data. Immediately after the 2015 fire, an additional 22 quadrats were established in the areas of previously long unburnt vegetation that burnt in the fire (Fig. 3). The 27 quadrats were stratified across three habitats: (1) shrublands on granite outcrops; (2) shrublands/thickets on the slopes below the granite outcrops; and (3) heath/sedgelands on plains in deep sands, with nine quadrats in each of the three habitats.
Recent fire history of Two Peoples Bay Nature Reserve (Reserve outline in red) from DBCA fire mapping (in black with fire year in black), including the location of the ‘Strategic Fuel Reduction Zone’ (in green) and the vegetation quadrats (yellow squares).
In addition, opportunistic records of species outside of quadrats were documented. Through the use of published information sources (Supplementary material file S1) and field observations, the species were classified according to their fire response (Miller and Dixon 2014), into the following categories:
Annual;
Fire-ephemeral annual (also known as monocarpic fire ephemerals);
Fire-ephemeral perennial (also known as polycarpic fire ephemerals);
Non-sprouter: non-persistent seed;
Non-sprouter (also referred to as obligate seeders): soil stored seed;
Non-sprouter: canopy stored seed;
Resprouter: non-persistent seed;
Resprouter: soil stored seed; and
Resprouter: canopy stored seed.
Fire response of fauna
Monitoring of the Djimaalup/noisy scrub-bird population involved recording of territorial males as an index of population trends between 1965 and 2024 (Smith and Forrester 1981; Danks 1997; Roberts et al. 2020; S. Comer, unpubl. data). For Maardjitup Gurlin/Mt Gardner, this was completed annually. In this paper, we provide additional data on the response of Djimaalup/noisy scrub-birds to fire from pre-2015 fire to 2024. Monitoring of leaf-litter invertebrates in occupied Djimaalup/noisy scrub-bird territories provides an indication of available food resources. The invertebrate taxa preferred by Djimaalup/noisy scrub-birds were identified by Danks and Calver (1993), and Welbon (1993) with ants, crickets, cockroaches, larvae and spiders, and beetles most frequently consumed. Documenting the availability of leaf litter invertebrates before and after the 2015 fire followed the methods established by Danks and Calver (1993), with wet pit traps collecting samples from core areas of scrub-bird territories. For this paper, leaf-litter invertebrate data from four Djimaalup/noisy scrub-bird territories on Maardjitup Gurlin/Mt Gardner headland are presented, with these sites all sampled at six intervals between 1994 and 2023. These territories were burnt in 2015, and data used to investigate changes in the abundance of Djimaalup/noisy scrub-bird preferred invertebrates following the 2015 fire.
Continuous camera trap monitoring, established in 2014 to monitor feral cat (Felis catus) and red fox (Vulpes vulpes) activity and response to management (Comer et al. 2020), resulted in a significant set of quokka images. With these data straddling the 2015 fires, they provide an opportunity to use simple abundance measures to track quokka activity and occupancy pre- and post-fire (O’Connell and Bailey 2011). While activity indices are often biased by spatial variability and detectability (O’Brien 2011), they are useful in understanding quokka activity in response to fire. A monthly activity index for the 15 cameras situated on the Maardjitup Gurlin/Mt Gardner headland was derived as the number of independent events per month per 100 trap nights. Trends in naïve occupancy, or changes in the proportion of the 15 camera sites occupied in each month quokkas were detected, were also calculated to look at changes in the amount of habitat occupied on Maardjitup Gurlin/Mt Gardner headland following the 2015 fires.
The responses of other threatened fauna are discussed briefly, with details captured in separate papers for Dading/western whipbirds, Booderitj/western bristlebirds, and Ngilgaitch/potoroos (Burbidge et al. 2025; Friend et al. 2025).
Results
Fire history and management
In the early period after European settlement, fires tended to be small and numerous on the Maardjitup Gurlin/Mt Gardner headland and more extensive in other parts of the Reserve, where fires entered the Reserve from adjacent farmland or settlements (Table S1). There were few fires in the 1950s, but a high incidence of large, hot fires prior to the establishment of the Reserve in 1966 and for a few years afterwards.
The first management plan for Two Peoples Bay Nature Reserve (Burbidge and Evans 1970) acknowledged the importance of excluding fire from scrub-bird habitat. This plan led to implementing strategies aimed at reducing the risk of fire, including restricting visitors to day-use only, the establishment of water tanks on the Maardjitup Gurlin/Mt Gardner headland, and the construction of strategic firebreaks. Also significant was the employment of an on-site reserve officer (Danks et al. 2024).
In 1975, the Department of Fisheries and Wildlife (the predecessor of DBCA) established a strategic fuel reduction zone (Fig. 3) across the narrowest part of the isthmus to isolate the Djimaalup/noisy scrub-bird population on the Maardjitup Gurlin/Mt Gardner headland from sources of fire to the west. This work was completed in 1976. Later, it served to isolate scrub-bird territories that established around the Lakes in the early 1980s from any fires on Maardjitup Gurlin/Mt Gardner. The fuel reduction zone initially consisted of 12 manageable blocks 100–200 m wide (Fig. 3), Sinker Reef Road, and firebreaks. The plan was to burn two blocks each year for the first 6 years and then review the situation. Eventually the fuel reduction zone had a total width of about 400–600 m and was maintained through prescribed burning carried out by Fisheries and Wildlife staff from Woodvale Wildlife Research Centre. The formation of the Department of Conservation and Land Management (CALM) in 1985 meant that there was now access to extensive fire-fighting expertise and resources, including heavy duty trucks, training, and spotter aircraft. This effectiveness was demonstrated in the 1988 and 1989 fires, where hand crews were deployed from other CALM districts to support ground-based fire suppression. However, it is worth noting that fire exclusion from the Maardjitup Gurlin/Mt Gardner area was an exception to the established CALM fire policies of the time.
From the mid-1980s, the fuel reduction zone was dominated by peppermints (Agonis flexuosa), because of selective grazing by western grey kangaroos. Electric fences were used to exclude kangaroos, but this was ineffective and abandoned. Consequently, grazing pressure on the flats resulted in the loss of leaf litter, annuals, and low shrubs and prescribed fire could no longer be applied during the cooler months. This led to the use of mechanical slashing of the buffer to maintain low fuel loads, which commenced in the early 1990s (Danks et al. 2024).
Other fire management measures on the Maardjitup Gurlin/Mt Gardner headland included maintaining the water points on the headland and the maintenance and slashing of access tracks. Between 1996 and 2014, a number of wildfires occurred in the Reserve west of the fuel reduction zone, including one north of Kaiupgidjukgidj/Moates Lake (1996) that burnt 242 ha, and one in 1999 that burnt in a similar area to the 1996 fire (144 ha) (Table S1, Fig. 3). Approximately 1300 ha of the central part of the Reserve burnt in 2000 in a large wildfire that entered the Reserve from the adjacent Angove Water Reserve, but none of the Djimaalup/noisy scrub-bird territories occupied in 1999 were impacted. The fuel reduction zone assisted in preventing this fire from heading east to the Maardjitup Gurlin/Mt Gardner headland by the low levels of fuel in this zone. In 2012, the Black Cat Creek fire burnt more than 1000 ha of the western part of the Reserve. The fuel reduction zone was identified as a strategic fall-back line to support the management of this fire.
