New breakthrough study by Ganzenmüller et al., published in August 2025 in the journal One Earth, reveals that human activity has led to depletion of global terrestrial carbon stocks by significant 24 %, which represents a deficit of 344 gigatons carbon (PgC). This depletion, which is comparable to all fossil fuel emissions over the past 50 years (361 PgC between 1975 and 2023), contributes significantly to global warming and highlights the profound human impact on Earth's ecosystems.

Understanding the terrestrial carbon deficit

The terrestrial carbon deficit is defined as the difference between the actual and potential carbon stocks in vegetation and soils. “Potential” refers to the hypothetical amount of carbon that would exist in the absence of direct human changes to the Earth’s surface under current environmental conditions. Vegetation and soils are key components of the global carbon cycle, storing more carbon than the atmosphere and all fossil fuel reserves combined. Human impact on these stocks began more than 10,000 years ago with agricultural practices and the use of fire, accelerating exponentially in recent centuries.

The study by Ganzenmüller et al. uses a combination of semi-empirical data and machine learning to estimate this deficit with an unprecedented resolution of approximately 1 km². They estimate that the actual global carbon stock in vegetation is 420 PgC and in soil (to a depth of 0–30 cm) 642 PgC. The potential carbon stock in vegetation ranges from 617 to 683 PgC and in soil from 679 to 723 PgC. The observed deficit is 234–344 PgC, while Carbon stocks in vegetation account for 76 % – 84 % of the total deficit, while the soil organic carbon deficit accounts for the remaining 16 %–24 %. Areas with the largest deficits include China, Brazil, Europe, and the United States.

Main causes of carbon depletion

The study identifies three main drivers of the terrestrial carbon deficit:

• Pasture expansion: Responsible for 30 % deficit.
• Expanding arable land: Responsible for 24 % deficit.
• Forest management: Responsible for 23 % of the deficit, including logging, afforestation and agroforestry. Urbanized areas contribute 3 % to the carbon deficit, while another 20 % is due to other unattributable causes. Cropland and grassland have the highest relative carbon deficits in vegetation per hectare. Specifically, the carbon density in vegetation on cropland is 81 % lower than the potential density and on grassland it is 73 % lower.

Underestimation by global vegetation models (DGVMs)

A critical finding is that dynamic global vegetation models (DGVMs), which are key to climate predictions, underestimate the terrestrial carbon deficit by an average of 37 % (range: 2 % – 58 %). This underestimation is particularly pronounced in all ecological zones except tropical rainforests. The reasons for these discrepancies are diverse and include differences in model structure, implementation of processes (such as nutrient cycling, CO2 fertilization, vegetation dynamics and natural disturbances), and especially in the insufficient consideration of anthropogenic processes such as forest management, crop rotation, artificial fires and cultivation. This underestimation means that DGVMs likely underestimate emissions from land-use change, which may indicate that the remaining carbon budget to limit global warming to 2°C may be exhausted sooner than previously thought.

Consequences and political implications

The study's findings are crucial for assessing anthropogenic impacts on ecosystems, refining global vegetation models, and effectively planning natural carbon removal. The enormous scale of human impact on terrestrial carbon stocks highlights the urgent need for effective climate change mitigation measures.

• Priority of protection and restoration: The study provides valuable insights for policymakers to prioritize efforts to protect and restore ecosystems. The concept of protecting "virgin" and vulnerable ecosystems is crucial because carbon stocks can take decades to centuries to recover, or may never be fully restored.
• Sustainable land use decision-making: The findings call for careful land use decisions, particularly regarding the expansion of cropland and pasture. Strategies such as increasing land use efficiency, minimizing food waste, and considering alternatives to bioenergy are suggested.
• The importance of dietary changes: Large areas of pasture emphasize the need to also focus on dietary changes.
• Improving climate models: The data provided offer new opportunities to refine estimates of carbon losses from land-use change in DGVMs, thereby reducing uncertainties in future climate projections.

Overall, the study confirms that depletion of terrestrial carbon stocks is almost as important an anthropogenic disturbance to the global carbon cycle as fossil fuel emissions, underscoring the need for a holistic approach to climate change mitigation that includes not only emissions reductions but also extensive efforts to restore Earth's ecosystems. JRi