Boreal forests are crucial components of the global carbon cycle. These ecosystems act as substantial carbon sinks, storing large quantities of carbon in both plant biomass and soil, particularly in the organic layers found in peatlands. In Finland, a large proportion of boreal forests is under some form of management, typically focused on wood production. In the last decade, the preferred forest management approach in Finland has been rotation forestry. In this system, forests are periodically thinned before eventually being clear-cut and replanted. More recently, however, continuous cover forestry (CCF), which relies on selective harvest while maintaining permanent forest cover, has gained attention as a more sustainable alternative.
Thinning is an integral silvicultural practice within both rotation forestry and CCF. This practice involves the removal of individual trees from a forest stand to harvest wood biomass and improve growing conditions for the remaining trees, by for example increasing the availability of water, nutrients, and light. Although thinning is known to enhance the growth of the remaining trees over time, its effects on the carbon budget of different forest components, such as soil and ground vegetation, depend strongly on stand conditions.
To better understand these effects, we studied two Finnish forests with different soil conditions, thinning intensities, and dominant tree species. The first site, located at the Hyytiälä Forest Station, is a Scots pine dominated upland forest, established on mineral soil in the 1960’s. It was commercially thinned in 2020, removing 40% of the tree biomass. The second site, Ränskälänkorpi, is a drained peatland forest dominated by Norway spruce. It was drained before 1960 for forestry purposes, and it was intensively thinned in 2021 with the aim of applying CCF, removing 75% of tree biomass, primarily from dominant trees.
Using field measurements, tree allometric equations, and soil carbon flux measurements we examined how thinning affected carbon accumulation rates in tree stems, bark, branches, foliage, coarse and fine roots, and forest floor vegetation, as well as carbon emissions from soil respiration and litter decomposition in both the upland forest (Hyytiälä) and drained peatland forest (Ränskälänkorpi). These measurements allowed us to estimate the forests' net ecosystem productivity (NEP), which represents the balance between carbon sequestered and carbon released each year.
Our results showed that thinning had significant effects on several forest components from soil processes to how tree responded to the intervention (de Quesada 2026), which impacted the balance of the NEP. After thinning, both forests experienced a reduction in annual carbon accumulation rates after thinning, mainly due to the removal of individuals that were previously storing carbon. At the same time, carbon emissions increased as branches, needles, and other harvesting residues left on the forest floor began to decompose
A year after thinning, the upland forest experienced a recovery in the annual carbon accumulation rates of the aboveground tree components (stem, foliage, bark, branches), and a significant increase in carbon accumulation in forest floor vegetation, as more sunlight reached the ground, promoting the growth of mosses and vascular plants. The belowground components (coarse and fine roots, and stump) took an additional year to experience a recovery in their annual carbon accumulation rates. Carbon emissions slightly decreased compared to the year of the thinning, while soil heterotrophic respiration remained unchanged.
In the drained peatland forest, the annual carbon uptake did not experience any significant recovery a year after the intervention. However, we also observed a significant increase in the forest floor vegetation carbon accumulation rates. As in the upland forest, carbon emissions from harvest residue decomposition slightly decreased one year after thinning, with soil heterotrophic respiration remaining constant.
These changes had important consequences for the forests' overall carbon balance. In the upland forest, thinning temporarily shifted the ecosystem’s NEP from a carbon sink to a carbon source, meaning it released more carbon than it absorbed. However, the forest recovered within a year and returned to being a net carbon sink, although carbon uptake remained below pre-thinning levels. The drained peatland forest started from a different situation. Even before thinning, it was already releasing more carbon than it absorbed. Thinning further increased this carbon loss, and only limited recovery was observed during the first year after the intervention.
In the HIKET project, we aim to develop knowledge that supports climate-sustainable land-use planning and management. Our findings highlight that the short-term effects of thinning depend on forest type, soil conditions, and thinning intensity. Understanding these short-term changes is essential for developing forest management practices that balance sustainability with wood production. These findings also support HIKET project’ efforts to reduce uncertainties in peatland soil emissions and improve soil models. However, further research, particularly in drained peatland forests, is needed to better understand how these complex ecosystems respond to management over time. Therefore, expanding the current measurement network is essential to better represent a wider range of forest stands and to improve understanding of their responses to management.
Figure 3. Net ecosystem production (NEP) of Hyytiälä upland forest.
Positive values (orange) represent carbon uptake (carbon sink) in the forest ecosystem, while negative values (magenta) represent carbon emissions (carbon source). The dotted line indicates the timing of thinning. Error bars represent the standard error (SE) of NEP for each year. Modified from figure 7 in de Quesada et al. 2025.
Figure 4. Net ecosystem production (NEP) of Ränskälänkorpi drained peatland forest, showing that the forest was a net carbon source before thinning and remained a net carbon source after the intervention.
The dotted line indicates the timing of thinning. Error bars represent the standard error (SE) of NEP for each year. Modified from figure 8 in de Quesada et al. 2025.