Deforestation releases old carbon

Deep soil carbon in tropical catchments can be rapidly mobilized to rivers upon land-use change to agriculture, suggest analyses of dissolved organic carbon. Such carbon stocks had been thought stable for millennia.

Tropical forests hold about a third of the global soil organic carbon (SOC)1. When land-use changes from forests to agricultural lands, carbon stocks are reduced2. SOC below the top soils tends to get older with increasing depth and can be thousands of years old3. This deep, old SOC has been thought to be stable, but whether land-use change has an effect on this old carbon remains largely unknown. Writing in Nature Geoscience, Drake et al.4 studied dissolved organic carbon (DOC) in rivers of catchments with differing degrees of deforestation. Results indicate that land-use change led to mobilization of old organic matter to the river as DOC, which turned out to be more biodegradable than the DOC from pristine forests.

Globally, the organic carbon stored in soils amounts to more than the carbon in the atmosphere and biomass taken together. The stability of soil carbon stock is therefore important but has been under threat by land-use change. Deforestation results in a much lower input of plant litter, an important source of SOC. And it also leads to a greater loss of SOC in several ways2 (Fig. 1). The removal of plant cover is particularly problematic in tropical forests, because intensive rains lead to a high degree of erosion of unprotected soil2. Trees in the tropics protect the soil in two ways, by canopy interception — which slows down the flow of water through the system — and by roots keeping the soil particles together.

Fig. 1: Effects of land-use change from forest to agricultural land in a tropical hilly landscape.

Removal of trees exposes the soil to heavy rain, which increases the rate of surface run-off and soil erosion. SOM from those soils flows into rivers. Drake et al.4 found that DOC in rivers of pristine catchments, with 100% forest cover, was young and decomposed comparably slowly, whereas DOC from catchments with deforestation was old and decomposed comparably easily. Young carbon (newly photosynthesized) in green; old carbon in black; DOC, dissolved organic carbon; SOM, soil organic matter; DIC, dissolved inorganic carbon; POC, particulate organic carbon. Some riverine carbon may not reach the atmosphere, and can then be stored in downstream sediments9.

The effects of deforestation on forest SOC have been investigated mostly with a focus on the more carbon-rich soils in the upper 20–30 cm of depth2. However, despite a lower SOC concentration, similar amounts of SOC may be stored below this depth3. In addition, SOC below this depth changes very slowly and can be thousands of years old3. Compared to a global carbon cycle normally assessed over timescales of hundreds of years, this SOC can be considered permanently removed and stored. However, human disturbance may reintroduce such old SOC into the modern carbon cycle5.

Drake et al. study the impact of deforestation and subsequent conversion to agricultural lands on carbon losses from tropical soils via riverine transport in the rift valley of the Congo Basin, near Lake Kivu. They analyse radiocarbon age (Δ14C), stable carbon isotopes (δ13C) and molecular composition of the dissolved organic matter (DOM) in 19 rivers from catchments with 0 to 100% deforestation. All three independent data types suggest that DOC from forested catchments was derived from modern plant litter, whereas the fully deforested catchments had a DOC age of ~1.5 kyr. The δ13C signature and molecular composition suggest that the SOC input of this old DOC is dominantly from soil depths of 50 cm or below.

DOM from deforested catchments contained smaller molecules, compared to that from pristine forests, and a composition that indicates it to be more bioavailable and of a microbial origin. Indeed, incubation experiments showed that this DOM was more biodegradable compared to the modern DOM from the forested areas. SOC that is closely attached onto particles in soils and sediments can prevent microbial degradation; but if particle-bound carbon is dissolved into water as DOC, it can be rapidly mineralized6,7. Hence, mobilization of SOM into water as DOC means that this carbon can potentially reach the atmosphere as CO2, and a carbon sink then becomes a greenhouse gas source.

This study is based on a single sampling campaign, given the costs of 14C analyses and difficulties in sampling in the Congo Basin. Yet, surface run-off and soil erosion as well as the form of carbon that transports into streams vary considerably over time8. Other pathways, such as CO2-release directly from land via soil respiration, and the lateral export of other carbon forms, such as dissolved inorganic carbon and particulate organic carbon, are also likely to play a role. Further studies are therefore needed to assess the relative importance of the proposed DOC pathway for SOC losses. Additionally, after quantifying SOC losses from all possible routes, it would also be important to track the ultimate fate of the exported carbon: emission to the atmosphere or deposition and storage in lake or sea sediments.

Drake and co-authors conclude that deforestation can result in a rapid loss of surface soils that have accumulated over hundreds to thousands of years in the Congo Basin. In order to reduce further losses of soil carbon, they also propose modifications to the agricultural techniques presently used, and stress that the retention of the biodegradable SOC that is currently stabilized and isolated from the modern carbon cycle is ultimately dependent on forested land remaining forested.


  1. 1.

    Köchy, M., Hiederer, R. & Freibauer, A. SOIL 1, 351–365 (2015).

    Article  Google Scholar 

  2. 2.

    Don, A., Schumacher, J. & Freibauer, A. Glob. Chang. Biol. 17, 1658–1670 (2011).

    Article  Google Scholar 

  3. 3.

    Trumbore, S. E., Davidson, E. A., Barbosa de Camargo, P., Nepstad, D. C. & Martinelli, L. A. Glob. Biochem. Cycles 9, 515–528 (1995).

    Article  Google Scholar 

  4. 4.

    Drake T. W. et al. Nat. Geosci. https://doi.org/10.1038/s41561-019-0384-9 (2019).

  5. 5.

    Butman, D. E., Wilson, H. F., Barnes, R. T., Xenopoulos, M. A. & Raymond, P. A. Nat. Geosci. 8, 112–116 (2015).

    Article  Google Scholar 

  6. 6.

    Hedges, J. I. & Keil, R. G. Mar. Chem. 49, 81–115 (1995).

    Article  Google Scholar 

  7. 7.

    Cole, J. J. & Caraco, N. F. Mar. Freshwater Res. 52, 101–110 (2001).

    Article  Google Scholar 

  8. 8.

    Johnson, M. S. et al. Hydrol. Proc. 20, 2599–2614 (2006).

    Article  Google Scholar 

  9. 9.

    Drake, T. W., Raymond, P. A. & Spencer, R. G. M. Limnol. Oceanogr. Lett. 3, 132–142 (2018).

    Article  Google Scholar 

Download references

Author information



Corresponding authors

Correspondence to Alf Ekblad or David Bastviken.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Ekblad, A., Bastviken, D. Deforestation releases old carbon. Nat. Geosci. 12, 499–500 (2019). https://doi.org/10.1038/s41561-019-0394-7

Download citation


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing