Drained forest peat soils emit less than assumed in cold-summer climate zone, synthesis finds

Forest soils on drained peat in the cold, humid Dfb climate zone of northeastern Europe emit far less carbon dioxide than current international default values suggest, according to a synthesis of greenhouse gas measurements published in Biogeosciences. A team led by Aldis Butlers of the Latvian State Forest Research Institute compiled 297 annualised flux values from 131 forest sites with organic soils, drawing on measurements from the Baltic States together with aggregated site data from a previous synthesis that overlapped the same climate zone. The work was designed to produce emission factors tailored to the Köppen–Geiger Dfb zone, which is defined by cold winters, warm summers and no dry season, and which covers the Baltic States along with southern parts of Finland and Sweden.

The headline result concerns carbon dioxide. Across drained sites, mean net soil CO2 emissions were 0.22 ± 0.18 tonnes of CO2-carbon per hectare per year, while undrained sites averaged −0.60 ± 0.37 tonnes, meaning the undrained soils acted as net sinks on average. For comparison, the Intergovernmental Panel on Climate Change default emission factor for drained organic soils in temperate forest land stands at 2.60 tonnes of CO2-C per hectare per year, a figure built from only eight study sites. The new synthesis drew on 81 study sites for its drained-soil CO2 estimate, and the resulting value of 0.14 tonnes per hectare per year for Latvia’s drained forest organic soils, after area weighting, is substantially lower than both the IPCC default and the 0.52 tonnes per hectare per year currently used in Latvia’s national greenhouse gas inventory.

The synthesis also challenges a widespread assumption about drainage. On average, emissions from drained forest organic soils in this zone were not substantially higher than those from undrained soils, a finding that runs counter to the expectation that drained areas consistently produce elevated emissions. Undrained soils did skew toward acting as CO2 sinks, with the strongest mean removals observed in birch stands at −1.20 ± 0.76 tonnes of CO2-C per hectare per year, but soils at some undrained sites across all dominant species categories still acted as sources. The authors caution that the net drainage impact on total greenhouse gas emissions remains unknown, because undrained-soil estimates carry uncertainties several times higher, driven largely by highly variable methane and nitrous oxide fluxes.

Stratifying the data by dominant tree species revealed patterns the authors consider ecologically meaningful. Drained coniferous stands, split roughly evenly between Scots pine and Norway spruce in the dataset, were roughly carbon neutral, emitting 0.03 ± 0.55 tonnes of CO2-C per hectare per year, while drained deciduous stands, which occurred only on nutrient-rich sites, emitted 0.47 ± 0.29 tonnes. Drained nutrient-rich soils overall were a source of 0.51 ± 0.17 tonnes per hectare per year, whereas drained nutrient-moderate sites, dominated by Scots pine in 80 percent of observations, showed a tendency toward lower fluxes with a mean of −1.13 ± 0.52 tonnes, effectively a sink. In mixed-effects models, nutrient status explained about 13 percent of the variance in net CO2 fluxes, while species type accounted for only about 3 percent within nutrient-rich sites, and no statistically significant differences in CO2 fluxes were detected among species groups when stratified by drainage and nutrient status.

Stand age emerged as the strongest stand-level influence on the CO2 balance. Soils tended to shift from a source to a sink beyond approximately 25 years of stand age, and the dataset showed a moderate correlation of r = 0.41 between stand age and soil carbon balance, suggesting an increasing sink potential of about 0.02 tonnes of carbon per hectare per year with subsequent aging. Younger stands averaged a source of 0.61 ± 0.44 tonnes of CO2-C per hectare per year, and clearcuts showed mean fluxes of 0.90 ± 0.28 tonnes, well above the overall drained-soil mean. Because the flux data with stand age information had a mean age of 56 years and an interquartile range of 31 to 73 years, the authors warn that national inventories risk underestimating emissions if the elevated fluxes from clearcuts and young stands are not properly accounted for.

Methane told a different story, one dominated by water rather than trees. A linear regression between mean water table level and annualised methane fluxes identified a threshold of about 20 centimetres below the surface separating sources from sinks, with methane uptake increasing by 0.102 kilograms of CH4 per hectare per year for each additional centimetre of water table drawdown. Sites with mean water tables below that threshold averaged −3.91 ± 0.41 kilograms per hectare per year, while wetter sites above it averaged 70.88 ± 44.30 kilograms. Apparent species effects dissolved once water levels were considered: alder stands in drained sites appeared as large methane sources because their mean water table sat only 16 centimetres below the surface, and when a single methane emission hotspot in undrained alder stands was excluded, mean undrained emissions fell from 178.2 to 12.8 kilograms per hectare per year. The authors therefore do not consider dominant tree species a suitable single proxy for methane upscaling and recommend relying on water table data, or on mean fluxes by drainage status where spatial water table information is unavailable.

