Fire regimes in southeastern Australia are projected to become more frequent and intense, increasing the likelihood of ecosystems burning outside their ecological tolerance. Shorter fire return intervals can negatively affect fire tolerant resprouting Eucalyptus species, leading to resprouting failure, tree mortality, and ecosystem collapse. To measure the implications of such regime shifts, we compared Eucalyptus forests that retained their canopy under tolerable fire regimes (reference forests) with those that lost their canopy following short-interval fires (alternative forests) and persisted through subsequent fires.
Alternative state forests differed significantly in composition and structure, resulting in altered ecosystem function. Non-metric multidimensional scaling (NMDS) indicated that understory species and growth form composition differed between states. Alternative forests were dominated by small trees (<20 cm DBH), with 97% fewer large trees (>20 cm DBH) per hectare than reference forests. These forests exhibited structural simplification, including reduced tree diameter variability
These changes impacted ecosystem services such as carbon storage and habitat provision. The absence of large trees capable of forming hollows, combined with restricted Eucalyptus dispersal, suggests that recruitment of hollow-bearing trees would be slow, limiting long-term habitat availability for hollow-dependent fauna. Furthermore, the mass of coarse woody debris resources were 38% lower. Forests in an alternative state stored 59% less carbon, with 98% of carbon that in stems <20 cm DBH compared to 20% in reference forests. Thus, alternative forests are more vulnerable to mortality and consumption during subsequent fires and potentially become a carbon source.
Given the close link between ecosystem function and the services they provide, increased fire frequency threatens to disrupt carbon dynamics and biodiversity in these forests. Our findings highlight risk to forest structure and function posed by short-interval fires and some of the long-term consequences of state changes for carbon sequestration, carbon stability and biodiversity conservation in fire-prone landscapes.

Background
Drought related tree mortality is rising globally. In Australia, extreme drought from 2017-2019, compounded by heatwaves in 2019, was associated with widespread canopy browning, or dieback, across native eucalypt forests. Eucalypts may remain dormant following severe stress, emphasising the need for long-term recovery to distinguish delayed recovery from mortality.

Objectives
We aimed to quantify the extent of tree mortality resulting from the 2019 drought and dynamics of recovery in surviving trees. We sought to understand where recovery occurs, and why recovery differs across sites.

Methods
We quantified post-drought recovery across 35 plots using ground-based surveys and satellite imagery at three timepoints: drought (2019), short-term recovery (2020), and long-term recovery (2023-2024). Canopy health was assessed using a 20-point index based on canopy extent, density, browning, and dead branches. Hydraulic function was measured using percentage loss of conductivity (PLC). Sentinel-2-derived Normalised Burn Ratio (NBR) was used to evaluate return to baseline levels of greenness.

Results
At the long-term recovery time point, 14±2.8% of basal area was dead, reduced from 24±5% at short-term recovery, but higher than 5.2±0.86% during drought. Small trees (<10cm DBH) were most vulnerable and least able to resprout; large trees (>30cm DBH) exhibited continued epicormic resprouting, indicating delayed, active recovery. NBR returned to baseline levels at long-term recovery. Health scores were highest at wetter sites and where rainfall was high in 2022.

Conclusions
Our findings underscore the need for long-term monitoring to capture delayed tree recovery. Integration of ground surveys with remote sensing expanded our understanding of forest resilience and recovery patterns. Under increasing frequency of drought and heatwaves, these insights are critical to inform forest management.

Australia has already warmed by 1.5 °C since 1910, and warming may reach 3 °C or more by the second half of the century. Climate warming is associated with more extreme events, such as heatwaves during drought. Tree species differ in their ability to tolerate these extreme events but it is difficult to predict species vulnerability, in part because key plant functional traits associated with drought-induced tree dieback often fail to capture responses to the combination of heat plus drought stress. To address this challenge, we selected ten Eucalyptus species originating along an aridity gradient ranging from arid to humid habitats, with mean annual precipitation of 340-1240 mm. Trees were exposed to moderate drought stress followed by a five-day heatwave with a maximum temperature of 45 °C in a controlled whole-tree chamber experiment. Half of the species were vulnerable, having >10% topkill and/or >20% crown dieback. Most functional traits were correlated to the observed differences in maximum quantum yield of PSII (Fv/Fm) among plants under drought stress alone, including xylem pressure inducing 50% loss of hydraulic conductivity due to embolism (R²=0.26), leaf size (R²=0.21), and leaf mass per area (R²=0.13). Fewer traits were correlated to the Fv/Fm of heat- and drought-stressed plants, although correlations existed for the leaf water potential at turgor loss point (πtlp; R²=0.17), wood density (R²=0.13), and leaf temperature (R²=0.11). The πtlp was also correlated with crown dieback among trees (R²=0.20) and species (R²=0.70), while maximum leaf temperature (Tleaf_max) was correlated with topkill among trees (R²=0.24) and species (R²=0.53). These results highlight the importance of combined heat and drought stress for Eucalyptus vulnerability and recovery, emphasising that functional traits such as πtlp and Tleaf_max are critical predictors of tree dieback under future climate extremes in Australia.

