Blackout burning in dry conditions increases long-term fire severity risk

Introduction

Changes to land uses in Australia as a result of European settlement largely prevented the continuation of Indigenous cultural burning and frequent, low intensity burning, leading to altered fire regimes (Mariani et al. 2022). This change has caused a loss of open forest structures (Williams et al. 2020; Mariani et al. 2022) and key habitats for species such as eastern bristlebird (Dasyornis brachypterus; Stone et al. 2022) habitat in Southeast Queensland and northern New South Wales. Many areas in Queensland were historically burned by Aboriginal peoples (Fensham 1997), including on Minjerribah (North Stradbroke Island) by the Quandamooka people (Ngugi et al. 2020). The frequent and purposed use of fire limited the severity of wildfire, maintained structure, kept travel ways open, provided a mosaic of burn diversity and facilitated longer unburned places. Changed fire regimes on Minjerribah and other areas in Southeast Queensland has likely influenced the overall vegetation structures and inherent fire risk at a landscape scale (State of Queensland (Department of Environment and Science) 2022). These changed landscape structures likely play a role in the outcomes and impacts of both unplanned and planned fire events.

Recent extreme fire events across the globe have had broad impacts (Wang et al. 2020; Ward et al. 2020; Pivello et al. 2021) and are predicted to become more frequent and more severe (Shi et al. 2021). This has led to calls for an increase in the extent and frequency of prescribed burning to reduce risk (State of Queensland (Inspector-General Emergency Management) 2019; DPC 2020). Land managers often prioritise burning vegetation adjacent to assets (i.e. infrastructure and habitation areas) in programs variously described as Hazard Reduction, Asset Protection or Wildfire Mitigation burning. These patches of vegetation are often considered ‘sacrificial areas’ since the ecological values are secondary to the priority of reducing fuel hazard (Marlow 1994; Rose et al. 1999; Corey et al. 2020). Fire operators often delay burning until fuel and soils are sufficiently dry to ensure a comprehensive burn in these areas to remove all or most of the surface, near surface and elevated fine fuels. This delay in burning (until conditions are dry) may also be perceived as more efficient and less time-consuming than burning under moist, mild conditions, where fires travel more slowly and multiple spot ignitions may require more time to implement. However, whether delaying burning until conditions are dry results in reduced or heightened long-term fuel hazard, has not been adequately evaluated.

We explore various case studies in Southeast Queensland’s open, dry eucalypt-dominated or heathy forest/woodlands representing different fire regimes to discuss the impact of changed fire regimes on vegetation structure and fuel risk.

Materials and methods

Five case studies are presented whereby vegetation plots were surveyed and/or Overall Fuel Hazard (OFH) (sensu Hines et al. 2010) measured. Each vegetation community is allocated to a Regional Ecosystem (RE), which is the basis of Queensland’s vegetation classification system (Neldner et al. 2022). In our case studies, saplings were defined as small single stemmed trees up to 6 m tall. Shrubs are multi-stemmed woody plants, typically up to 6 m tall. The case studies are outlined below with a summary provided in Table 1.

Results and discussion

The case studies presented highlight patterns in wildfire and planned burn outcomes across various locations in Southeast Queensland. While most of the case studies (excluding case study 4) present site data with limited replication, combined they show a consistent pattern across the Southeast Queensland landscape. The following describes the patterns observed using our data, with a focus on Acacia density due to key patterns that were observed from this genus.

Multiple studies following the 2019–2020 wildfires of eastern Australia found the severe wildfires in dry conditions, combined with wet conditions in the following year, promoted an increase in shrubs and sapling density (Barker et al. 2022; Qin et al. 2022), and therefore elevated fuel loads (Barker et al. 2022).

