Terrestrial ecosystems represent the most important carbon (C) sink with their capacity to store almost three times that of the atmosphere (Trumper et al., 2009). Further, approximately 40% of terrestrial C is stored in tropical forests, sequestering large amounts of carbon dioxide from the atmosphere (Beer et al., 2010; Pan et al., 2011). However, these forests are threatened by high rates of conversion to other land uses, constituting a major source of greenhouse gas (GHG) emissions and contributing to climate change (Fearnside, 2000; Houghton, 2005). The UN initiative, Reducing Emissions from Deforestation and Forest Degradation (REDD+), represents one path aimed at mitigating the impacts of climate change by conserving tropical forests threatened by deforestation or degradation (Day et al., 2013). It aims to reduce carbon dioxide emissions from developing countries through the sustainable management of forests, while providing co-benefits of biodiversity conservation and livelihood support (Danielsen et al., 2011). Accurate carbon stock quantification represents one important step in ensuring the successful implementation of REDD+, as such information is needed for validation and verification of emissions reductions. (Gibbs et al., 2007; Saatchi et al., 2011).
In eastern Madagascar, deforestation is mainly due to slash-and-burn agriculture (Styger et al., 2007), which results in a mosaic of land use types where fallows are prevalent (Nambena, 2003). To address deforestation in one area of this region, many activities have been implemented, including the development of a REDD+ project initiated by the Government of Madagascar in 2008 (Conservation International, 2013). REDD+ demands a precise estimation of the amount of C stored in forest and other land use types in order to accurately calculate, for example, the emissions avoided due to the presence of a REDD+ project (Andriamananjara et al., 2016). This is needed because the contribution of the C pools may vary across the landscape. The majority
of studies to date on C accounting in different forest ecosystems in Madagascar considered separately the C pools, while studies of the estimation of C stock that consider all five C pools are scarce (Andriamananjara et al., 2016; Grinand et al., 2017; Razakamanarivo et al., 2011, 2012). In this chapter, we firstly review a recent study that accounted C stocks in all five pools recognized by the IPCC (2003), including AGB, BGB, litter, deadwood (DW), and SOC. Afterwards, we identify their dynamics across land uses following deforestation.