In Autumn 2024, following an extremely hot and dry summer and autumn, southwestern Australia experienced a widescale vegetation die-off, spanning over 1000km and a range ecosystem types. Although this caused a sense of anxiety in the broader community, in this presentation we will outline how work following previous die-off events in Australia, and globally, can help develop steps forward in terms of mapping, understanding, and managing die-off events. Following forest die-off in southwestern Australia in 2011, we tracked forest structural development and investigated flow-on effects. Our data shows that many forest die-off sites have undergone historical structural changes, and others have experienced multiple disturbance events, such as fire and additional drought/heat events. A range of post-drought die-off consequences have been reported upon, from carbon dynamics, fuels and fire potentials, reptiles, birds, wood boring insects, and microbial communities. Management options for these vulnerable sites range from a do-nothing approach to detailed ongoing research to understand how, where and when forests are vulnerable, and how the forest responds over time, to proactive intervention like ecological thinning, to finding climate-hardy genotypes for assisted migration. Outreach and engagement are also key components of forest conservation and management, particularly during die-off events, and, over the past year, due to the high level of anxiety across the community regarding drought and heat-induced forest die-off, a range of public presentations have been conducted by the team.

Background: The World Heritage-listed Gondwana rainforests of eastern Australia include vast areas of warm temperate, cool temperate, and the world’s most extensive subtropical rainforests. These forests have evolved under relatively stable rainfall patterns and are adapted to withstand seasonally dry conditions. However, climate change is now driving more frequent and extreme droughts, significantly increasing the risk of bushfires and placing these ecosystems under growing environmental pressure.

Objectives: The objective of this study was to establish essential baseline data on the hydraulic traits and drought vulnerability of key rainforest tree species in the Gondwana rainforests, and to assess post-fire recovery following the 2019/20 bushfires.

Methods: Physiological measurements were conducted to quantify hydraulic vulnerability in dominant rainforest tree species. Post-fire recovery was evaluated by comparing basal area, resprouting, and tree mortality in burnt and unburnt sites over a five-year period.

Results: Most species operate within moderate hydraulic safety margins but show lower drought tolerance compared to Australia’s sclerophyll forests. Species with restricted distributions, such as Nothofagus moorei, are particularly vulnerable, losing hydraulic function under relatively mild water deficits. Five years on, recovery remains slow and uncertain. Burnt areas still have reduced basal area relative to unburnt sites, and resprouting has proven to be an unreliable predictor of long-term survival. Ongoing tree mortality and the lasting impacts of fire severity are reshaping canopy health and forest structure.

Conclusions: These physiological vulnerabilities indicate that drought preceding the 2019/20 fires likely heightened both the susceptibility of these forests to fire and the severity of fire impacts. While the Gondwana rainforests demonstrate notable resilience, their recovery is complex and slow. Full restoration may require decades to centuries. Ongoing monitoring and predictive modelling, informed by this baseline data, are essential for guiding effective conservation and management as these ancient forests face mounting environmental threats.

Global climate change is intensifying wildfires and driving more frequent, severe droughts and pathogen invasions, fundamentally reshaping native plant communities in forest ecosystems. While the effects of these disturbances on aboveground vegetation are well documented, their combined impacts on belowground function, essential for nutrient uptake and survival, remain poorly understood.

The Northern Jarrah Forest in southwestern Western Australia has experienced marked declines in rainfall and rises in temperature since the 1970s. Fires now occur more frequently, and the invasive soil pathogen Phytophthora cinnamomi is gradually spreading throughout the forest. As some areas remain disease-free, we investigated how drought and fire history influence the invasion and impact of P. cinnamomi in these ecosystems.

In a glasshouse experiment, we grew three native species with distinct nutrient acquisition strategies in soils collected from sites with different fire histories: burned 0–3, 4–6, and >10 years ago. P. cinnamomi was inoculated into half the pots to assess its effects across disturbance gradients.

P. cinnamomi and fire history had distinct, species-specific impacts. Banksia grandis was highly susceptible to infection, with earlier mortality in recently burned soils. In Eucalyptus marginata, recent fire reduced beneficial mycorrhizal symbionts, and pathogen infection compounded this decline. In Acacia acuminata, nodulation, critical for nitrogen fixation, declined with fire alone but increased when both disturbances were present.

By simulating P. cinnamomi invasion across post-fire soils, this study shows how disturbance history and species traits influence plant–soil relationships, revealing key belowground mechanisms that shape ecosystem recovery under compounding threats.

Climate change is expected to drive shifts in ecosystem composition for a wide range of canopy species. Species that historically occur in any given location may experience changes in environmental suitability, which in turn will drive turnover in species composition across geography. The environmental impacts of climate change may not be contiguous across the landscape, resulting in areas of climate refugia as species distributions contract from the limits of their ecological niche. The alpine and montane regions of Australia are already experiencing signs of both climate stress (both ecological and environmental), while also being an area of national cultural significance. This represents a particular problem for land managers and conservation practitioners, as they are responsible for maintaining ecosystem function, values, and services while the ecosystem changes at unprecedented rates. Given the uncertainty, it can be difficult to identify priority areas to target limited resources and conservation effort. National scale species distribution modelling can often lack model specificity and resolution to guide this decision making at finer geographic scales. This study aims to model projected changes in environmental suitability for canopy species and identify climate refugia across alpine and montane Australia. In turn, helping form a targeted spatial modelling product to guide resist, adapt, and direct (RAD) decision making for stakeholders in the region. I produced MaxEnt species distribution models under a variety of future climate scenarios for E. pauciflora species complex, and canopy species from surrounding ecosystems. My results highlight significant changes in potential distribution of historical canopy species across the region, identifying areas of climate refugia and range contraction for conservation management.