<p>Soils of Northern Hemisphere forests contain a large part of the global terrestrial carbon (C) pool. Even small changes in this pool can have large impact on atmospheric [CO2] and the global climate. Soil respiration is the largest terrestrial C flux to the atmosphere and can be divided into autotrophic (from roots, mycorrhizal hyphae and associated microbes) and heterotrophic (from decomposers of organic material) respiration. It is therefore crucial to establish how the two components will respond to changing environmental factors. In this thesis I studied the effect of elevated atmospheric [CO2] (+340 ppm, ¹³C-depleted) and elevated air temperature (2.8-3.5 oC) on soil respiration in a whole-tree chamber (WTC) experiment conducted in a boreal Norway spruce forest. In another spruce forest I used multivariate modelling to establish the link between day-to-day variations in soil respiration rates and its δ¹³C, and above and below ground abiotic conditions. In both forests, variation in δ¹³C was used as a marker for autotrophic respiration. A trenching experiment was conducted in the latter forest in order to separate the two components of soil respiration. The potential problems associated with the trenching, increased root decomposition and changed soil moisture conditions were handled by empirical modelling. The WTC experiment showed that elevated [CO2] but not temperature resulted in 48 to 62% increased soil respiration rates. The CO2-induced increase was in absolute numbers relatively insensitive to seasonal changes in soil temperature and data on δ¹³C suggest it mostly resulted from increased autotrophic respiration. From the multivariate modelling we observed a strong link between weather (air temperature and vapour pressure deficit) and the day-to-day variation of soil respiration rate and its δ¹³C. However, the tightness of the link was dependent on good weather for up to a week before the respiration sampling. Changes in soil respiration rates showed a lag to weather conditions of 2-4 days, which was 1-3 days shorter than for the δ¹³C signal. We hypothesised to be due to pressure concentration waves moving in the phloem at higher rates than the solute itself (i.e., the δ¹³C–label). Results from the empirical modelling in the trenching experiment show that autotrophic respiration contributed to about 50% of total soil respiration, had a great day-to-day variation and was correlated to total soil respiration while not to soil temperature or soil moisture. Over the first five months after the trenching, an estimated 45% of respiration from the trenched plots was an artefact of the treatment. Of this, 29% was a water difference effect and 16% resulted from root decomposition. In conclusion, elevated [CO2] caused an increased C flux to the roots but this C was rapidly respired and has probably not caused changes in the C stored in root biomass or in soil organic matter in this N-limited forest. Autotrophic respiration seems to be strongly influenced by the availability of newly produced substrates and rather insensitive to changes in soil temperature. Root trenching artefacts can be compensated for by empirical modelling, an alternative to the sequential root harvesting technique.</p>

Soils of Northern Hemisphere forests contain a large part of the global terrestrial carbon (C) pool. Even small changes in this pool can have large impact on atmospheric [CO2] and the global climate. Soil respiration is the largest terrestrial C flux to the atmosphere and can be divided into autotrophic (from roots, mycorrhizal hyphae and associated microbes) and heterotrophic (from decomposers of organic material) respiration. It is therefore crucial to establish how the two components will respond to changing environmental factors. In this thesis I studied the effect of elevated atmospheric [CO2] (+340 ppm, 13C-depleted) and elevated air temperature (2.8-3.5 oC) on soil respiration in a whole-tree chamber (WTC) experiment conducted in a boreal Norway spruce forest. In another spruce forest I used multivariate modelling to establish the link between day-to-day variations in soil respiration rates and its δ13C, and above and below ground abiotic conditions. In both forests, variation in δ13C was used as a marker for autotrophic respiration. A trenching experiment was conducted in the latter forest in order to separate the two components of soil respiration. The potential problems associated with the trenching, increased root decomposition and changed soil moisture conditions were handled by empirical modelling. The WTC experiment showed that elevated [CO2] but not temperature resulted in 48 to 62% increased soil respiration rates. The CO2-induced increase was in absolute numbers relatively insensitive to seasonal changes in soil temperature and data on δ13C suggest it mostly resulted from increased autotrophic respiration. From the multivariate modelling we observed a strong link between weather (air temperature and vapour pressure deficit) and the day-to-day variation of soil respiration rate and its δ13C. However, the tightness of the link was dependent on good weather for up to a week before the respiration sampling. Changes in soil respiration rates showed a lag to weather conditions of 2-4 days, which was 1-3 days shorter than for the δ13C signal. We hypothesised to be due to pressure concentration waves moving in the phloem at higher rates than the solute itself (i.e., the δ13C–label). Results from the empirical modelling in the trenching experiment show that autotrophic respiration contributed to about 50% of total soil respiration, had a great day-to-day variation and was correlated to total soil respiration while not to soil temperature or soil moisture. Over the first five months after the trenching, an estimated 45% of respiration from the trenched plots was an artefact of the treatment. Of this, 29% was a water difference effect and 16% resulted from root decomposition. In conclusion, elevated [CO2] caused an increased C flux to the roots but this C was rapidly respired and has probably not caused changes in the C stored in root biomass or in soil organic matter in this N-limited forest. Autotrophic respiration seems to be strongly influenced by the availability of newly produced substrates and rather insensitive to changes in soil temperature. Root trenching artefacts can be compensated for by empirical modelling, an alternative to the sequential root harvesting technique.