Around the world forest ecosystems are exhibiting dieback events that are likely to herald widespread species turnover. These changes are particularly impactful in systems, like the Australian sub-alpine, where the forest canopy is largely monospecific. Decline or loss of the dominant canopy species in such systems will alter the ecological, hydrological and cultural values of the system, and in the case of snow gums (Eucalyptus pauciflora group) stands to impact state and national water-supply and power-generation systems. Our multi-disciplinary investigation draws together collaborators across institutions and disciplines to: 1) assess the future geography of snow gum dieback in the high country and identify priority locations for pro-active management, 2) quantify the impact of snow gums on high country water and carbon budgets and thus the socio- economic and biodiversity values, and 3) determine options for mitigation across the sub-alpine and montane snow gum taxa. Our research is developing solutions including spatially predictive tools, candidate seedstocks for pro-active restoration, and scenario analyses to assess value propositions. Together we are building understanding of the diverse values of a treed ecosystem and supporting evidence-based examination of what and whether mitigation or restoration is warranted. These outputs will enable up-scaling for high country management efforts that are co-designed with conservation and management partners and supported by knowledge and tools that maximise the prospect for long term success.

Native Australian trees are increasingly impacted by drought-induced dieback, threatening forest ecosystems. Refugia, places where the impacts of climate change are less severe, can constitute an option for in situ persistence of forests. However, refugia will still experience some impacts of climate change. Management approaches that utilize the buffering provided by climate-change refugia, while recognizing their vulnerability, are needed.

In a unique population of Eucalyptus macrorhyncha in South Australia, we investigated the refugial potential of topography from climate change-driven hot droughts. We established a gradient from good (putative microrefugia) to poor canopy health, and identified microrefugia in locations that received less solar radiation and were cooler and moister than other habitats. Physiological measurements (percent loss of conductivity) of trees indicate that microrefugia are already impacted by water stress during droughts, but less than more exposed habitats. Strong regeneration was observed in habitats with canopy dieback between 25% and 70%, but was absent in areas dying back heavily. Therefore, in situ persistence and recovery of populations should be feasible in the shorter term, but we recommend interventions during periods of acute stress. Watering of targeted microrefugia and selected adjacent areas of high regeneration during extreme heatwaves and droughts could prevent hydraulic damage that triggers canopy defoliation and maintain a buffer around selected microrefugia.

High-elevation subalpine woodlands of southeastern Australia are characterized by monospecific stands of snow gum (Eucalyptus pauciflora), making this species a keystone in its ecosystem. A widespread insect-induced dieback, associated with the native longicorn beetle (Phoracantha mastersi), is occurring throughout the Alps. However, the spatiotemporal dynamics of dieback remain poorly understood. Using a time-series of satellite-derived canopy condition from Sentinel-2 imagery between 2017–2025 combined with clustering analysis, we mapped insect-induced dieback with high accuracy (F1-score = 0.80; precision = 1.00; recall = 0.67). We estimated that 11.3% of the subalpine woodlands declined during this period, with a higher rate in long-unburnt areas (21.0%) compared to areas affected by the 2003 fire (6.7%). Excluding areas burnt in the 2019–2020 fires, forests exhibiting decline had significantly greater mean canopy height than other clusters (p < 0.05), suggesting that mature stands are more susceptible to insect-induced dieback. Additionally, 95% of more recent decline occurred in close proximity (distance < 662 m) of earlier decline areas, indicating strong spatial correlation. This study provides a robust estimate of the area affected by insect-induced dieback in high-elevation subalpine woodlands and offers insights into the spatial and temporal patterns of its occurrence.

Improved long term monitoring of ecosystem function is desperately needed to understand the impacts of environmental change and land management on Australian ecosystems. Realtime sensor networks are invaluable for monitoring plant function, providing continuous and remote measurement of plant growth, water use, and water stress. These systems deliver high-resolution data on stem growth, soil moisture, transpiration, plant water potential and microclimatic variations, enabling the greater insight into plant response to stress events than is possible with snapshots from manual measurements at discrete timepoints. In the face of extreme climate events, such networks help identify early signs of stress, supporting timely management actions to enhance resilience. Fine grained data also contribute to understanding how plant communities are impacted by environmental stress and are essential to improve process-based models and machine leaning approaches aimed at predicting the response of plants to major disturbance events. We are establishing the Blue Mountains Hub for Ecology and Conservation, a smart sensor network in forest and woodland sites across a broad environmental gradient in the Greater Blue Mountains World Heritage Area. In this seminar, I will describe the structure of the network, target variables, sensor types, and wireless interface.

Significant shifts in vegetation patterns worldwide as a response to global changes underscore the urgent need for effective monitoring. Rapid detection of changes in vegetation health is required for timely responses and to facilitate the development of conservation strategies. Remote sensing allows brisk data collection over large spatial scales and is increasingly used to detect changes in vegetation cover and health by gathering detailed information about vegetation structure, composition, and condition.
There is increasing documentation of the vulnerability of eucalypt vegetation to global change in the form of canopy dieback occurrences across Australia. Exploring various methods and techniques used in remote sensing will increase our knowledge of dieback across the landscape. Using aerial imagery, this study combined the detailed spectral information from hyperspectral imagery and the structural information from the LiDAR imagery in a fusion process monitoring a population of Eucalyptus macroryhncha vulnerable to drought-related dieback. We found 28% of the overall canopy had complex LiDAR returns and high greenness (NDVI > 0.6), indicating relatively healthy stands. However, 26% of the canopy included areas with less complex LiDAR returns and moderate greenness (NDVI < 0.6), indicating areas likely to be suffering from dieback. Fusing spectral and structural information, can therefore provide clearer indicators of where dieback is occurring and lead to clearer management recommendations for this vulnerable population.

Vegetation die-off events are being reported with increasing frequency across all continents and ecosystem types. In Australia, die-off events have occurred at least once in all states in the past ~6 years. The 2023-2024 warm season in Western Australia generated a die-off footprint stretching over 1000 km and spanning myriad ecosystems. In virtually all systems, die-off was intense yet patchy varying by ecosystem, site, and species, often with abrupt boundaries. During and following the die-off event, we quantified extent, location, site attributes, and species impacted across SWWA. Here, we report specific impacts from locations with longitudinal data combined with extensive data to understand the complex interactions of geology, climate, fire history, and hydrology leading to enhanced vulnerability of vegetation to die-off. These complex, often contingent responses to extreme climate events are critical to preparing for the future and identifying mitigation and response options. Altered fire rotations, exclusion of sites from burning, real time changes to management, and species specific responses are examples of nuanced, bespoke options worthy of development. The 2023-2024 season left decision makers and managers reeling, underscoring the urgency of studying the event to prepare for the future.