Soil carbon ages are biased: ancient radiocarbon-free carbon inflates mean residence times (MRT). By estimating its contribution across soil types, parent material & depth, we show how to correct this bias and refine global soil carbon projections.
Soils store a large amount of carbon and play a key role in mitigating climate change. However, their actual capacity to accumulate carbon remains uncertain: climate models predict a strong soil carbon sink by 2100, while radiocarbon-based studies often suggest a much more limited potential.
Part of this discrepancy is due to the presence of very old carbon, derived from rocks or fossilized organic matter. This “radiocarbon-free” carbon no longer takes part in current soil processes but makes soils appear much older than they really are.
By analyzing 313 soils worldwide, we quantified the proportion of this ancient carbon depending on parent material, soil type, and depth. On average, soils contain 2.4 mg/g (± 3.2 SD) of such carbon, representing up to 11% of organic carbon in topsoils (0–30 cm), 25% in subsoils (30–100 cm), and more than half in deep layers (>100 cm).
After removing this bias, we estimate the mean age of the truly active soil carbon at about 290 years, instead of several millennia as previously reported. This much younger estimate agrees better with independent isotopic evidence and helps refine climate models.
Reference: Copard Y., Hatté C., Cécillon L., Barré P., Chenu C., Cornu S., 2025. Soil carbon dynamics reshaped by ancient carbon quantification. Global Change Biology, 31:e70482 – doi: 10.1111/gcb.70482