A lightning strike ignited a fire on the northern slopes of the Maardjitup Gurlin/Mt Gardner headland in February 2006. It is notable in that this was the first time that aerial suppression with fixed-wing water bombers was used at Two Peoples Bay. The combination of aerial water bombing and suppression by fire crews on the ground resulted in this fire being contained to only 1.07 ha adjacent to Bishops Gully. A further two lightning strikes are known to have ignited fires on the Maardjitup Gurlin/Mt Gardner headland, one in the summer of 1998–1999 and one in 2014. The 1998–1999 fire on Tick Flat was detected early in the summer of 1999, and had self-extinguished and burnt an area of less than 5 m2. The 2014 fire was started from a lightning strike in Gardner Gully, but with rapid detection and mild weather conditions, this fire was quickly extinguished by ground crews, burning around 1 ha each. The majority of the headland remained long unburnt until a lightning strike in 2015 ignited a large fire that burnt 1138 ha of the headland. The firebreak network was used to defend the headland, and the low fuel reduction zone provided an opportunity to backburn and contain the western edge of the fire. Aerial suppression also contributed to efforts to contain this fire, with fixed wing water bombers deployed. The fire-fighting efforts resulted in a patch of around 30% on the headland that remained unburnt. Additionally, a number of smaller patches of between 1 and 14 ha on some of the granites and wet gullies did not burn.
The extent of different fire ages across the Reserve can be summarised by separating the Reserve into three areas based on the different fire management approaches for each area: Maardjitup Gurlin/Mt Gardner headland, the Strategic Fuel Reduction Zone and the rest of the Reserve. The percentage of each area that burnt in each year (since 1978 when detailed fire mapping became available) is detailed in Fig. 4. The extent and location of vegetation at less than 10 years since fire, between 10 and 30 years since fire, and greater than 30 years since fire is in Fig. 5.
Percentage of each area that burnt in each year from 1978 to 2024 in Maardjitup Gurlin/Mt Gardner headland, the Strategic Fuel Reduction Zone and the rest of Reserve. Actual area burnt (ha) are above each column.
Fire ages at Two Peoples Bay Nature Reserve, <10 years since fire, 10–30 years since fire, and >30 years since fire in (a) 2004, (b) 2014, and (c) 2024 (blank areas are non-flammable zones such as lakes and sand-dunes).
A tabulated history of fires at Two Peoples Bay Nature Reserve is in Table S1.
Vegetation biomass and impact of grazing in the fuel reduction zone
The fire in September 1976 in the fuel reduction zone reduced the above ground biomass at the Isthmus mixed dense low heath from 15.5 to 3.2 tonnes per ha (oven dry weight). This remaining biomass consisted entirely of woody material, which was too thick to burn under the prevailing conditions; the woody material was generally greater than 0.5 cm in diameter.
In the absence of grazing by kangaroos, the aboveground biomass of the vegetation returned to pre-fire levels within about 7 years (Fig. 6). After 10 years, non-grazed vegetation weighed 17.4 tonnes per ha and the rate of accumulation appeared to be tailing off. Grazing, however, had a pronounced effect on the accumulation of biomass. At Year 5, the grazed vegetation was only half the weight of ungrazed vegetation. At Year 10 this proportion was 74%. It was likely that the intensity of kangaroo grazing at the site at this time was above average for post-fire situations because of the small size of the area burnt (12 ha) and because, at the time, there was no kangaroo culling on adjacent private property. Extremely heavy grazing pressures were observed in all the blocks burnt in the Strategic Fuel Reduction Zone (Hopkins 1985), with 50–60 kangaroos concentrated in areas of 10–15 ha at sunset (A. J. M. Hopkins, pers. obs.).
Fire sensitive communities
The assessment identified vegetation communities present at Two Peoples Bay Nature Reserve that are fire sensitive:
Swamp margin thickets and forest around the lakes and in low-lying areas. These communities include a dominant serotinous seeder (Taxandria juniperina) vulnerable to frequent fire with medium to slow post-fire recovery and flowering. Additionally, intense fire removes organic soil and can impact reed and rush nesting habitat. Historically these areas burnt less as they are wetter and less flammable. However, sections around Kaiupgidjukgidj/Moates Lake burnt in the 2012 Black Cat Creek fire.
Vegetation communities that contain dominant serotinous seeders with medium to slow post-fire recovery and flowering may be vulnerable to both frequent fire and infrequent fire (i.e. Melaleuca cuticularis, Banksia seminuda, Banksia coccinea, Banksia occidentalis, Banksia verticillata, H. elliptica, Allocasuarina lehmanniana). Most of these communities have only burnt once in the past 50 years, such as in the 2015 Maardjitup Gurlin/Mt Gardner fire or the 2012 Black Cat Creek fire. However, there are woodland areas containing Banksia coccinea (which has also declined due to the fungal pathogen P. cinnamomi) north of Kaiupgidjukgidj/Moates Lake that have burnt up to three times between 1992 and 2016.
Maardjitup Gurlin/Mt Gardner Granites, Thickets and Gullies includes shrub, herb and moss species in shallow pockets on granite outcrops, dense thickets fringing the outcrops, and forests or thickets in the gullies. These communities are refugial habitats, with short range endemic invertebrates, threatened species, and a high seeder/sprouter ratio (including serotinous seeders such as H. elliptica and B. verticillata) vulnerable to inappropriate fire regimes. These areas were long unburnt and managed as a fire exclusion area until the wildfire in 2015.
Hakea elliptica senescence in long unburnt vegetation
Of the 159 H. elliptica plants monitored in long unburnt vegetation, 22% were dead, most of which were from Size Class 3. Only one was Size Class 2. Of the living plants the majority were Size Class 3, 10% were Size Class 1, and 28% were Size Class 2 (Fig. 7). Limb death was highest in the oldest plants and lowest in the two young age classes (Fig. 8).
Hakea elliptica regeneration after fire
H. elliptica was recorded in the nine quadrats where it was present pre-fire, with the canopy cover of this species at 7 years after fire differing between quadrats, ranging between 5 and 60%. Two of the quadrats were monitored prior to the 2015 burn and senescing H. elliptica individuals were recorded within the quadrats. At 7 years post-fire, the visual estimate of canopy cover of H. elliptica in both of these quadrats increased from pre-fire (from >2% to 5%, and from >30% to 60%).
Fire response of flora
Plant fire responses are in Supplementary material file S1, which also includes a list of the species recorded in the quadrats after the 2015 fire at the Maardjitup Gurlin/Mt Gardner headland and their fire response classification. The proportions of species with each fire response is in Fig. 9. Being able to re-sprout and having soil-stored seed was the dominant fire response for plants in all three habitat types. The granite outcrops had greater numbers of annuals and fewer resprouters (with canopy-stored seed) than the other two habitat types. Canopy-stored seed (resprouters and non-sprouters) made up a small proportion of the species in all three habitats.
Proportions of each fire response for species recorded within the quadrats in each habitat (slopes, deep sands, granite outcrops) at Two Peoples Bay Nature Reserve.
