+ de critères

Are forest soils a carbon sink?

Mathieu Jonard (Catholic University of Louvain) takes the floor

The role of forest soils in carbon sequestration: measurements, findings and explanatory hypotheses

Summary

Soils play an important role in climate regulation. They store considerable amounts of carbon in the form of organic matter, which, if increased by only 4/1000 a year, could possible stop the increase in atmospheric CO2 concentrations. Inversely, decreasing organic matter in the soil could accelerate climate change.

Many countries, including France, have shown their commitment to reducing greenhouse gas (GHG) emissions by signing the Kyoto protocol, and more recently the 2015 Paris Agreement, in the framework of international negotiations on climate change. To prove that they are honouring their commitments, signatory countries must track their emissions. For the 2008-2012 period, France committed to reducing its GHG emissions by 5% compared to 1990 levels and chose to incorporate forestry in their calculations, based on the assumption (unproven to date) that forest soils were either neutral or positive carbon sinks.

In this light, results from the first repeated soil samples taken at a 15-year interval on the 102 RENECOFOR plots were analysed. The objective of the study, supported by the Ministry in charge of Agriculture, aimed to detect and quantify the changes that had taken place in the organic carbon stocks found in the forest soils and litter. Two approaches were used to try to understand the causes of the observed changes: a statistical procedure to select explanatory factors, and an estimated carbon status report based on incoming and outgoing fluxes.

Soil data were collected during two sampling campaigns, one between 1993 and 1995, and the second between 2007 and 2012. On each plot, five square sub-plots were set up (one in the centre and four near the edges of the plot). Each sub-plot was further divided using an overlay grid to determine 16 nodes, five of which were chosen for sampling to guarantee good spatial distribution. For each soil layer, a composite sample was created by combining the soil in that layer from the five sub-plot samples. This sampling design made it possible to quantify intra-plot variability and to reveal possible changes over time at the plot scale. Depending on the sampling campaign and the plot, soil samples contained four to six layers. Up to three layers were described for forest floor litter; then the underlying mineral soil was sampled in successive pre-determined layers: 0-10cm, 10-20cm, and 20-40cm.

Laboratory analysis methods were the same for both sampling campaigns. Organic carbon content was determined by dry combustion (after removing carbonates) for the litter layers and the first mineral layer (0-10 cm), and by the Anne method for the other mineral layers (10-20 cm and 20-40 cm). Total carbon content for the litter layers was calculated by multiplying the mass of each layer by its organic carbon content, then by adding the results for all the layers. For the mineral layers, organic carbon content was calculated according to the apparent density of the soil layer while the percentage of coarse elements was accounted for.

Since the time between the two sampling campaigns varied from one plot to another, statistical analyses were carried out on the differences in carbon stocks between the two campaigns proportionate to the time elapsed. Annual variations in carbon stocks were significant only for the litter layers, the 0-10 cm soil layer and total stocks (see Table below). The variations were positive, thus indicating that carbon sequestration took place, reaching 0.35 tC per ha per year, which, in terms of soil carbon sequestration, equates to around 4/1000. Assuming that forest soils throughout France have gone through comparable changes, this results in a carbon sink equal to approximately 5% of the GHG emissions due to the use of fossil fuels in France.

Soil organic carbon stock and its annual average variation, by layer. The changes are significant if the probability P is less than 0.05 (in bold).
Soil organic carbon stock and its annual average variation, by layer. The changes are significant if the probability P is less than 0.05 (in bold).

To identify the factors explaining the variability in carbon sequestration rates among the plots in the network, a selection methodology was applied to a series of 34 potential explanatory variables; two factors were highlighted: stand age and stand structure.
Soil carbon sequestration rates slow down as the stand ages and are higher in irregular than in regular (even-aged) stands. However, these two factors combined explain only 14% of the variability. This small proportion is due to high within-plot variations in soil carbon sequestration. In addition, stand age effect could have been confounded by a possible species effect or an effect due to how long the plot had been forested: in the data set, softwood stands were on average younger and, in terms of historic land use, appeared later than hardwood stands. These results do suggest, however, that forest management practices may have an effect on soil carbon sequestration.

To assess which processes might underlie forest soil carbon sequestration, a carbon balance of incoming and outgoing carbon was built for a virtual plot representative of the whole network. The balance was estimated for above-ground (litter) and below-ground (mineral layers) soil components and assumed that there was initially an equilibrium between carbon intake and release. The results indicate that the carbon sequestration rate found for the litter layers could be due to slowed decomposition resulting from a deterioration in the quality of the organic matter (increase in the carbon/nitrogen ratio).
For the mineral layers, the rate of sequestration estimated in the balance was much lower than actual observed levels, thus indicating that there was probably no state of equilibrium at the time of the first sampling campaign and that litter was building up faster than it could decompose. One wonders, for example, if a reforestation campaign similar to the one that began at the turn of the 19th century were to be undertaken today, whether stocks of organic material in the mineral soil would increase with the change-over from previous agricultural use.

Following this study, many questions remain. Will forest soils continue to store carbon over the long term? How stable is the newly stocked carbon? What are the processes underlying forest soil carbon sequestration?

Pour en savoir plus

Ressources