Land-use change and greenhouse gas emissions in the tropics : Forest degradation on peat soils

Forest conversion and degradation are important contributors to worldwide anthropogenic greenhouse gas (GHG) emissions. In the tropics, this contribution is disproportionally large and reducing forest conversion and degradation can substantially reduce GHG emissions. If such GHG reduction efforts are driven by some kind of performance-based payment scheme (e.g. REDD+, The Green Climate Fund), an exact quantification of emissions is crucial in order to prevent over- or underestimation of such reduction efforts. However, for the tropics default IPCC Tier 1 emissions factors are generally based on few studies and on short-term measurements, sometimes from other climatic zones and/or different continents. Another source of low accuracy in GHG emission estimates occurs when emission factors for specific tropical land uses are missing and those emissions are not included in national GHG emission budgets. In this thesis I focus on both of these problems by increasing the mechanistic understanding of the effects of forest conversions on GHG emissions in the tropics, and to contribute to the derivation of robust emission factors for land-use change in the tropics.In a meta-analysis I show that tropical forest conversion to cropland significantly increased N2O emissions, irrespective of the region or type of crops. Nitrogen inputs from fertilizers and animal manure are useful proxies for general IPCC Tier 1 approaches in the tropics, while more detailed IPCC Tier 2 and 3 approaches could be extended with soil cultivation-induced soil organic nitrogen mineralization effects during the first years after conversion, as well as soil moisture and nitrogen availability indices.Subsequently, I focus on a land use not studied to date: forest degradation on tropical peat swamp forest in the Peruvian Amazon. Forest degradation on peat consists of the harvesting of female fruit-baring Mauritia flexuosa palms in natural stands. This type of forest degradation was observed throughout three regions that were studied in the Peruvian Amazon. However, the intensity of degradation was not significantly related to soil carbon stock variability between sites. I then conducted a four-year field study to further investigate the impact of palm harvesting on the soil carbon balance. In sites with >80% of fruit-bearing palms harvested (heavy degradation), litter production and composition altered and resulted in less carbon input into the soil. Carbon output, on the other hand, increased in heavily degraded situations due to faster peat and/or litter decomposition. The combined effects of more carbon output and less carbon input turned the soil at the heavily degraded peat swamp forest into a net source of carbon to the atmosphere of -7.1 Mg CO2-C ha-1 yr-1, while it remained at -0.1 Mg CO2-C ha-1 yr-1 in undisturbed sites.In conclusion, I show that forest degradation in tropical peat swamps can cause significant soil carbon dioxide losses – even without drainage or fertilization practices – and I present a first emissions estimate for this specific land use practice that can be included in national GHG emission estimates, if combined with a quantification of the level of degradation and the area where it occurs