Plant species richness was high in the first spring following the fire because of the presence of annuals and fire ephemerals. Five of the annuals were species that are abundant after fire. Fire ephemerals and perennials were recorded across all habitats. All the habitats were dominated by resprouters with soil stored seed. The resprouters also produced seed (i.e. recorded as both resprouting and seedling), particularly within the Myrtaceae and Proteaceae families. Most facultative re-seeders produced lower numbers of seedlings than the non-sprouters, with the exception of Taxandria marginata and Dodonaea ceratocarpa on the granite outcrops, which produced large numbers of seedlings during the first winter after the fire, many of which died in the first summer.
Non-sprouters with soil stored seed was the second most common fire response strategy in all habitats. Proportions of non-sprouters with canopy stored seed were low in all habitats. They were highest on the granites and lowest in the deep sands.
Twelve woody species recorded after fire were not recorded in the quadrats pre-fire. These included canopy-stored seed species and species with soil-stored seed.
Fire response of fauna
Fire impacts on Djimaalup/noisy scrub-bird habitat were minimal in the 50 years post-rediscovery, and the population grew with the policy of fire exclusion from the Maardjitup Gurlin/Mt Gardner headland. In 1966, the year of the first scrub-bird census, all birds recorded were in areas that had not been burnt in the 1940s (Fig. 10). On Maardjitup Gurlin/Mt Gardner, which remained unburnt over this period, the consequent accumulation of litter and litter invertebrates provided additional Djimaalup/noisy scrub-bird habitat.
(a) 1946 orthophoto showing Djimaalap/noisy scrub-bird territories occupied in 1966, (b) the absence of fire and territories occupied in 1994, and territories occupied following the 2015 fires in (c) 2016 and (d) 2024 (c and d with March 2016 orthophoto showing area burnt).
By 1994, when the population peaked on Maardjitup Gurlin/Mt Gardner, there had been no large fires impacting scrub-bird habitat for at least 30 years, and in most cases longer. A decline in the Djimaalup/noisy scrub-bird population on Maardjitup Gurlin/Mt Gardner commenced in the mid-1990s and by the time the 2015 fire impacted the headland the population index was only 78 (Fig. 11) (Roberts et al. 2020). Immediately post-fire, there was evidence of a small number of males persisting in unburnt pockets, and some seeking refuge in the Lakes sub-area (Fig. 10c).
Djimaalup/noisy scrub-bird population trends, estimated by the number of territorial males recorded each year, in the two sub-areas of Two Peoples Bay Nature Reserve.
The numbers of territorial male scrub-birds recorded post-fire on the Maardjitup Gurlin/Mt Gardner headland in 2016 and 2017 were 22 and 21, respectively, with 15 birds recorded in 2018 (Fig. 11). Around the Lakes, 12 territorial birds were recorded in 2016, 21 in 2017, and 17 in 2018. In 2019 (4 years post-2015 fire), 44 territorial males were recorded on the Maardjitup Gurlin/Mt Gardner headland, and five around the Lakes, but the following year only nine male scrub-birds were singing on the headland and one around the Lakes (Fig. 11). Numbers remain low, and in 2024 only eight males were defending territories on the Maardjitup Gurlin/Mt Gardner headland. No birds have been heard around the Lakes since 2021 (Fig. 11).
Abundance of leaf litter invertebrates on the four sites declined prior to the 2015 fires. Pooled samples for the four sites on the Maardjitup Gurlin/Mt Gardner area contained over 12,000 individual invertebrates in 1994, and 7000 in 1995 (Fig. 12). By 2005 numbers of individuals in the six groups had decreased to 2000, and these low levels of abundance were evident pre- and post-2015 fires (Fig. 12). An increase in the abundance of ants was observed in 2017, 18 months after the 2015 fire (Fig. 12).
Raw abundance of Djimaalup/noisy scrub-bird preferred invertebrates from four scrub-bird territories on Maardjitup Gurlin/Mt Gardner headland sampled between 1994 and 2023.
Quokka activity on camera traps on the Maardjitup Gurlin/Mt Gardner headland provides an index of activity and naïve occupancy from August 2014 through to November 2022. A period of low activity and detection occurred in the 4 months following the fire, with an increasing trend to pre-fire activity within 5 months. Seven years post-fire, quokka activity levels had increased to levels observed pre-fire (Fig. 13). Naïve occupancy also decreased following the fire with quokka detected on fewer than a quarter of cameras, but within 4 months quokka were detected on 40% of cameras (Fig. 13). Two years post-fire, naïve occupancy was comparable with pre-fire data, with quokka regularly detected on over 50% of cameras (Fig. 13).
Monthly quokka camera trap success (±s.e.) and naïve occupancy for Maardjitup Gurlin/Mt Gardner headland between August 2014 and November 2022.
In the mid-1980s, Booderitj/western bristlebirds were resident in the low-fuel buffer on the isthmus, with breeding activity recorded in this area. Since the fuel reduction zone maintenance commenced, the vegetation change due to slashing has resulted in the loss of bristlebird home ranges in this area. In the months immediately following the 2015 fire, Booderitj/western bristlebirds were recorded within the burn area on the Maardjitup Gurlin/Mt Gardner headland, which was unexpected given the extent and intensity of impact on the birds’ habitat. Dading/western whipbirds were rarely detected in the burnt area, but by 2019 numbers were increasing. The fire response of these two threatened birds is discussed in more detail in Burbidge et al. (2025).
An estimated 90% of Ngilgaitch/Gilbert’s potoroo habitat was impacted in the 2015 fire, and post-fire surveys failed to detect animals on the Maardjitup Gurlin/Mt Gardner headland (Friend et al. 2025). Following an assessment of vegetation recovery and intensive introduced predator management, six potoroos were reintroduced to the headland in 2022. Expectations were that the small, long-unburnt patch of habitat would be the main area occupied, but the translocated animals used the 7-year-old regrowth extensively. The impact of the 2015 fire on potoroo is discussed in more detail in Friend et al. (2025).
Discussion
Fire history and management
The fuel reduction zone on the isthmus and the implementation of rapid suppression measures for fires on the headland meant that the Maardjitup Gurlin/Mt Gardner headland was unburnt for 60 years and allowed populations of scrub-birds and other fire sensitive species to increase. Habitat remained suitable for the potoroo. However, fuel accumulation means future fires on the Maardjitup Gurlin/Mt Gardner headland could burn larger areas and be harder to control. Additionally, fire suppression is complicated in an area such as the headland, which is steep with limited access (Danks et al. 1996). Consequently, despite a considerable amount of effort being put into suppression, the 2015 fire burnt over 70% of the Maardjitup Gurlin/Mt Gardner headland.
Vegetation biomass and impact of grazing in the fuel reduction zone
The biomass measurements before and after fires in the 1970s in the fuel reduction zone show how historical fuel reduction burning within a buffer zone worked as a fire management tool at Two Peoples Bay. To assess the likely effect of fuel reduction, we assumed that the biomass remaining unburnt after the 1976 fire was unavailable fuel and taking a value of 8 tonnes per ha available fuel as a limit above where fires might become uncontrollable (this is the value used in the jarrah Eucalyptus marginata forest; Underwood et al. 1985). Broad scale fuel reduction burning in this kind of vegetation would thus have been effective for no more than 6 years, although intensive grazing by kangaroos extended the period of effectiveness in this small-scale study. Viewed in terms of biomass, the vegetation in the zone recovered after fire, returning to pre-fire levels within 10 years. Thereafter, the rate of accumulation declined as growth slowed and there was an equilibration between litter accumulation and decay.
