Carbon dynamics of past terrestrial and marine ecosystems
The present day marine environment in the Baltic Sea is affected by multiple forcing factors, such as land use and climate change, changes in fisheries, urbanization and pollution, which have caused an alarming deterioration of the status of coastal and marine ecosystems. Efforts during the last decades to reduce nutrient loads to mitigate the harmful effects of eutrophication have so far not been rewarded with a healthier Baltic Sea, probably due to the multiple stressors effect (Duarte et al, 2009).
Past periods of hypoxia have often been attributed to climatic variations, disregarding the potential effects of changes in land cover and of preindustrial human impact (Andrén et al, 2000; Leipe et al, 2008), but other studies suggest that even preindustrial land use change and agricultural activities have affected aquatic ecosystems by altering organic carbon and nutrient cycles and contributing to hypoxia (Zillén & Conley, 2010). These effects may have been enhanced by interactions with climate and other environmental stressors.
Forest clearance and land use alters carbon storage in vegetation and soil and can lead to an increased quantity and mobility of dissolved organic matter in surface waters where it can cause brownification and contribute to oxygen depletion (Erlandsson et al, 2008, Bragée 2013). Expanding agriculture has also lead to increased erosion and input of other nutrients to the aquatic system, which has caused eutrophication of lakes (Bradshaw et al. 2006) and may have contributed to environmental changes in coastal waters (Filipsson & Nordberg, 2010; Polovodova et al, 2011, 2013) and potentially triggered hypoxia in the Baltic Sea (Zillén et al, 2008; Zillén and Conley, 2010).
The project intends to utilise the potential of newly developed methods for pollen based landscape reconstruction and dynamic ecosystem modelling to widen the spatial scale to the entire Baltic Sea catchment, and the temporal scale to the whole period of agricultural impact in the region (i.e. 6000 years).
We will apply models for quantitatively reconstructing regional vegetation to existing, well dated pollen records from lakes and bogs in the Baltic Sea catchments. The estimates of past land usethus derived from the pollen records can be incorporated into the dynamic vegetation model LPJ-Guess to assess the effects of land use on ecosystem properties, including terrestrial Carbon pools (Olofsson, 2013; Olofsson et al, 2013; Nielsen et al, in press). The methodology for including different land use/land cover classes (such as grassland, cropland, heathland and wetland) in LPJ-Guess are under constant development, and the our research aims at contributing to this model development as well as providing essential input data for applying the method to modeling the past. The reconstructed land use changes and modelled changes in carbon and nutrient exports to the sea will then be related to changes in the marine ecosystem, as reflected in palaeoecological proxies.