Our case studies representing wildfires across Southeast Queensland (Brisbane, Mulgumpin, Minjerribah and Bauple) promoted shrub growth, sapling density and elevated fuel thickening (see example: Figs 1 and 2c), dramatically increasing the elevated fuel loads and OFH up to high to extreme (Table 2). For example, in the Brisbane Eucalyptus racemosa forest, a low intensity planned burn under mild conditions in 2013 only stimulated moderate numbers of wattle (Acacia concurrens) seedlings, with their density increasing from three to seven plants in the first year after fire and to 38 in the second year after fire, after which their numbers began to decrease via natural thinning. In contrast, a subsequent high intensity wildfire at the same transect greatly increased wattle recruitment to 799 seedlings. Four years after the wildfire, approximately 539 wattle plants remained, averaging 2.5 m tall, dramatically increasing the elevated fuel density (Fig. 1). The fuel hazard correspondingly increased from moderate in the year after the 2019 wildfire, to extreme hazard by 4 years after wildfire, due to extreme elevated fuel loads (Table 2). Mulgumpin data showed that wildfire impacted areas had an average sapling density of 434.5 ± 42.5 standard error (s.e.), whereas the regularly burned areas adjacent had a density of 31.5 ± 28.5 s.e. In addition, a higher fuel hazard was recorded in the wildfire impacted plots (Table 2). These data confirmed the effect of wildfires in dry conditions promoting excess saplings, especially of wattles, epacrids, eucalypts and bracken fern (Pteridium esculentum), and therefore fuel hazard and bushfire risk.

Fig. 1.

The cluttering of a Brisbane eucalypt forest through excessive wattle recruitment following a high intensity wildfire. Each of the three graphs represents an aerial view of a single 100 m² plot in different years, with each dot representing the position of wattle plants, showing (a) the original four Acacia concurrens plants in 2009; (b) wattle density peaked at 38 plants in the second year after a low intensity 2013 burn; (c) approximate spread of 539 wildfire-promoted Acacia concurrens plants across the plot that survived into the fourth year (2023) after a 2019 wildfire.

Fig. 2.

Differences in (a) proportion burned and (b) burn patch heterogeneity between annual and triennial burn treatments in different seasons (mean ± s.e.) at Bauple State Forest. Graph (c) shows shrub canopy cover within spotted gum woodland at Bauple State Forest with significant differences among fire type (representative of WF = wildfire, PB = prescribed/planned burn, and LU = long unburned) from Kruskal–Wallis tests (P < 0.05), showing mean ± s.e. The dashed line represents the benchmark value for this RE. Where significant differences occurred from post hoc tests, these are shown with lettering (P < 0.05).

High intensity wildfires promote excessive shrubs and saplings through increased topsoil heat penetration that breaks the dormancy of hard seeded species (Bradstock and Myerscough 1981; Walters et al. 2004) and via increased light following canopy damage. This wildfire-promoted elevated fuel load increases the risk of future higher intensity fires, described as a positive feedback loop by Karna et al. (2021). Dense elevated fuel loads reduce the gap between mid-strata and canopy, increasing forest flammability and potentially causing the loss of old canopy trees (Palmer et al. 2018; Williams et al. 2020).

A lack of fuel management post-wildfire in these areas means an increased fire risk which may lead to a promotion of further high-intensity wildfires due to the vegetation structure that is created, i.e. limited ground fuels and dense elevated fuels often requiring dry conditions and wind-driven fires to burn. Post-wildfire areas were also found to have limited surface fuels. While both shrubs and woody saplings were considered together in this study, and those woody saplings are likely to grow out of the elevated fuel layer (no longer forming a dense fuel load), it should be noted that limited reduction in fuel hazard occurred over up to 8 years post-fire, highlighting a long period of increased fuel hazard and potential wildfire risk. A similar process has been recorded across southeast Australia where wildfires promote further wildfire impact (Gordon et al. 2017; Barker et al. 2022). The greater mid-strata shrub and sapling density shades out grasses, making fires under moist, mild conditions more difficult to implement, causing further planned burns to be delayed until conditions are drier (Baker et al. 2021; Williams et al. 2023), thereby contributing to the positive feedback loop. The excessive elevated fuel load following high intensity wildfires therefore increases the longer-term bushfire risk.