While this provided the basis for planning fuel reduction burns in the 1970s to prevent fire incursion to the Maardjitup Gurlin/Mt Gardner headland, the assumptions on fuel levels and fire behaviour may not apply to present conditions because of climate change. In the late 1980s and early 1990s, fuel reduction measures in the zone were achieved by vegetation slashing rather than burning (Danks et al. 2024). From the early 1990s, it became apparent that management of vegetation in the fuel reduction zone had changed the vegetation community and structure to the point where it no longer provided suitable habitat for the Booderitj/western bristlebird, highlighting the trade-off between management of a low-fuel buffer and habitat for threatened species.
Departmental fire management practices and the presence of resident staff, which meant earlier suppression, reduced the incidence of wildfires on the Reserve following its establishment. The fuel reduction zone assisted in protecting the Maardjitup Gurlin/Mt Gardner headland from fire from the west but there is a risk of fire igniting on the headland from lightning. Ignition of fires from lightning strikes occurred at least five times since 1989 (with some undetected). At the time of the 2015 lightning storm, weather conditions, available fuel, and terrain led to a high intensity fire that burnt approximately 70% of the headland. The trend of increasing fires at Two Peoples Bay Nature Reserve aligns with a spatial analysis of the fire history of south coastal Western Australia, which has shown an upward trend in the incidence and area of bushfires (Barrett et al. 2009).
Fire sensitive communities
Fire sensitive vegetation communities occur at Two Peoples Bay, including on the Maardjitup Gurlin/Mt Gardner peninsula, which mostly avoided fire between the early 1960s and 2015. With future higher temperatures and lower rainfalls, these communities may burn more frequently. These also include wetland areas that may be at higher risk of more frequent and intense fires as the drying climate will lower the water table and reduce the ability of these areas to withstand fire.
In the 1980s, changes to the structure and floristics of long unburnt vegetation (older than 30 years) of Two Peoples Bay Nature Reserve could be observed in aerial photographs and in photographs from around the Reserve (Hopkins et al. 2024b). Many of these changes, particularly the increasing height and cover of the vegetation, could be related to the fire regime of the area. An increased dominance of Agonis flexuosa in heath areas in the north-west part of the headland and height and cover of Banksia sessilis on limestone areas was attributed to a long absence of fire (Danks et al. 1996). Senescence of a number of the non-sprouting canopy-stored seed species, including H. elliptica, in the long unburnt vegetation at the Maardjitup Gurlin/Mt Gardner headland had been noted anecdotally. There had been a decline in many of the Banksia species on the headland and isthmus, but this is because of P. cinnamomi dieback rather than fire (Danks et al. 1996; Hart et al. 2024). Since the mid-1990s, the Critically Endangered B. verticillata has become functionally extinct on the headland in long unburnt vegetation (more than 50 years old) (Hopper et al. 2024). The impact of these changes in habitat characteristics on fauna are unknown, and further study is required.
Hakea elliptica senescence in long unburnt vegetation
While senescence and mortality of H. elliptica was recorded in long unburnt vegetation, recruitment of this species at the Maardjitup Gurlin/Mt Gardner headland occurred even in the absence of fire, with the presence of seedlings and a range of size classes indicating that there is some level of inter-fire recruitment. The ability to release and germinate seed in the absence of fire may assist in mitigating the impacts of senescence of this species in long unburnt vegetation. Further studies in older vegetation would be beneficial to show whether the canopy-stored seed bank declines with time and whether seedlings established in the inter-fire period are sufficient for population replacement. Alternatively, H. elliptica may decline further in long unburnt vegetation, with lower densities now occurring in long unburnt vegetation.
Hakea elliptica regeneration following fire
In previously long unburnt habitat on the Maardjitup Gurlin/Mt Gardner headland that burnt in 2015, canopy seed banks of H. elliptica were sufficient for regeneration. Seedling density after fire was lower than that recorded in similar vegetation with shorter fire intervals at Torndirrup National Park (M. Dilly, unpubl. data). However, there was replacement of the species and increased cover after fire. This reflects work on other serotinous species potentially facing a ‘senescence risk’ in long-unburnt vegetation. After long intervals without fire these species also showed reduced seedling recruitment but still with seedling densities far greater than the number of plants present pre-fire vegetation. This indicates these species have resilience to a range of fire intervals (Ne’eman et al. 1999). Other serotinous species in long unburnt populations on the south coast have poor regeneration after fire, such as Banksia brownii at South Sister Nature Reserve and Vancouver Peninsula (S. Barrett, unpubl. data).
Fire response of flora
The post-fire response examined here reflects fire ecology studies in a range of vegetation communities in south-west Australia that have shown high plant species richness immediately following fire because of the germination of annuals and ephemerals (Yates et al. 2003; Burrows 2008; Gosper et al. 2012). Many fire ephemerals had not been recorded in floristic surveys at Two Peoples Bay and some were extensions of the species range. This included Hydrocotyle serendipita, Trachymene coerulea, and Kennedia glabrata. Within the previously known range of T. coerulea on the west coast of Western Australia this species varies from being a fire ephemeral on the mainland to an annual in its island habitat where fires are infrequent (Rye 1999). K. glabrata, which is a rare species that germinates after fire and had not been recorded in the area, was recorded on the granites after fire. H. serendipita was a previously undescribed fire-ephemeral species recorded for the first time in 2015 at Two Peoples Bay Nature Reserve and Torndirrup National Park. It has a restricted distribution, associated with fire and has long-lived seeds (Perkins and Dilly 2017).
In addition to the annuals and perennials appearing following fire, there were also a number of woody perennials that appeared after fire in quadrats in which they had not been recorded pre-fire. In some cases, this could be a result of seed dispersal, but there were a number of species with soil-stored seed and limited dispersibility that were consistently absent pre-fire and appeared following the burn. It is likely these plants had declined in the above ground vegetation pre-fire. In long unburnt vegetation, the species richness of the above-ground vegetation declines with time since fire, but this does not mean that species have been lost from the system; rather, these species remain in the soil seed bank and germinate following fire (Keith 1996; Burrows and Wardell-Johnson 2003; Gosper et al. 2012).
There are large numbers of long-lived seeds within the Australia flora, with bradysporous species being particularly long-lived (Merritt et al. 2014). Fire ephemeral and perennial species appeared after the 2015 fire at the Maardjitup Gurlin/Mt Gardner headland, which had not been recorded in previous surveys (including an undescribed species, H. serendipita:Perkins and Dilly (2017)), indicating that the seed of these species have long viability, as the areas where they germinated had gone at least 80 years without a fire.
A post-fire study in eastern Australia showed few resprouter species occurred on rock outcrops (Benwell 2007). Another study on the New England Batholith of eastern Australia showed that while non-sprouters were more common on rock outcrops, the resprouters were more common than non-sprouters on the outcrops and in the surrounding matrix (Hunter 2003). This was replicated at Two Peoples Bay after the 2015 fire, which showed almost half of the plant species on the granites post-fire were resprouters. However, the number of resprouters was less on the granites than in the deep sands or on the slopes. The proportion of resprouting species on these granite outcrops were higher than recorded on granite outcrops in the wheatbelt of Western Australia, on which 25% of species were recorded as resprouters, including geophytes (Yates et al. 2003).
