Epsilon Archive for Student Projects

Abstract

In boreal Sweden, continuous-cover forestry (CCF) is increasingly considered as an alternative method to rotation forestry (RF), but it remains unclear how different CCF systems vary in plant carbon inputs, soil carbon losses and carbon pools. This thesis compared four CCF systems near Vindeln: Åheden and Kulbäcksliden checkerboards, Botsmark gap-felling and Botsmark unmanaged forest (set-aside forests stands). The 2025 carbon flux data of litterfall, vascular understory production, soil microclimate, soil heterotrophic respiration (Rₕ) and mineral-soil organic carbon (SOC) were analyzed, as well as the tree increment data of different availability. Krycklan Catchment Study from 2016–2018 provided regional RF carbon cycle references, and 2025 Sveaskog RF inventory near Vindeln provided the tree increment reference. Modellings were used when the direct measured data were unavailable, hence the annual carbon balance was not the observed net ecosystem production.
Central values of current tree volume increment ranged from 4.85 to 7.64 m³sk ha⁻¹ yr⁻¹ among the four study systems. Tree litterfall in harvested gaps was 19.9–89.4% lower than that in retained forest, whereas understory vascular plants did not show consistent aboveground net primary production (ANNP) compensation with this reduced C input. Heterotrophic Respiration (Rₕ) increased non-linearly with soil temperature and was proved strongly associated with litterfall supply in Åheden checkerboard (R²= 0.824, p<0.01). In this checkerboard system, gaps were warmer and wetter, yet temperature-standardized Rₕ at 10°C was 42% lower than that in the forest cells. The 2025 mineral SOC results did not show any gap or forest cell superior mechanisms among all the CCF systems. NEP estimates ranged from 146.7 to 347.0 g C m⁻² yr⁻¹, where Botsmark unmanaged forest (347.0 g C m⁻² yr⁻¹) and Åheden checkerboard (315.3 g C m⁻² yr⁻¹) had the greatest carbon sequestration among all the systems. Tree volume increment estimates of all the four CCF systems fell within the range of 2025 regional Sveaskog RF P10–P90 (3.26–8.29 m³sk ha⁻¹ yr⁻¹), while the carbon increment contained one of the biggest uncertainties in this thesis due to estimated volume-C conversion rate. Krycklan transferred root production, annualized soil respiration, and missing deadwood pools and fluxes (Krycklan deadwood respiration 4.2 g C m⁻² yr⁻¹) all increased the uncertainty in final NEP estimates among all the CCF systems. None of the study system could make the measured carbon fluxes all contributing significantly to their own carbon sink strength while maximizing the tree increment. For these CCF rankings, more comparable results were found in the within-system mechanisms than the cross-systems ones. Hence, I recommend we repeated measuring the complete carbon inputs (e.g. tree inventories), heterotrophic carbon losses, and soil carbon pools in more CCF systems with different time intervals (e.g. observation of SOC changes take decades). I also suggest doing direct measurements (pools and fluxes) of regional roots and deadwood, to make carbon balance and productivity contrast of CCF-RF more reliable in Vindeln or other further studies. Researchers can make more rigorous and practical management implications based on these more detailed experiments.