The broadscale impacts of the 2019–2020 wildfires across eastern Australia, burning more than 8 million ha of vegetation (Godfree et al. 2021), means that a broad geographic range now has a relatively consistent fuel age and likely vast areas impacted by woody thickening (due to the conditions under which the burns occurred - extreme fire weather). This has implications for the current and future fire risk across the east coast of Australia and specifically in Southeast Queensland, with the potential for further unplanned wildfires to occur across broad geographic regions when conditions allow. This phenomenon has been observed across New South Wales, where areas burned in 2019–2020 impacted by woody thickening have recently been observed to promote large intense wildfires across the state (Field 2024).

Our case studies provide evidence that planned burns in dry conditions can cause similar excessive sapling densities as severe wildfires. Planned burns under dry conditions measured at two locations (Brisbane and Minjerribah) promoted the recruitment of large numbers of shrubs, including wattles and bracken fern. OFH increased up to extreme by the third or fourth year after fire due to increased shrub or elevated fuel density (Table 2). An assessment of burn patchiness conferred that later season (spring burns) had higher percentage cover and lower burn heterogeneity (Fig. 2a, b).

Shrub and sapling thickening can smother out grasses and herbs, reducing plant diversity (Baker et al. 2020; Williams et al. 2020; Stone et al. 2022). Where grasses have disappeared, the capacity to implement fires under mild and moist conditions is reduced. This is because native grasses, such as kangaroo grass (Themeda triandra), fuel fires that carry under moist conditions. In contrast, forests with only leaf litter and lacking grass cover are difficult to burn until conditions dry further (Williams et al. 2023). Therefore, thickened vegetation can further lead to a hot fire trap, whereby hot, dry planned fires are implemented repetitively over time to continuously reduce the fuel risk in the short term, as previously implemented on Minjerribah. The consequence of this to the land managers is a cost inefficiency in the long-term, through repetitively burning high risk areas whilst not maintaining a low fuel hazard or healthy forest structure and likely impacting on biodiversity values.

Both planned and unplanned fires in areas of extreme elevated fuels or thickened woody vegetation are likely to impact on ecological and cultural values by increasing the flame height and fire intensity, leading to potential impacts on sensitive species, large significant trees and other values. An extensive loss of significant, large trees on Minjerribah has been observed in areas of wildfire impact over recent decades, including eucalypts and culturally significant cypress pines (Kington et al. 2016). The implementation of planned burns under dry conditions with increased elevated fuels are also a higher risk due to the increased flame heights and intensity, making the burns more difficult to control.

All case studies across Southeast Queensland showed that low intensity burns did not promote excessive regrowth (i.e. Fig. 1b) and maintained lower OFH (primarily low-moderate) for up to 4-years post-burn (Table 2). Sapling density was consistently lower in frequent, low intensity burn areas compared to wildfire impacted areas. This was evidenced even when patchy burns were implemented such as those at the frequently burned (1–3 years) Bauple experimental sites. A low intensity planned burn in Brisbane eucalypt forest thinned the number of pre-existing wattles (from 16 to 7 by 4 years after) with only moderate wattle recruitment. The average height of wattles dropped from 2.2 m prior to the burn, to 0.71 m 3 years after fire, thereby reducing the elevated fuel height. This highlights the ecologically beneficial result of a mild fire in restoring an open forest structure. This matches other studies highlighting the benefits of low intensity burning (e.g. McKemey et al. 2019, 2021; Ngugi et al. 2020; Radford et al. 2020).

Excessive elevated fuel loads can be avoided through regular low intensity prescribed burning with good soil moisture. Frequent, low intensity burning is akin to historical land management practices of Aboriginal peoples (McKemey et al. 2019; Greenwood et al. 2022) including those of the Quandamooka peoples of Minjerribah and Mulgumpin (Ngugi et al. 2020), and therefore likely similar to the needs of Southeast Queensland’s cultural and ecological landscapes. In open, dry eucalypt forests and heathy woodlands, frequent burning is recommended to both maintain lower fuel hazards as well as healthy, open forest systems with healthy grass and forb layers (State of Queensland (Department of Environment and Science) 2022; Williams et al. 2023). Whether elevated fuel thickening can be reversed requires further detailed assessments, however the following section touches on preliminary findings from one case study.