Granite outcrops such as the Maardjitup Gurlin/Mt Gardner headland on the south coast may provide an environment that favours seedling survival more than adult survival when compared to surrounding habitats. These outcrops have high variability in water availability and granite plants are exposed to high water stresses (Yates and Ladd 2004), which favours non-sprouters (Pausas et al. 2016). Additionally, these habitats may be subject to a lower frequency of fires than the surrounding landscape and act as refugia for species vulnerable to frequent fire (Hopper et al. 1997, Hopper 2000, Clarke 2002).
Many of the slope quadrats were dominated by non-sprouters with canopy stored seed, with species such as H. elliptica forming thickets on the water-gaining slopes below the granites. Non-sprouters (obligate seeders) often form thickets (Burrows et al. 2008) and have an advantage in thickets over resprouters, which may be less likely to colonise because of reduced recruitment potential (Bond and Midgley 2001).
Fire response of fauna
The challenges of managing habitat for multiple co-occurring species with different fire responses can be problematic. This is particularly relevant when compared with management of invasive species that are more adaptable to different fire regimes and the opportunities these provide (Hradsky et al. 2017; Doherty et al. 2022). For example, Booderitj/western bristlebirds can re-occupy an area after fire sooner than Djimaalup/noisy scrub-birds. Once the vegetation gets higher and denser, it is not suitable for Booderitj/western bristlebirds, but Djimaalup/noisy scrub-birds can recolonise the taller vegetation. This means that a fire regime to suit Booderitj/western bristlebirds is different from the regime required for Djimaalup/noisy scrub-birds, so there will be a trade-off in attempts to manage both species in the same area. This is evident on Yilberup/Mt Manypeaks where Booderitj/western bristlebirds were found along the ridge in the early 1980s following bushfire in 1979, with this habitat now dominated by Hakea cucullata and H. elliptica and no longer suitable for Booderitj/western bristlebirds, but with numerous Djimaalup/noisy scrub-bird territories (A. Danks and S. Comer, unpubl. data). At Two Peoples Bay, Booderitj/western bristlebirds are not found in areas of dense H. elliptica but occupy Anarthria sedgeland and low heath. These areas are contained within the mixed dense low heath vegetation communities described by Hopkins et al. (2024b) and remain structurally suitable for Booderitj/western bristlebirds in the absence of fire for over 60 years (Burbidge et al. 2025).
In the absence of fire on the Maardjitup Gurlin/Mt Gardner headland, the Djimaalup/noisy scrub-bird population increased until the mid-1990s, when the population started to decline. This declining trend was established prior to the 2015 bushfires and was likely related to multiple factors including the decline in the abundance of leaf litter invertebrates, a decrease in rainfall, and an increase in feral cat activity following the commencement of fox baiting through DBCA’s Western Shield Program (Comer et al. 2020; Roberts et al. 2020). In 2016, Djimaalup/noisy scrub-birds occupied unburnt habitat on Maardjitup Gurlin/Mt Gardner, including small refuges on the south side of the headland. The Lakes area also had a number of males defending territories. This spatial distribution of territorial males was observed from 2016 to 2018. However, in 2019 the distribution was skewed, with over 44 territorial males on the Maardjitup Gurlin/Mt Gardner and only five detected around the lakes. The increase in territorial calling activity on the headland in 2019 was reflective of birds exploring regenerating habitat and coming from outside of the Reserve. But the fact that this growth was not sustained cannot be explained easily. Possibly the birds could not access sufficient resources to sustain territories, and that increased predation with lack of cover occurred (S. Comer, unpubl. data). The importance of monitoring the response of habitat parameters is supported by the lack of a clear causal factor in the decline of Djimaalup/noisy scrub-birds at Two Peoples Bay pre-2015 fire and warrants further long-term studies (Legge et al. 2018; Roberts et al. 2020).
With the success of the Djimaalup/noisy scrub-bird translocation, the species’ population is both larger and much more widespread than it was in the 1970s when restricted to Two Peoples Bay (Comer et al. 2025). Consequently, the introduction of small fires to the Maardjitup Gurlin/Mt Gardner headland has the potential to ensure that some long unburnt habitat can be retained to ensure the long-term viability of Djimaalup/noisy scrub-birds at Two Peoples Bay.
The recovery of quokka following the 2015 fires was encouraging, and similar to observations on Yilberup/Mt Manypeaks following the 2004 bushfires (Department of Environment and Conservation 2013). It is likely that this recovery was enhanced by the intensive management of introduced predators, but also the connectivity and presence of a number of refugia on the headland and nearby Lakes were instrumental in the recovery of this species post-fire at Two Peoples Bay. These factors are important for the recovery of metapopulations in the southern forest areas (Bain et al. 2023). Further investigation of quokka response to fire on the south coast is recommended, with occupancy modelling and abundance measures standardised to allow for comparability with other quokka metapopulations (Bain et al. 2023).
Studies of the post-fire recovery of Ngilgaitch/Gilbert’s potoroo have found that this species used burnt areas as early as 5 years following fire (Friend et al. 2025). This led to efforts to re-establish the species on Maardjitup Gurlin/Mt Gardner and a re-assessment of its habitat requirements.
Conclusion
The decision to exclude fire from Maardjitup Gurlin/Mt Gardner was made in the early 1970s, based on a rapidly developing understanding of the ecology of the Djimaalup/noisy scrub-bird (Smith and Robinson 1976; Danks et al. 2024). Management decisions in the Reserve were focused on the exclusion of fire from the headland, and improvements in surveillance and rapid suppression had conservation benefits for a suite of threatened and non-threatened native species (Danks et al. 2024).
The fuel reduction zone on the isthmus and the implementation of rapid suppression measures for fires on the headland at Two Peoples Bay meant that the headland was largely unburnt for 60 years, consequential as long unburnt vegetation is a rapidly disappearing resource across Australia. The long unburnt vegetation allowed populations of Djimaalup/noisy scrub-birds and other fire sensitive species to increase, with habitat suitable for the subsequently rediscovered Ngilgaitch/Gilbert’s potoroo (Friend et al. 2025; Burbidge et al. 2025; Comer et al. 2025). This fire management strategy was unusual in that it was a deliberate policy of fire exclusion, but resulted in remarkable conservation outcomes. However, the accumulation of fuel meant that future fires on the headland had the potential to burn large areas, and be difficult to control, having significant negative impacts on Ngilgaitch/Gilbert’s potoroos, Djimaalup/noisy scrub-birds, Booderitj/western bristlebirds, and Dading/western whipbirds, as illustrated in the aftermath of the extensive 2015 fire. Future management will need to find an appropriate balance, mitigating the risk of extensive fire while providing appropriate fire regimes for threatened species.
The increase in temperatures and decrease in rainfall at the Reserve has led to a higher risk of wildfire (Hopkins et al. 2024b) and the risk will increase in the future with a predicted drying and warming climate (van Oldenborgh et al. 2021). Changes to fire regimes combined with existing threats and the impact of climate change increases the risks to Two Peoples Bay Nature Reserve and the conservation significant species that rely on it (Harvey and Enright 2022).
