Ecosystems in the Anthropocene: the role of cropland management for carbon and nitrogen cycle processes

Abstract: Through deforestation and conversion of natural ecosystems to croplands and pastures, have humans released vast amounts of carbon (C) dioxide (CO2) into the atmosphere. Roughly one third of the cumulative anthropogenic emissions until today stem from these practices, the remainder being mostly due to fossil fuel combustion. The extra CO2 has caused an increase in global air temperature, and both the enhanced CO2 as well as the enhanced temperatures can trigger feedbacks by affecting the C cycle in terrestrial ecosystems. Warmer temperatures stimulate decomposition of soil organic matter, such that C stored in soils will be released - thus causing a positive feedback to CO2 and climate. Increased \CO2 stimulates photosynthesis in plants, which fosters CO2 uptake and causing negative feedback to CO2 and climate. While under present-day conditions the land ecosystems take up more CO2 than they release, it is not yet known whether this uptake will persist also in the future since many relevant processes are not yet fully understood. The continuation of this sink is of societal importance: land management for C has become one of the widely discussed climate change mitigation options (e.g., growth of bio-energy, afforestation), while climate change affects societies in manifold ways, ranging from food production, to rising sea levels. Humans have changed the global nitrogen (N) cycle to a similar degree as the global C cycle. Nitrogen is the chief nutrient for photosynthesis and growth of plants. By using fertilisers, especially in the form of mineral N, the input of N to ecosystems has increased drastically over the last 100 years. While this has led to increased crop yields in many parts of the world, the large application of N also had negative effects. Enhanced ecosystem N levels often affect species composition and thus biodiversity on land, and also lead to N leakage which causes algal blooms downstream of agricultural regions. Human activities have also altered the natural N cycle through combustion of fossil fuels, which release nitrous oxides to the atmosphere which are subsequently deposited to land and ocean through rainfall. How the future C and N cycles interact with climate is an important question for managing land sustainably from a production, pollution and climate mitigation perspective. In order to examine and better understand these interactions, global ecosystem models are used, tools that simulate how functioning of natural ecosystems and croplands might change in response to global environmental changes. In this thesis, the development of the land-use change version of the global ecosystem model LPJ-GUESS, will be described, with special focus on the C-N interactions on croplands. LPJ-GUESS is applied , focussing in particular on the interactions of the environment and human land use with crop productivity and the C cycle, including a number of different climate change mitigation strategies on croplands. It was shown that global cropland soils can be managed from a C sequestration perspective without risking large yield losses and at the same time reduce N leaching, if the management is adapted locally. However, these effects were minor, compared to C losses from deforestation and fossil fuel combustion. Overall, the thesis demonstrates a way forward to explore a number of ecosystem services and their spatial and temporal variation, and how state-of-the-art ecosystem models can be adopted to increasingly address questions that bridge between the natural and socio-economic domain.