An important question the results of the case studies’ pose is: ‘how do land managers deal with post-wildfire landscapes or areas with woody thickening?’. A trial low intensity planned burn in a wildfire-impacted area on Minjerribah (transect 3c) reduced post-fire regrowth (measured 3 years after wildfire) from extreme OFH to moderate (measured 4 years after the planned burn) (Table 2). This burn began to re-instate open forest structures and reduce recurring fuel risk. Keenan et al. (2021) highlights that mechanical thinning in combination with fire could be a successful tool to reduce fuel loads and fire hazard in eucalypt forests.

Recently, burning has been used in combination with mechanical thinning on Minjerribah and Russell Island. Burning in wildfire-impacted areas is recommended to be done with thorough consideration and management of conditions to ensure sufficient soil moisture is present and conditions will thereby allow for a small, patchy, low intensity fires. This should be done before woody vegetation is too tall to manage through burning, i.e. 1–4 years post-fire, and with consideration of combining mechanical thinning to break up sections or alter fuel structure for burning. However, where thickening inhibits low intensity fire, a burn in dry conditions may be the only option. This would require follow up of subsequent burns conducted that are more favourable to grasses. While grasses and sparse shrubs persist, managing for these by use of frequent low intensity burning while there is good soil moisture is ideal.

Conclusion

The hot fire trap caused by woody thickening promotes excessive elevated fuel making it difficult to break up fuel continuity and manage forest structure, necessitating delays in burning until fuel is particularly dry and flammable. Hazard reduction burns in dry conditions promote excessive woody seedlings which can grow into a dense elevated fuel load. This increases future fire severity risk adjacent to assets, requiring further burning under dry conditions, that puts the environment, large trees and assets at risk. In a changing climate, where bushfires are predicted to be more frequent and of higher intensity, this outcome is not sustainable.

Low intensity, patchy mosaic burning while the remains good soil moisture, every 1–4 years, is recommended in open, dry eucalypt communities in Southeast Queensland to restore vegetation condition. Burning at this frequency would mean generally the patches of fuel that are a few years old would burn and therefore different patches of vegetation would burn at various intervals, allowing for mosaics of fire intervals required for appropriate landscape scale fire management that promotes biodiversity (i.e. Wills et al. 2020; Partridge et al. 2023). Overabundant native saplings would be reduced, including wattles and eucalypts, recognised as an indicator of unhealthy subtropical eucalypt forests (Williams et al. 2020; Melzer and Hines 2022).

Our experience is that very frequent burning can be carried out even when it is a bit dry without detrimental impacts on kangaroo grass, whereas allowing 4 years and above between fires increases the surface fuel, thereby increasing fire intensity and impacting grass clumps. This also increasing blady grass (Imperata cylindrica) and bracken fern, which regenerate via rhizomes, to the detriment of kangaroo grass. Under the current recommendations of 4–6 years (and often 4–25 years), it has been observed that kangaroo grass dominated open forest shifts to a blady grass dominated understorey. This builds and carries higher fuel loads more rapidly, subsequently worsening the surface/near surface fuel hazard, and also stimulates thickening and allows time for excess establishment of resprouters. Healthy open forest with a kangaroo grass dominated ground layer presents an ideal risk management option for asset mitigation and protection of ecological and cultural values. This will lower fire risk to the communities, produce better ecological outcomes and reduce the cost of implementation over the long-term. In addition, frequently burned areas are more resilient to extreme fire events, limiting woody thickening post-wildfire in those areas (Williams et al. 2022).

The initial implementation costs of applying low intensity, slow moving prescribed burns is higher; however, the long-term outcome is a more sustainable approach that reduces longer-term fire hazard. To achieve this, land managers must be nimble, carrying out burning that is adjusted to conditions and changing seasonality. Using a triage approach to prioritise healthy country will facilitate cost effective management where land managers work backwards from healthy areas aiming to return forests to a lower fuel risk and healthier vegetation profile. This study is a preliminary exploration of the influence of fire regimes on woody vegetation thickening with a focus on Acacia species, and therefore further research is required to better quantify the influence of fire regimes on woody thickening and forest condition in eucalypt forests as well as best practice management outcomes.