Dedication
This paper is dedicated to the memory of Judith Harvey who was an author on one of the original unpublished fire papers (Fire History of Two Peoples Bay Nature Reserve), as well as an author on two other papers in this special collection of Pacific Conservation Biology. Judith died before seeing this important work updated and published.
Conflicts of interest
Judith Harvey died in 2024, Angas Hopkins died in 2016, and Graeme Smith died in 1999. They had no known conflicts of interest, nor do the living authors of this paper.
Declaration of funding
The monitoring of floristic quadrats was funded by a grant from South Coast Natural Resource Management. Monitoring of Djimaalup/noisy scrub-birds and other fauna mentioned has been funded by the Western Australian Department of Biodiversity, Conservation and Attractions and its predecessor departments, the Commonwealth Government. In 2024 funding from the Australian Government Saving Native Species Program supported Djimaalap/Noisy Scrub-bird census work. South Coast Natural Resource Management have provided funding support for some historical monitoring of scrub-birds as part of their delivery of the Commonwealth Government funding streams, and currently support Gilbert’s potoroo monitoring.
Acknowledgements
Judith Harvey, Angas Hopkins, and Graeme Smith were authors of ‘Fire history of Two Peoples Bay Nature Reserve’ and Angas Hopkins and Graeme Smith were authors of ‘Fire: effects and management implications’, which were written for a special bulletin of CALMScience on the natural history of Two Peoples Bay Nature Reserve. The papers were subject to peer review, revised, and accepted for publication in 1991. However, neither the papers nor the special bulletin were published. In resurrecting the papers collected for the special bulletin, in order to publish them over 30 years later in the special collection of Pacific Conservation Biology, with the three original authors dead, Megan Dilly, Sarah Barrett, Sarah Comer, Allan H. Burbidge, and Alan Danks brought the combined papers up to date. The three original authors would have met criteria for authorship if alive, so they are included as authors. The original authors thanked Ben Baylis, Mike Ellis, Lekkie Hopkins, Linda Mason, David Mitchell, Les Moore, and Chris Robinson for their assistance in the field, and Reserve officers Alan Danks, Graeme Folley, Dick Grayson, and Ron Sokolowski, who were also helpful during the surveys. DBCA’s Albany District and South Coast Regional staff have continued to support pro-active management of fire in Two Peoples Bay Nature Reserve. In particular, we thank Kelly Gillen, Terry Maher, Roger Armstrong, Greg Broomhall, Peter Hartley, Greg Mair, Dave Atkins, Vince Hilder, Greg Freebury, Jeremy Friend, Deon Utber, and Grantly Morton for their ongoing support for fire exclusion on Maardjitup Gurlin/Mt Gardner and their efforts in suppressing the 2015 bushfires, and their ongoing commitment to fire management at Two Peoples Bay. We thank Steve Hopper for input into earlier drafts, and Harry Recher and John Woinarski for reviewing the paper and providing valuable comments. We also thank Denis Saunders for editorial assistance.
References
Abbott I (2001) Aboriginal names of mammal species in south-west Western Australia. CALM Science 3, 433-486.
| Google Scholar |
Abbott I (2009) Aboriginal names of bird species in south-west Western Australia, with suggestions for their adoption into common usage. Conservation Science Western Australia 7, 213-278.
| Google Scholar |
Bain K, Halley M, Wayne A (2023) Severe, large-scale bushfire threatens metapopulation function of quokka (Setonix brachyurus) in south-western Australia. International Journal of Wildland Fire 32, 1175-1186.
| Crossref | Google Scholar |
Benwell A (2007) Response of rock-outcrop and fringing vegetation to disturbance by fire and drought. Australian Journal of Botany 55, 736-748.
| Crossref | Google Scholar |
Bond WJ, Midgley JJ (2001) Ecology of sprouting in woody plants: the persistence niche. Trends in Ecology & Evolution 16, 45-51.
| Crossref | Google Scholar | PubMed |
Bradstock RA, Bedward M, Scott J, Keith DA (1996) Simulation of the effect of spatial and temporal variation in fire regimes on the population viability of a Banksia species. Conservation Biology 10, 776-784.
| Crossref | Google Scholar |
Burbidge AH, Danks A, Comer S, Smith GT (2025) Booderitj/western bristlebird and Dading/western whipbird: ecology and management of two of the lesser-known threatened species at Two Peoples Bay Nature Reserve, Western Australia. Pacific Conservation Biology 31, PC25021.
| Crossref | Google Scholar |
Burrows ND (2008) Linking fire ecology and fire management in south-west Australian forest landscapes. Forest Ecology and Management 255, 2394-2406.
| Crossref | Google Scholar |
Burrows ND, Wardell-Johnson G, Ward B (2008) Post-fire juvenile period of plants in south-west Australian forests and implications for fire management. Journal of the Royal Society of Western Australia 91, 163-174.
| Google Scholar |
Chatfield GR, Saunders DA (2024) History and establishment of Two Peoples Bay Nature Reserve. Pacific Conservation Biology 30, PC24004.
| Crossref | Google Scholar |
Clarke PJ (2002) Habitat islands in fire-prone vegetation: do landscape features influence community composition. Journal of Biogeography 29, 677-684.
| Crossref | Google Scholar |
Clarke MF (2008) Catering for the needs of fauna in fire management: science or just wishful thinking? Wildlife Research 35, 385-394.
| Crossref | Google Scholar |
Comer S, Burbidge AH (2006) Manypeaks rising from the ashes. Landscope 22(1), 51-55.
| Google Scholar |
Comer S, Clausen L, Cowen S, Pinder J, Thomas A, Burbidge AH, Tiller C, Algar D, Speldewinde P (2020) Integrating feral cat (Felis catus) control into landscape-scale introduced predator management to improve conservation prospects for threatened fauna: a case study from the south coast of Western Australia. Wildlife Research 47, 762-778.
| Crossref | Google Scholar |
Comer S, Danks A, Berryman A, Cowen S, Burbidge AH, Smith GT (2025) The role of conservation translocations in the recovery of the endangered Djimaalap/Noisy Scrub-bird (Atrichornis clamosus): from Two Peoples Bay Nature Reserve and back. Pacific Conservation Biology 31, PC25017.
| Crossref | Google Scholar |
Danks A (1997) Conservation of the Noisy Scrub-bird: a review of 35 years of research and management. Pacific Conservation Biology 3, 341-349.
| Crossref | Google Scholar |
Danks A, Calver MC (1993) Diet of the Noisy Scrub-bird Atrichornis clamosus at Two Peoples Bay, south-western Western Australia. Emu - Austral Ornithology 93, 203-205.
| Crossref | Google Scholar |
Danks A, Comer S, Burbidge AH (2011) Back from the brink: 50 years of conservation at Two Peoples Bay. Landscope 27(1), 32-38.
| Google Scholar |
Danks A, Burbidge AA, Coy NJ, Folley GL, Sokolowski RES (2024) Management of Two Peoples Bay Nature Reserve from 1967 to 1999. Pacific Conservation Biology 30, PC24076.
