Soils contain the largest terrestrial pool of carbon (C), but the magnitude and distribution of the soil C sink may be sensitive to climate change. My dissertation aims to identify key processes that mediate patterns of belowground carbon storage across the globe and quantify the effect of environmental perturbations associated with global change on existing soil carbon stocks in peatland ecosystems. Using meta-analysis, I show that the relationship between plant growth, C allocation, and soil nutrient availability varies on a global scale and high-latitude ecosystems allocate >60% of fixed C to belowground structures. As high latitude ecosystems are warming faster than the global mean, the future of this belowground C store is potentially sensitive to climate change. In high latitude ecosystems in particular, I further show that belowground warming increases the rate of peatland carbon dioxide (CO2) and methane (CH4) losses, although CH4 emissions are more sensitive to warming than CO2 emissions, which is likely to shift the nature of greenhouse gas emissions and increase the importance of CH4 as a radiative forcing agent in the near-term. I also use a natural peatland water table gradient to identify the effect of water table reduction on peatland C and N cycling and find that microbial community shifts in C and N demand may attenuate production of C-degrading enzymes and C mineralization in the presence of plant roots and in areas with low water tables. Together, my dissertation work highlights the important role of belowground plant and microbial processes in high latitude ecosystems, and identifies the potential influence of factors associated with global change on belowground C and nutrient cycling.

Abstract Mycorrhizal fungi are important contributors to the functioning of boreal forests, since they act in the bilateral carbon and nutrient transport between above- and belowground parts of the ecosystem. In ectomycorrhizal (ECM) symbiosis of woody host plants, both fungal and plant partners depend on resources provided by the other. A single tree may simultaneously host several ECM fungal partners, which greatly enhance the host's nutrient uptake. At the same time nearly 20% of host primary production is allocated to mycorrhizal fungi. Although fungi depend on host-derived carbon, it is poorly understood how reduced carbon availability, e.g., due to herbivory, affects the ECM fungal symbionts. In this thesis I studied the impact of simulated insect defoliation or mammal browsing on mycorrhizal fungi of boreal woody hosts. Quantitative and qualitative changes in biomass partitioning in different fungal compartments were detected. None of the experiments showed that defoliation or shoot clipping treatments reduced the intensity of ECM colonisation, while treatments often shifted fungal composition towards less biomass producing ECM morphotypes. Above- and belowground diversity in ECM symbionts tended to decrease due to shoot or foliar damage. In addition, in some cases defoliation also reduced fungal biomass in fine roots and decreased ECM sexual reproduction by reducing the number of sporocarps produced. Defoliation induced a similar response pattern in the host and in ECM fungi with a stronger response to increasing severity of treatment (e.g. degree of removed foliage or repeated years of defoliation). This was also confirmed when relating the effects of host and ECM fungal symbionts to defoliation using present and previously published data. The present results suggest that belowground adaptation of boreal trees to the changing environment is mediated by changes in fungal community or biomass partitioning. The lack of response in the intensity of ECM colonisation further emphasises the importance of the symbiosis to boreal trees.