Tropical forests currently store more than 60% of the world’s vegetation biomass, making them one of the most critical regulators of the global carbon cycle. Yet new research suggests the Amazon’s ability to retain that carbon may weaken faster than previously understood, not primarily because of declining tree growth, but because climate-driven mortality is accelerating the turnover of forest biomass.
A study published in Nature Climate Change by researchers from South China Botanical Garden, in collaboration with Cornell University and other international institutions, found that atmospheric drying and increasing convective storm activity are likely to shorten carbon residence times across Amazonian forests during this century.
The distinction is significant because much of the existing climate literature has focused heavily on tropical forest productivity, measuring how much carbon forests absorb through photosynthesis and biomass growth. Far less attention has been paid to biomass turnover, the rate at which stored carbon returns to the atmosphere through tree mortality and decomposition.
That gap matters because a forest can remain highly productive while simultaneously losing long-term carbon storage capacity if trees die more frequently or biomass cycles more rapidly through the ecosystem.
The study estimates that biomass carbon turnover time in Amazonian forests could decline by approximately 3% under a low-emissions climate scenario by the end of the century. Under a high-emissions pathway, the decline could reach 15%, materially reducing the forest’s ability to function as a stable long-term carbon sink.
Researchers identified convective storms as a particularly influential driver of biomass turnover. These storms, typically characterized by intense rainfall, lightning, and strong wind events over short durations, were found to exert greater influence on biomass carbon turnover than drought stress indicators alone.
That finding complicates the dominant narrative surrounding Amazon forest degradation, which has often emphasized drought and heat stress as the principal climate threats. While those factors remain important, the study suggests that extreme storm dynamics may play a larger role in tree mortality and biomass destabilization than previously incorporated into large-scale climate models.
The implications extend beyond regional ecology because the Amazon rainforest remains deeply interconnected with global climate regulation. Carbon residence time determines how long atmospheric carbon remains locked within vegetation before returning to the atmosphere. Shorter residence times weaken the buffering role forests play against rising greenhouse gas concentrations.
One of the study’s key methodological contributions lies in how researchers approached the problem of scale. Previous analyses of tropical forest carbon cycling relied heavily on field plot observations, which provide high-quality localized measurements but often struggle to capture the enormous spatial heterogeneity of Amazonian ecosystems.
According to corresponding author Wu Donghai, limited field plots cannot fully represent the environmental variability and ecological complexity across the Amazon basin. To address this limitation, researchers combined long-term forest plot observations with satellite remote sensing data to estimate tree mortality patterns across Amazonian forests at regional scale. They then applied interpretable machine learning models within a non-equilibrium carbon cycle framework to map biomass carbon turnover time and identify environmental drivers.
The results revealed strong nonlinear relationships between environmental conditions and turnover dynamics. In practical terms, this means that forest responses to climate stress may not progress gradually or proportionally. Certain climatic thresholds could trigger disproportionately large increases in tree mortality or declines in carbon retention capacity.
That possibility has important implications for Earth system modeling. Many global climate models still simplify vegetation dynamics or inadequately represent mortality-driven carbon cycling processes, particularly in tropical ecosystems where observational coverage remains incomplete.
The study also reinforces growing concerns that the Amazon’s ecological resilience may be increasingly vulnerable to interacting climate pressures rather than isolated stressors alone. Atmospheric drying, rising temperatures, altered precipitation patterns, and intensifying storm regimes do not operate independently. Their combined effects may accelerate forest instability in ways difficult to detect through productivity metrics alone.
This dynamic is particularly relevant as portions of the Amazon already show signs of weakening carbon sink performance. Several studies over the past decade have suggested that some areas of the forest are absorbing less carbon than in previous decades, while others may already be transitioning toward net emissions due to deforestation, degradation, and climate stress.
The new findings shift further attention toward mortality processes as a central component of future forest carbon dynamics. While reforestation and conservation efforts continue to focus heavily on increasing biomass accumulation, the stability and longevity of stored carbon may become equally important indicators of ecosystem resilience under intensifying climate conditions.