| Crossref | Google Scholar |
Department of Parks and Wildlife (2014) South Coast Threatened Birds Recovery Plan. Western Australian Wildlife Management Program No. 44. Department of Parks and Wildlife, Perth, WA. Available at http://www.environment.gov.au/system/files/resources/52c306c7-9085-4b62-a1dc-4d98c6ebae41/files/south-coast-threatened-birds-2014.pdf
Doherty TS, Geary WL, Jolly CJ, Macdonald KJ, Miritis V, Watchorn DJ, Cherry MJ, Conner LM, González TM, Legge SM, Ritchie EG, Stawski C, Dickman CR (2022) Fire as a driver and mediator of predator–prey interactions. Biological Reviews 97, 1539-1558.
| Crossref | Google Scholar | PubMed |
Doherty TS, Macdonald KJ, Nimmo DG, Santos JL, Geary WL (2024) Shifting fire regimes cause continent-wide transformation of threatened species habitat. Proceedings of the National Academy of Sciences 121, e2316417121.
| Crossref | Google Scholar |
Driscoll DA, Macdonald KJ, Gibson RK, Doherty TS, Nimmo DG, et al. (2024) Biodiversity impacts of the 2019–2020 Australian megafires. Nature 635, 898-905.
| Crossref | Google Scholar | PubMed |
Enright NJ, Marsula R, Lamont BB, Wissel C (1998) The ecological significance of canopy seed storage in fire-prone environments: a model for non-sprouting shrubs. Journal of Ecology 86, 946-959.
| Crossref | Google Scholar |
Fontaine JB, Westcott VC, Enright NJ, Lade JC, Miller BP (2012) Fire behaviour in south-western Australian shrublands: evaluating the influence of fuel age and fire weather. International Journal of Wildland Fire 21, 385-395.
| Crossref | Google Scholar |
Friend AJ, Sinclair EA, Courtenay JM (2025) From rediscovery towards recovery: a recent history of Australia’s most critically endangered marsupial, Gilbert’s potoroo (Potorous gilbertii). Pacific Conservation Biology 31, PC24103.
| Crossref | Google Scholar |
Gioia P, Hopper SD (2017) A new phytogeographic map for the Southwest Australian Floristic Region after an exceptional decade of collection and discovery. Botanical Journal of the Linnean Society 184, 1-15.
| Crossref | Google Scholar |
Gosper CR, Yates CJ, Prober SM, Parsons BC (2012) Contrasting changes in vegetation structure and diversity with time since fire in two Australian Mediterranean-climate plant communities. Austral Ecology 37, 164-174.
| Crossref | Google Scholar |
Hart RP, Freebury G, Barrett S (2024) Phytophthora cinnamomi: extent and impact in Two Peoples Bay Nature Reserve, Western Australia (1983–2024). Pacific Conservation Biology 30, PC24028.
| Crossref | Google Scholar |
Harvey BJ, Enright NJ (2022) Climate change and altered fire regimes: impacts on plant populations, species, and ecosystems in both hemispheres. Plant Ecology 223, 699-709.
| Crossref | Google Scholar |
Hopkins AJM (1985) Fire in woodlands and associated formations of the semi-arid region of south-western Australia. In ‘Fire ecology and management in Western Australian ecosystems’. (Ed. JK Ford) pp. 83–90. (Western Australian Institute of Technology Environmental Studies Group Report No. 14.: Bentley)
Hopkins AJM, Smith GT, Saunders DA (2024a) Introduction to the special issue of The Natural History of Two Peoples Bay Nature Reserve, Western Australia. Pacific Conservation Biology 30, PC24023.
| Crossref | Google Scholar |
Hopkins AJM, Williams AAE, Harvey JM, Hopper SD (2024b) A new vegetation classification for Western Australia’s Two Peoples Bay Nature Reserve and its significance for fire management. Pacific Conservation Biology 30, PC24036.
| Crossref | Google Scholar |
Hopper SD (2000) Creation of conservation reserves and managing fire on granite outcrops – a case study of Chiddarcooping Nature Reserve in the Western Australian wheatbelt. Journal of the Royal Society of Western Australia 83, 173-186.
| Google Scholar |
Hopper SD (2009) OCBIL theory: towards an integrated understanding of the evolution, ecology and conservation of biodiversity on old, climatically buffered, infertile landscapes. Plant and Soil 322, 49-86.
| Crossref | Google Scholar |
Hopper SD (2023) Ocbil theory as a potential unifying framework for investigating narrow endemism in Mediterranean Climate Regions. Plants 12(3), 645.
| Crossref | Google Scholar |
Hopper SD, Brown AP, Marchant NG (1997) Plants of Western Australian granite outcrops. Journal of the Royal Society of Western Australia 80, 141-158.
| Google Scholar |
Hopper SD, Harvey JM, Hopkins AJM, Moore LA, Smith GT (2024) Plant diversity on the edge: floristics, phytogeography, fire responses, and plant conservation of Two Peoples Bay Nature Reserve in the context of OCBIL theory. Pacific Conservation Biology 30, PC24024.
| Crossref | Google Scholar |
Hradsky BA, Mildwaters C, Ritchie EG, Christie F, Di Stefano J (2017) Responses of invasive predators and native prey to a prescribed forest fire. Journal of Mammalogy 98, 835-847.
| Crossref | Google Scholar |
Hunter JT (2003) Persistence on inselbergs: the role of obligate seeders and resprouters. Journal of Biogeography 30, 497-510.
| Crossref | Google Scholar |
Keith D (1996) Fire-driven extinction of plant populations: a synthesis of theory and review of evidence from Australian vegetation. Proceedings of the Linnean Society New South Wales 116, 37-78.
| Google Scholar |
Keith D, Lindenmayer D, Lowe A, Russell-Smith J, Barrett S, Enright N, Fox B, Guerin G, Paton D, Tozer M, Yates C (2014) Heathlands. In ‘Biodiversity and environmental change: monitoring, challenges and direction’. (Eds D Lindenmayer, E Burns, N Thurgate, A Lowe) pp. 215–285. (CSIRO Publishing: Melbourne)
Knapp L, Cummings D, Cummings S, Fiedler PL, Hopper SD (2024) A Merningar Bardok family’s Noongar oral history of Two Peoples Bay Nature Reserve and surrounds. Pacific Conservation Biology 30, PC24018.
| Crossref | Google Scholar |
Legge S, Rumpff L, Woinarski JCZ, Whiterod NS, Ward M, Southwell DG, Scheele BC, Nimmo DG, et al. (2022) The conservation impacts of ecological disturbance: time-bound estimates of population loss and recovery for fauna affected by the 2019–2020 Australian megafires. Global Ecology and Biogeography 31, 2085-2104.
| Crossref | Google Scholar |
Lullfitz A, Pettersen C, Reynolds R, Eades A, Dean A, Knapp L, Woods E, Woods T, Eades E, Woods S, Dortch J, Guilfoyle D, Hopper SD (2021) The Noongar of south-western Australia: a case study of long-term biodiversity conservation in a matrix of old and young landscapes. Biological Journal of the Linnean Society 133, 432-448.
| Crossref | Google Scholar |
McArthur WM, Bartle GA, Evans KA, Tyler IM (2024) Landforms and soils of Two Peoples Bay Nature Reserve, Western Australia. Pacific Conservation Biology 30, PC24092.
