The effect of climate change on the health and productivity of Australia’s temperate eucalypt forests

The release of the Sixth Assessment Report of the Intergovernmental Panel on Climate Change (IPCC 2021) is a galvanising moment globally for concerted action to stem climate change. The report provides stronger evidence of previously identified changes to Earth systems as a result of anthropogenic greenhouse-gas emissions. If current trends for rising emissions continue beyond mid-century, the global mean temperature will be 3.3–5.7 degrees warmer by the end of this century compared with the period 1850–1900. Warming across Australia is predicted to be higher than the global average (Australian Academy of Science 2021). The Sixth Assessment Report (IPCC 2021) also gives greater prominence to extreme events, particularly heatwaves, fire weather and compound events (heatwaves coupled with droughts). Although temperature increases drive most of the extreme events that will affect temperate eucalypt forests, a trend of increasing aridity across southern Australia due to climate change (CSIRO and Bureau of Meteorology 2020) has amplified these compound events (Matusick et al. 2018) and fire risk (van Oldenborgh et al. 2021). Despite more confident predictions of future climate states under different emissions scenarios, the impacts of climate change on the productivity of temperate eucalypt forest remain uncertain. This is because of a lack of understanding about how changing temperature and moisture conditions will affect tree establishment, growth rates and mortality due to extreme events and disturbances (e.g. fire, pests and diseases). There is also considerable uncertainty about the effects of increasing atmospheric concentrations of carbon dioxide (CO2). Roxburgh et al. (2004) concluded that existing simulation models used in Australia have insufficient capability to predict the consequences of climate change for Australia’s forests. Haverd et al. (2013) made considerable advances in reducing the uncertainty of continent-scale predictions of net primary productivity by using multiple observation types to constrain BIOS2 model predictions. Two models, CABALA and 3-PG, have been used extensively in Australia to predict forest productivity under simulated future climates (Pinkard, Bruce et al. 2014; Battaglia and Bruce 2017; Wang et al. 2021). Calibration studies show that both 3-PG (Potithep and Yasuoka 2011) and CABALA (Battaglia et al. 2004; Battaglia et al. 2011) can accurately predict forest productivity when growing conditions can be defined, although 3-PG underpredicts productivity in temperate eucalypt forests, particularly the most productive of these (Haverd et al. 2013; Volkova et al. 2015). Predictions of productivity in future climates under different emissions scenarios made using CABALA and 3-PG vary widely, however, because of uncertainties surrounding climate projections, responses to higher concentrations of atmospheric CO2 (Battaglia and Bruce 2017; Wang et al. 2021) and mortality from drought (Pinkard, Bruce et al. 2014). Models need to evolve to enable more confident predictions of productivity in Australian temperate eucalypt forests under future climate scenarios. Our understanding of the key environmental drivers of productivity in eucalypt native forests has progressed in the last decade. Using independent methods, growth rates of eucalypt forests have been found to have a consistent relationship with temperature – a convex parabola with the vertex corresponding to the temperature optimum at which maximum growth occurs. Bowman et al. (2014), using an inventory approach, found an optimum (mean annual) temperature for diameter growth rates of 11°C in eucalypt forests growing on mesic sites in temperate and subtropical regions of Australia. Using data from a network of eddy covariance towers in eucalypt forests and woodlands, Bennett et al. (2021) found that temperature optima for gross primary productivity (GPP) in southern Australia ranged between 15°C and 23°C and were directly related to the historical climate of the site. There is great potential to make more use of Australia’s network of eddy covariance sites in temperate eucalypt forests (potentially extending them to eucalypt plantations) to provide observational data that could be used to constrain growth model predictions (sensu Haverd et al. 2013) and to fieldtest model parameters (sensu Potithep and Yasuoka 2011). Most eucalypt forests have some capacity to maintain productivity with a certain amount of warming because their optimum temperatures for GPP are above their longterm average temperatures (Bennett et al. 2021) and they have potential to acclimate to short-term (seasonal) temperature fluctuations (Zhu et al. 2018; Zhu et al. 2021). However, eucalypt forests are unable to maintain productivity during more extreme warming events, such as heatwaves. Our understanding of how eucalypt forests respond to heatwaves has advanced considerably in the last five years, and three main patterns of response have been seen: