Methane Emissions from Paludified Boreal Soils in European Russia as Measured and Modelled

Spatial or temporal forest–peatland transition zones were proposed as potential hot spots of methane (CH₄) emissions. Consequently, paludified soils are an important component of boreal landscape biogeochemistry. However, their role in the regional carbon cycle remains unclear. This study presents CH₄ fluxes from two forest–peatland transition zones, two wet forest sites and two clear-cut sites which were compared to fluxes of open peatlands and dry forest. The median fluxes measured using the closed-chamber technique varied from ? 0.04 to 12.6 mg m^?2 h^?1 during three climatically different years. The annual mean CH₄ emissions of the forest–peatland transition zone were significantly lower than the fluxes of the open peatland sites, 7.9 ± 0.5 and 21.9 ± 1.6 g m^?2a^?1, respectively. The dry forest site was characterized by a small uptake of CH₄ (? 2.3 ± 0.2 g m^?2a^?1). Although clear-cut forest area drastically increased in European Russia during the last two decades, if water level depths in these forests remains below 10 cm they do not act as CH₄ sources. Fluxes of CH₄ from the transition zone sites showed a higher response to soil temperature than to water table level. Fluxes of CH₄ between the atmosphere and the two investigated peatlands were not significantly different, although a significant difference in water table level could be observed. The meteorological conditions of the investigated summers changed from being hot and dry in 2013 to cold and wet in 2014; the summer of 2015 was characterized as warmer and drier in the first half and colder and wetter in the second half. Significant differences in CH₄ fluxes were measured only between 2014 and 2013. Significant differences in CH₄ fluxes and in nonlinear regressions showed that the CH₄ fluxes of the different site types such as dry forests, transition zones and open peatlands need to be modelled separately on a landscape level. Obviously, underlying processes vary with the ecosystem and (along with regional aspects) have to be understood first before large-scale modelling is possible.