Forest management effects on soil carbon storage and soil CO₂ and CH₄ fluxes in high latitude forests

Forests are important storages and sinks of carbon (C) and high latitude forests specifically store the majority of total ecosystem C in soils, and still accumulate C. Thus, when aiming to preserve or enhance the forest C sink in boreal or northern temperature forests, it is vital that soils are not neglected. The forests of the Nordics and Canada are to a large extent managed, but whether soil C storage and accumulation is affected positively or negatively depends on the specific management practice. Forest harvesting is one of the management practices that has the largest impact on soil C storage and accumulation, but application of biochar and nitrogen fertilizer have also been found to affect the soil C cycle. In Paper I of this thesis the effect of clear-cut harvesting on soil C storage is investigated based on a chronosequence approach using National Forest Soil Inventory data from Sweden, Denmark, Norway, Finland and Canada.

In Paper II, the effect of the chamber closure period length on linear regression and HutchinsonMosier (non-linear) based CO2 and CH4 flux rate estimates is tested. We applied both linear regression and Hutchins-Mosier models using nine different chamber closure periods (60–300 seconds, at 30 second increments) and examined the differences in estimated flux rates. Paper II supports Paper III, in which we analyzed soil CO2 and CH4 fluxes across a forest age-management gradient as well as in response to application of biochar and nitrogen fertilizer. The analyses in Paper II and Paper III are based on soil CO2 and CH4 fluxes measured during three consecutive years of the doctoral program (2021–2023).

Clear-cut harvesting led to continuously decreasing forest floor C stocks for 28–35 years, after which stocks remained unchanged for the rest of the chronosequence (up to 53 years after clearcutting) in stands dominated by Picea species, while stocks started to increase in forest stands dominated by Pinus species. Mineral soil C stocks did not change during the chronosequence, except in deeper mineral Podzol soils (~55–65 cm below mineral soil surface) in which the C stock increased throughout the chronosequence. The topsoil (forest floor + 0–20 cm mineral soil) C stock was larger at the end than at the start of the chronosequence in stands dominated by Pinus tree species, but not in Picea dominated stands. In this study we demonstrate that National Forest Soil Inventory data can be used to examine forest management effects on soil C stocks but that it requires large amounts of data. Our key finding is that clear-cut harvesting in high latitude forests can lead to large and persistent decreases in soil C storage.

The length of the chamber closure period had a large effect on the estimated soil CO2 and CH4 flux rates, with the most substantial differences between chamber closure periods observed when fluxes were high. Both linear regression and Hutchinson-Mosier based flux rate estimates decreased as the chamber closure period increased. In addition, linear regression and HutchinsonMosier based flux rate estimates became more similar to each other as the chamber closure period decreased and increased, respectively. This strongly indicates that the most accurate flux rate estimates are achieved by either using linear regression for flux rate estimation when the chamber closure period is kept short (≤60 seconds using a measurement set-up similar to ours), or by using the Hutchinson-Mosier model when the chamber closure period is kept sufficiently long (≥240 seconds using a measurement set-up similar to ours). Our results have implications for the interpretation and comparison of already published flux rates, upscaled budget estimates and process rates as they may be substantially affected by the length of the chamber closure period and the model used for flux rate estimation. We also show that the accuracy of flux rate estimates can be improved with simple methodological adjustments, and sometimes while alleviating logistical 6 and budgetary constraints. The findings of Paper II were used in the flux rate estimations in Paper III.

Soil CO2 and CH4 fluxes were site dependent, with lower CO2 efflux at the clear-cut, matureunthinned and mature-thinned sites than at the thinned and old-unmanaged sites. Possible explanations for these patterns include C input limitations at the clear-cut site and high soil moisture levels limiting oxygen availability at the mature-unthinned and mature-thinned sites. Net soil CH4 consumption was highest at the clear-cut site and lowest at the mature-thinned site which we hypothesize is due to differences in soil moisture levels which regulate O2 availability and diffusion of atmospheric CH4 into the soil. Application of nitrogen fertilizer led to decreased soil CO2 efflux only at the mature-unthinned site, and while it led to a decreased mean net soil CH4 consumption at all sites, this effect was associated with large uncertainties. The effect of biochar was site dependent, resulting in decreased soil CO2 efflux at the clear-cut and thinned sites and positive effect sizes at the mature-unthinned and mature-thinned sites. The mean net soil CH4 consumption increased in response to biochar addition at the mature-unthinned and mature-thinned sites and decreased at the clear-cut and thinned sites, yet these effects were associated with large uncertainties. The largest reductions in the net soil CO2 and CH4 budgets in response to biochar were achieved at the sites believed to be the driest, and likely due to exacerbated soil moisture limitations. Such increased moisture limitations may have negative effects also on tree growth. Thus, trade-offs between the effects on soil greenhouse gas budgets and C sequestration by trees need to be carefully considered before conclusions on the climate change mitigation potential of biochar can be drawn. Our results suggest that application of nitrogen fertilizer and biochar has limited potential for reducing soil CO2 and CH4 budgets in boreal forests dominated by Norway spruce (Picea abies).

In short, this thesis provides evidence of the magnitude and persistence of soil C stock declines after clear-cut harvesting, it provides guidance on how to improve the accuracy of greenhouse gas flux rate estimates and shows that the effects of biochar and nitrogen fertilizer on soil CO2 and CH4 fluxes depend on the site-specific soil environment.