Nitrous oxide showed its own distinctive pattern. Excess emissions appeared in deciduous stands regardless of drainage conditions, though more pronouncedly under drained conditions, with mean fluxes of 5.58 ± 1.87 kilograms of N2O per hectare per year in drained deciduous sites and 3.51 ± 1.35 kilograms in undrained sites, the latter on average 37 percent lower. Conifer stands emitted on average 1.58 ± 0.35 kilograms per hectare per year. Statistical differences among the deciduous stands could not be identified because of outlier fluxes, and the authors note that high inputs of labile litter and nitrogen in deciduous stands could stimulate microbial N2O production even under wet conditions, although they stress that more data are needed to confirm this mechanism.

To demonstrate how the emission factors would be applied, the team used Latvia as a case study. Drained organic soils cover an estimated 529,000 hectares there, about 8 percent of the country, of which roughly 314,000 hectares, or 59 percent, is forest land. The choice of stratification materially changed the national estimate: the area-weighted CO2 emission factor for drained forest organic soils came to 0.14 tonnes of CO2-C per hectare per year when species-based factors were applied, rising to 0.23 tonnes if a single factor were used for all deciduous stands, and swinging to −0.71 tonnes if a single nutrient-based factor were applied, because nutrient-rich organic soils are underrepresented in Latvia’s area distribution. For nitrous oxide, species-type-specific factors yielded a weighted 3.80 kilograms per hectare per year, versus 2.82 kilograms for a single value. Without country-specific factors, Latvia would otherwise rely on the IPCC default of 2.60 tonnes per hectare per year or on regional temperate values of 1.61 and 1.46 tonnes derived from just 19 sites, which the authors consider poorly representative of Latvian conditions because the temperate zone extends to considerably warmer and drier parts of Europe.

The synthesis rests on careful data handling. The compiled fluxes came from peer-reviewed studies using the chamber method, with values either taken directly from a previous harmonised synthesis, which had assigned reliability weights to individual estimates, or verified as annualised net soil fluxes and given a weight of 1.0. Nutrient classifications were harmonised across datasets, though the authors acknowledge that different national forest site-type systems may not yield directly comparable indicators of nutrient status; nine of ten data points for nutrient-rich Scots pine sites came from the earlier synthesis, while such stands account for only about 4 percent of drained forest organic soils in Latvia. Soil property data were available for 93 percent of sites at least in part, and stand age was the most frequently reported stand parameter, available for 43 percent of sites.

The authors frame the work as empirical support for refining greenhouse gas accounting for forest organic soils, while acknowledging its limits. Currently available data do not allow robust quantification of drainage impacts on the total greenhouse gas balance, owing to residual flux variability that cannot yet be statistically explained, and the undrained-soil uncertainty remains largely driven by a single methane hotspot whose representativeness is unknown. They argue that expanding CO2 datasets linked to stand variables and quantitative soil properties offers strong potential for future site-specific assessments of soil carbon balance across stand development stages, and that stratification by drainage status and dominant tree species provides a practical first-order improvement for emission upscaling in the Dfb zone.

Subject of Research: Greenhouse gas emission factors for forest organic soils in the Köppen–Geiger Dfb climate zone

Article Title: Synthesis of greenhouse gas emission factors for forest organic soils in the Köppen–Geiger Dfb climate zone

Article References: Butlers, A., Bārdule, A., Līcīte, I., Kamil-Sardar, M., & Lazdiņš, A. (2026). Synthesis of greenhouse gas emission factors for forest organic soils in the Köppen–Geiger Dfb climate zone. Biogeosciences, 23(20), 7559-7575. https://doi.org/10.5194/bg-23-7559-2026

Image Credits: AI Generated

DOI: 10.5194/bg-23-7559-2026

Keywords: greenhouse gases, organic soils, peatlands, carbon dioxide, methane, nitrous oxide, forest drainage, emission factors, Köppen–Geiger Dfb, Baltic States, carbon balance, IPCC defaults