Browsing by Author "Nordblad, Nina"

Soil carbon stocks of arctic regions are globally a remarkable carbon storage and a fundamental component in the global carbon cycle, as they store more than twice the amount of carbon compared to the atmosphere. Environmental controllers and mechanisms behind soil carbon stocks should be comprehensively studied, as there is no clear agreement of the climate change impacts on the sensitive tundra ecosystems and the soil carbon storages. Therefore, there is an urgent need for reliable and extensive soil carbon data to evaluate both local and global impacts of ecosystem changes in tundra regions. Typically, the upscaling of soil organic carbon predictions made in arctic regions has been based totally or partly on vegetation and land cover classifications. This might be one reason why we are lacking a wide view of the potential of using topographical variables in the upscaling of carbon stocks. Even though, the theory has proven a strong indirect relationship between topography and soil properties. Consequently, only variables extracted from a terrain model were used in this study. The aim of the study was to produce as realistic regional soil organic carbon prediction as possible and to investigate the potential of modeling soil organic carbon with topographical variables. Additionally, the variation of soil carbon stocks in relation to the topographical position were examined closely. The landscape scaled subarctic research area located in northern Norway in a mountainous region with relative elevation reaching almost one thousand meters. Hence, the area has a wide range of environmental gradients, which makes it a great area for studying ecological impacts. The research material consisted of field measurements and soil samples of which carbon contents were analyzed in laboratory. Carbon storages were examined against topographical variables extracted from the terrain model using two different multivariate models: generalized additive model (GAM) and generalized boosted regression modeling (GBM). Models were assessed through correlations between observed and predicted values and through model residuals and their root mean square errors (RMSE). Based on the predictive models, soil carbon stocks varied on the research area between 0–34,5 kg C m-2 and the regional mean estimate was 4,2 kg C m-2. These estimates and the regional variation in stocks are in line with earlier inventories made in similar environments. The largest soil carbon stocks were found above the treeline in valleys, at gently sloping hillsides and in local water-logged peatlands. Soil carbon stocks were generally smaller in the mountain birch forest compared to the shrub and heath tundra areas. Local scale variability in carbon stocks were significant and a great portion of total storages was found on a limited area. Above a height of 700 meters, steep topography and harsh climate conditions limits soil formation leaving only barren ground, which explains the low observed carbon storages. Relationships between local topography and soil carbon stocks presented in theory were also recognized in the results of the multivariate models. Absolute height above sea level regulated soil carbon stock especially through the impact on vegetation and temperature conditions. Hence, these mechanisms made up the robust landscape scaled distribution in predicted carbon storages. Instead, soil moisture determined the fine scaled variation. As well, the results indicated soil moistures essential role in soil carbon accumulation. Earlier observations of topographical variables potential in soil carbon prediction modeling are supported by the fairly good models of this research. Nevertheless, large uncertainties are still associated with regional upscaling of soil carbon stocks, which should be paid attention to in future researches to improve the reliability of predictions. A holistic perspective will be necessary to understand the spatial distribution and environmental factors influence on soil carbon storages. Reliable estimations of soil carbon stocks are a key component when determining future climate change impacts and feedbacks as these mechanisms have globally reaching consequences.