Long-term changes in land cover and carbon storage in Tanzania, East Africa

• Main Author: Willcox, Simon
• Other Authors: Lewis, S. ; Phillips, O. ; Munishi, P. ; Lovett, J.
• Published: University of Leeds 2012
• Subjects: 577.1
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.581597

The carbon stored in vegetation varies spatially and temporally due to a complex mix of anthropogenic, climatic and edaphic variables. Thus, the success of climate change policy developments such as REDD+ (Reducing Emissions from Deforestation and Forest Degradation) relies heavily on measuring and understanding this variation in the past, present and future. Here, I first analyse the change in forest cover within a 33.9 million hectare tropical study area in eastern Tanzania. I develop both linear and non-linear baselines of deforestation, providing evidence that Tanzanian forest policy has resulted in forest transition. I then present an Intergovernmental Panel on Climate Change (IPCC) ‘Tier 2’ reporting-compliant look-up method to estimate regional carbon storage, and associated 95% confidence intervals (CI). Applying this method to my study area indicates that 1.58 (95% CI: 1.56-1.60) Pg of aboveground live carbon (ALC) was stored across the landscape in the year 2000. Combining these Tier 2-type values with the historical land use/cover data I derived, I estimate that my study area had a total committed carbon release of 0.94 (0.37-1.50) Pg C between 1908 and 2000. However, look-up methods are overly simplistic for heterogeneous landscapes. Using regression equations, including the effects of disturbance, my IPCC ‘Tier 3’ compliant estimate for the same region in the year 2000 is 1.32 (0.89-3.16) Pg ALC. The most influential variables of carbon storage in the region are human, the strongest impact variables being the nature of the local governance regime (land under national control contained only 40-65% of the ALC stored in areas under local control) and historical logging (areas that had previously experienced logging held 51-77% of the ALC of never-logged areas). Throughout, I provide spatially explicit estimates to aid decision-makers who, due to the influence of anthropogenic variables, could significantly affect landscape carbon storage across this important area.