In the face of escalating global temperatures and climate extremes, understanding how forests respond to these changes is critical for projecting the future of Earth’s carbon cycle. The year 2023 marked an unprecedented moment in climate history, being the hottest year on record globally, with particularly severe heat and drought conditions gripping Canada’s expansive boreal forests. These extreme environmental stressors brought forward a unique opportunity for scientists to investigate how such climatic anomalies influence net forest carbon uptake, a crucial component of global carbon balance and climate regulation.
Recent research combining satellite-based atmospheric CO₂ flux inversions with ground-based in situ CO₂ flux and concentration measurements has unveiled surprising insights into Canada’s forest carbon dynamics during the anomalously warm 2023. Contrary to expectations that extreme heat and drought would suppress forest carbon uptake by stressing vegetation and amplifying respiratory carbon release, the study revealed a significant enhancement in net carbon sequestration across Canadian boreal forests during this period. This counterintuitive result fundamentally challenges prevailing assumptions about ecosystem responses under climate extremes.
The study quantified an increase of 0.28 ± 0.23 petagrams of carbon (PgC) in net forest carbon uptake in 2023 compared with the 2015–2022 baseline. This uptick in carbon capture was substantial enough to offset between 38 and 48 percent of the carbon emissions released from wildfires that ravaged parts of Canada that same year. Wildfires, stimulated and exacerbated by heat and drought, have traditionally been associated with massive carbon releases, adding perturbations to the atmospheric carbon budget. Thus, the observation that forests simultaneously increased carbon absorption rates is striking and prompts a deeper exploration of underlying mechanisms.
Key to this enhanced net carbon uptake was a pronounced reduction in ecosystem respiration, a process through which carbon stored in plant and microbial biomass is released back into the atmosphere. The research identified severe root-zone soil moisture deficits as a primary driver behind suppressed respiration rates during the heat-drought conditions. Dry soils limit microbial activity and reduce root respiration, both of which contribute substantially to ecosystem carbon release. Moreover, the respiration response exhibited a unimodal temperature dependency—meaning there is an optimal temperature range for respiration beyond which rates decline—thus extreme heat effectively curtailed carbon efflux.
This nuanced understanding elaborates that during heat extremes, while photosynthetic carbon uptake may face constraints, the disproportionately large decrease in respiration creates a net positive carbon balance. Essentially, forests become carbon sinks not just through enhanced carbon fixation, but critically through dampened carbon loss pathways. This challenges traditional carbon cycle models which often consider respiration to linearly increase with temperature, underscoring the complexity of biosphere-atmosphere carbon exchange under non-equilibrium climate conditions.
Intriguingly, most existing dynamic global vegetation models (DGVMs), which simulate terrestrial ecosystem responses to climate variables, failed to accurately capture these respiration downturns and the complex interplay of hydrothermal stressors observed in 2023. This modelling shortfall highlights an urgent imperative for revising ecosystem carbon feedback frameworks within climate models to incorporate such nonlinear and threshold responses, ensuring better predictions of forest carbon dynamics under future warming scenarios.
The integration of satellite atmospheric CO₂ inversion data with meticulous ground-based flux measurements was pivotal in resolving this complex picture. Satellite data provided large-scale insights into atmospheric carbon variability, while in situ sensors allowed precise temporal resolution of carbon flux components in forest ecosystems. Together, this multi-disciplinary approach bridged observational scales and deepened understanding of how boreal forests mediate carbon exchange amid extreme climatic perturbations.
Importantly, this research also speaks to the resilience and adaptive capacity of Canada’s boreal forests. Despite exposure to unprecedented heatwaves, severe drought stress, and widespread wildfire, these ecosystems exhibited a capacity to modulate carbon release and uptake, potentially buffering climate feedback loops that could otherwise accelerate atmospheric CO₂ accumulation. This resilience, however, should not be misconstrued as immunity; rather, it highlights critical physiological and ecological processes that warrant further investigation to ensure forest management and climate mitigation efforts account for nuanced ecosystem responses.
As the frequency and intensity of heat extremes and drought episodes escalate under global warming, elucidating carbon cycle feedbacks from forested landscapes is increasingly vital for global climate policy. This study’s findings urge a reframing of assumptions about boreal forest carbon dynamics, advocating for enhanced monitoring infrastructure and improved predictive models that better mirror real-world responses. Bridging this gap can help refine mitigation pathways targeting terrestrial carbon sinks as natural allies in combating climate change.
Moreover, the observed disconnect between wildfire emissions and net forest carbon uptake accentuates the complexities inherent in ecosystem carbon budgets during climate extremes. While wildfires undeniably release large carbon quantities, the compensatory carbon uptake mechanisms observed may transitorily offset these losses. Understanding the temporal persistence and spatial heterogeneity of such compensations will be crucial in assessing the net climatic impact of fire-prone boreal regions under future warming regimes.
Future research opportunities abound in decoding the physiological underpinnings of respiration suppression under heat-drought stress, particularly through exploring root-soil-microbe interactions, hydraulic limitations, and plant metabolic adjustments. Integrating these mechanistic insights into dynamic global vegetation models could significantly reduce uncertainties in projecting terrestrial carbon fluxes amid accelerating climate variability.
Equally, expanding observational networks to incorporate additional biophysical measurements and expanding satellite remote sensing capabilities will enhance our capacity to detect and quantify ecosystem feedbacks in near real-time. These advances will facilitate adaptive forest management strategies that optimize carbon retention potential while mitigating climate-related disturbance risks.
In summary, the 2023 heatwave and drought event across Canadian boreal forests have provided a natural laboratory for uncovering divergent carbon cycle trajectories under unprecedented climate stress. The enhanced net forest CO₂ uptake driven by respiration declines amidst extreme hydrothermal conditions marks a paradigm-shifting insight with broad implications. It challenges existing ecosystem-climate models, underscores the complexity of biosphere feedbacks, and charts a path forward for integrating realistic vegetation responses into climate mitigation frameworks.
To weather the climatic storms ahead, incorporating these emergent findings into global carbon budgets and policy decisions will be paramount. As boreal forests serve as one of the Earth’s largest terrestrial carbon reservoirs and regulators, unlocking their climate resilience potential could prove pivotal in sustaining Earth’s carbon balance in an uncertain warming future.
Subject of Research: Boreal forest carbon dynamics and their response to heat and drought extremes
Article Title: Canadian net forest CO₂ uptake enhanced by heat drought via reduced respiration
Article References:
Dong, G., Jiang, F., Zhang, Y. et al. Canadian net forest CO₂ uptake enhanced by heat drought via reduced respiration. Nat. Geosci. (2026). https://doi.org/10.1038/s41561-025-01875-1
DOI: https://doi.org/10.1038/s41561-025-01875-1