| Crossref | Google Scholar |
Merritt DJ, Martyn AJ, Ainsley P, Young RE, Seed LU, Thorpe M, Hay FR, Commander LE, Shackelford N, Offord CA, Dixon KW, Probert RJ (2014) A continental-scale study of seed lifespan in experimental storage examining seed, plant, and environmental traits associated with longevity. Biodiversity and Conservation 23(5), 1081-1104.
| Crossref | Google Scholar |
Ne’eman G, Fotheringham CJ, Keeley JE (1999) Patch to landscape patterns in post fire recruitment of a serotinous conifer. Plant Ecology 145, 235-242.
| Crossref | Google Scholar |
O’Brien TG (2011) Abundance, density and relative abundance: a conceptual framework. In ‘Camera traps in animal ecology’. (Eds AF O’Connell, JD Nichols, KU Karanth) pp. 71–96. (Springer Japan: Tokyo) 10.1007/978-4-431-99495-4_6
O’Connell AF, Bailey L (2011) Inference for occupancy and occupancy dynamics. In ‘Camera traps in animal ecology’. (Eds AF O’Connell, JD Nichols, KU Karanth) pp. 191–204. (Springer Japan: Tokyo) 10.1007/978-4-431-99495-4_11
Pausas JG, Pratt RB, Keeley JE, Jacobsen AL, Ramirez AR, Vilagrosa A, Paula S, Kaneakua-Pia IN, Davis SD (2016) Towards understanding resprouting at the global scale. New Phytologist 209, 945-954.
| Crossref | Google Scholar | PubMed |
Perkins AJ, Dilly ML (2017) Hydrocotyle serendipita (Araliaceae), a new species of fire ephemeral from south-western Australia. Telopea 20, 269-275.
| Crossref | Google Scholar |
Povh LF, Willers N, Shephard JM, Fleming PA (2023) A conservation-significant threatened mammal uses fire exclusions and shifts ranges in the presence of prescribed burning. International Journal of Wildland Fire 32, 1291-1303.
| Crossref | Google Scholar |
Roberts JD, Danks A, Berryman A, Sidhu N, Burbidge AH, Comer S (2020) Population decline of the Noisy Scrub-bird is not correlated with territory size, marginal declines in rainfall or fire impacts. Pacific Conservation Biology 26, 230-238.
| Crossref | Google Scholar |
Russell-Smith J, McCaw L, Leavesley A (2020) Adaptive prescribed burning in Australia for the early 21st Century – context, status, challenges. International Journal of Wildland Fire 29, 305-313.
| Crossref | Google Scholar |
Rye BL (1999) A taxonomic revision of many-flowered species of Trachymene (Apiaceae) in Western Australia. Nuytsia 13, 193-232.
| Crossref | Google Scholar |
Sinclair EA, Danks A, Wayne AF (1996) Rediscovery of Gilbert’s Potoroo, Potorous tridactylus, in Western Australia. Australian Mammalogy 19, 69-72.
| Crossref | Google Scholar |
Smith GT (1977) The effect of environmental change on six rare birds. Emu - Austral Ornithology 77, 173-179.
| Crossref | Google Scholar |
Smith GT (1985a) Fire effects on populations of Noisy Scrub-bird (Atrichornis clamosus), Western Bristlebird (Dasyornis longirostris) and Western Whipbird (Psophodes nigrogularis). In ‘Fire ecology and management in Western Australian ecosystems’. (Ed. JK Ford) pp. 95–102. (Western Australian Institute of Technology Environmental Group Report No. 14: Bentley)
Smith GT (1985b) Population and habitat selection of the Noisy Scrub-bird, Atrichornis clamosus, 1962–83. Australian Wildlife Research 12, 479-485.
| Crossref | Google Scholar |
Smith GT, Forrester RI (1981) The status of the Noisy Scrub-bird Atrichornis clamosus. Biological Conservation 19, 239-254.
| Crossref | Google Scholar |
Smith GT, Robinson FN (1976) The Noisy Scrub-bird, an interim report. Emu - Austral Ornithology 76, 37-42.
| Crossref | Google Scholar |
Spencer PBS, Bain K, Hayward MW, Hillyer M, Friend JAT (2019) Persistence of remnant patches and genetic loss at the distribution periphery in island and mainland populations of the quokka. Australian Journal of Zoology 67, 38-50.
| Crossref | Google Scholar |
Syme K, Lebel T, Hilton RN (2025) Macrofungi of Two Peoples Bay Nature Reserve, Western Australia. Pacific Conservation Biology 31, PC24091.
| Crossref | Google Scholar |
Tyler IM, Playford PE, Evans KA (2024) Geology and landforms of Two Peoples Bay Nature Reserve, Western Australia. Pacific Conservation Biology 30, PC24084.
| Crossref | Google Scholar |
Underwood R J, Sneeuwjagt RJ, Styles HG (1985) The contribution of prescribed burning to forest fire control in Western Australia: case studies. In: ‘Fire ecology and management in Western Australian ecosystems’. (Ed. JK Ford) pp. 153–170. (Western Australian Institute of Technology Environmental Studies Group Report No. 14: Bentley)
van Oldenborgh GJ, Krikken F, Lewis S, Leach NJ, Lehner F, Saunders KR, van Weele M, Haustein K, Li S, Wallom D, Sparrow S, Arrighi J, Singh RK, van Aalst MK, Philip SY, Vautard R, Otto FEL (2021) Attribution of the Australian bushfire risk to anthropogenic climate change. Natural Hazards and Earth System Sciences 21, 941-960.
| Crossref | Google Scholar |
Ward M, Tulloch AIT, Radford JQ, Williams BA, Reside AE, Macdonald SL, Mayfield HJ, Maron M, Possingham HP, Vine SJ, O’Connor JL, Massingham EJ, Greenville AC, Woinarski JCZ, Garnett ST, Lintermans M, Scheele BC, Carwardine J, Nimmo DG, Lindenmayer DB, Kooyman RM, Simmonds JS, Sonter LJ, Watson JEM (2020) Impact of 2019–2020 mega-fires on Australian fauna habitat. Nature Ecology & Evolution 4, 1321-1326.
| Crossref | Google Scholar | PubMed |
Woinarski JCZ, Burbidge AH, Comer S, Harley D, Legge S, Lindenmayer DB, Partridge TB (2015) Fire and biodiversity in Australia. In ‘Austral Ark: the state of wildlife in Australia and New Zealand’. (Eds A Stow, N Maclean, GI Holwell) pp. 537–559. (Cambridge University Press: Cambridge, England) 10.1017/CBO9781139519960.027
Yates CJ, Ladd PG (2004) Breeding system, pollination and demography in the rare granite endemic shrub Verticordia staminosa ssp. staminosa in south-west Western Australia. Austral Ecology 29, 189-200.
| Crossref | Google Scholar |
Yates CJ, Ladd PG (2005) Relative importance of reproductive biology and establishment ecology for persistence of a rare shrub in a fragmented landscape. Conservation Biology 19, 239-249.
| Crossref | Google Scholar |
Yates CJ, Hopper SD, Brown A, van Leeuwen S (2003) Impact of two wildfires on endemic granite outcrop vegetation in Western Australia. Journal of Vegetation Science 14, 185-194.
| Crossref | Google Scholar |
